Perspective of content providers Barbora Benediktova Lukas Nevosad Luleå University of Technology Master Thesis, Continuation Courses Electronic… [630712]

2008:003
MASTER'S THESIS
Affiliate Marketing
Perspective of content providers
Barbora Benediktova
Lukas Nevosad
Luleå University of Technology
Master Thesis, Continuation Courses
Electronic Commerce
Department of Business Administration and Social Sciences
Division of Industrial marketing and e-commerce
2008:003 – ISSN: 1653-0187 – ISRN: LTU-PB-EX–08/003–SE

Affiliate Marketing
Perspective of content providers

Barbora Benediktová
Lukáš Nevosád

e-MBA program
Department of Business Administration and Social Sciences
Luleå University of Technology
Supervisor: Rickard Wahlberg
January, 2008

ii
ABSTRACT
This thesis depicts how content providers (aka publ ishers or affiliates) use affiliate
marketing, a performance oriented internet marketin g. Three different content providers
were interviewed in order to find out detailed info rmation about their usage of affiliate
marketing.
The thesis identifies main advantages and disadvant ages of affiliate marketing for
content providers and specifies conditions, under w hich affiliate marketing is more
beneficial than other types of online advertising. Moreover, the process of selecting
affiliate marketing program by content providers is described.
This paper concludes in a model of a typical affili ate marketing usage. Furthermore, the
work proposes using different terminology for conte nt providers, because it was found
out that the term content provider does not cover a ll aspects of the business any more, as
affiliate marketing can even be used without provid ing any content. Finally,
recommendations for companies, that are planning to start a new affiliate program, are
also added and the work suggests several topics for further research in this field.


iii
PREFACE
There are many people who helped us a lot during st udying and writing this thesis and
deserve to be named here. Our very special Thank yo u! goes to the supervisor of our
thesis, Rickard Wahlberg, who was always very helpf ul, provided us with excellent
support and overall lead us quickly and efficiently to the finish line.
We would also like to thank Petr Sršeň, Thomas Wilk inson and Jan Petr for their
willingness to share their experiences and knowledg e about affiliate marketing.
And finally we are thankful to our families and fri ends that supported us, kept our
thinking positive and altogether made our stay in L uleå, Sweden really special.


iv

"Half the money I spend on advertising is wasted;
the trouble is I don't know which half."

John Wanamaker axiom


v
TABLE OF CONTENTS
1 INTRODUCTION ………………………………………….. …………………………………………… ………………. 1
1.1 Emergence of Cybermediaries and Content Providers . …………………………………………… 1
1.2 Online Advertising ………………………….. …………………………………………… ………………………………… 3
1.3 Affiliate marketing …………………………. …………………………………………… …………………………………. 4
1.4 Problem Discussion…………………………… …………………………………………… ……………………………… 6
1.4.1 Research Purpose…………………………….. …………………………………………… ……………………….. 6
1.4.2 Research Questions ………………………….. …………………………………………… ………………………. 7
2 LITERATURE OVERVIEW ………………………………………….. …………………………………………… … 9
2.1 On-line Advertising …………………………. …………………………………………… ……………………………….. 9
2.1.1 Affiliate marketing …………………………. …………………………………………… ………………………. 10
2.1.2 Advantages of Online Advertising ……………… …………………………………………… ……….. 11
2.1.3 Disadvantages of Online Advertising …………… …………………………………………… …….. 13
2.1.4 Types of Online Advertising ………………….. …………………………………………… ……………… 14
2.1.5 Traffic Acquisition by Content Providers ………. …………………………………………… ….. 17
2.1.6 Compensation Models of Online Advertising ……… ………………………………………….. 18
2.2 Business Models of Content Providers ………….. …………………………………………… …………… 22
2.3 Affiliate program selection ………………….. …………………………………………… ……………………….. 25
2.3.1 Buying process ……………………………… …………………………………………… ………………………… 25
2.3.2 Buying situations …………………………… …………………………………………… ……………………….. 26
2.3.3 Supplier Selection Process (De Boer) ………….. …………………………………………… ……… 27
2.3.4 Supplier Selection and Supplier Management System ( Lee) ……………………….. 29
2.4 Summary ……………………………………. …………………………………………… ……………………………………. 30
3 METHODOLOGY ………………………………………….. …………………………………………… …………… 32
3.1 Research purpose ……………………………. …………………………………………… …………………………….. 32
3.2 Research strategy …………………………… …………………………………………… ……………………………… 32
3.3 Selection of cases ………………………….. …………………………………………… ……………………………….. 33
3.4 Data collection …………………………….. …………………………………………… ………………………………….. 33
3.5 Data analysis ………………………………. …………………………………………… …………………………………… 34
3.6 Validity and reliability……………………… …………………………………………… ……………………………. 34
4 DATA PRESENTATION ………………………………………….. …………………………………………… …. 36
4.1 Case no. 1………………………………….. …………………………………………… ……………………………………… 36
4.1.1 Examples of websites: ……………………….. …………………………………………… ………………….. 36

䜘/uni0641
vi 4.1.2 Research Question 1 …………………………. …………………………………………… …………………… 39
4.1.3 Research Question 2 …………………………. …………………………………………… …………………… 40
4.1.4 Research question 3 …………………………. …………………………………………… ……………………. 42
4.2 Case no. 2………………………………….. …………………………………………… ……………………………………… 43
4.2.1 Examples of websites: ……………………….. …………………………………………… ………………….. 43
4.2.2 Research Question 1 …………………………. …………………………………………… …………………… 44
4.2.3 Research Question 2 …………………………. …………………………………………… …………………… 45
4.2.4 Research question 3 …………………………. …………………………………………… ……………………. 46
4.3 Case no. 3………………………………….. …………………………………………… ……………………………………… 48
4.3.1 Examples of websites: ……………………….. …………………………………………… ………………….. 48
4.3.2 Research Question 1 …………………………. …………………………………………… …………………… 50
4.3.3 Research Question 2 …………………………. …………………………………………… …………………… 51
4.3.4 Research question 3 …………………………. …………………………………………… ……………………. 52
5 DATA ANALYSIS ………………………………………….. …………………………………………… ………….. 54
5.1 Case no. 1………………………………….. …………………………………………… ……………………………………… 54
5.1.1 Research Question 1 …………………………. …………………………………………… …………………… 54
5.1.2 Research Question 2 …………………………. …………………………………………… …………………… 55
5.1.3 Research question 3 …………………………. …………………………………………… ……………………. 58
5.2 Case no. 2………………………………….. …………………………………………… ……………………………………… 60
5.2.1 Research Question 1 …………………………. …………………………………………… …………………… 60
5.2.2 Research Question 2 …………………………. …………………………………………… …………………… 61
5.2.3 Research question 3 …………………………. …………………………………………… ……………………. 63
5.3 Case no. 3………………………………….. …………………………………………… ……………………………………… 65
5.3.1 Research Question 1 …………………………. …………………………………………… …………………… 65
5.3.2 Research Question 2 …………………………. …………………………………………… …………………… 66
5.3.3 Research Question 3 …………………………. …………………………………………… …………………… 68
5.4 Cross-case Analysis …………………………. …………………………………………… …………………………….. 70
5.4.1 Research Question 1 …………………………. …………………………………………… …………………… 70
5.4.2 Research Question 2 …………………………. …………………………………………… …………………… 72
5.4.3 Research Question 3 …………………………. …………………………………………… …………………… 76
6 CONCLUSIONS ………………………………………….. …………………………………………… ……………… 78
6.1.1 Research Question 1 …………………………. …………………………………………… …………………… 78
6.1.2 Research Question 2 …………………………. …………………………………………… …………………… 79
6.1.3 Research Question 3 …………………………. …………………………………………… …………………… 80
6.1.4 Other conclusions …………………………… …………………………………………… ……………………… 80

䎸/uni0639
vii 7 REFERENCES ………………………………………….. …………………………………………… ……………….. 83
APPENDICES ………………………………………….. …………………………………………… ……………………….. 86
Interview Guide …………………………….. …………………………………………… …………………………………………… 87
Screenshots ………………………………… …………………………………………… …………………………………………… …. 90


viii
LIST OF IMAGES
Figure 1 – Basic model of online advertising (Adapt ed from Chaffey, 2003:332) …………………… …………………………..4
Figure 2 – Overview of online advertising and its r elationship to affiliate marketing. Based on
categorization by Zeff, 1999:26-58, Chaffey, 2003 a nd Janoschka, 2004:51-72 …………………….. …………………………. 15
Figure 3 – Buying process (adapted from Palmer, 200 1) ………………………………………… …………………………………………… .. 25
Figure 4 – Evaluation process model for services, F isk, 1981 ………………………………….. …………………………………………… 26
Figure 5 – Model of supplier selection and supplier management system, Lee et al., 2001 ………….. ……………….. 30
Figure 6 – Case 1: letenkar.cz (larger image availa ble in thesis appendix) ……………………… …………………………………. 36
Figure 7 – Case 1: eurovikendy.cz (larger image ava ilable in thesis appendix) …………………… ……………………………. 37
Figure 8 – Case 1: i-noviny.cz (larger image availa ble in thesis appendix) ……………………… …………………………………. 38
Figure 9 – Case 1: birmingham.pl (larger image avai lable in thesis appendix)…………………….. …………………………… 38
Figure 10 – Case 1: manchester.cz …………….. …………………………………………… …………………………………………… ………………………. 39
Figure 11 – Case 2: loanexplorer.co.uk (larger imag e available in thesis appendix) ………………. ……………………….. 43
Figure 12 – Case 2: remortgageexplorer.co.uk (large r image available in thesis appendix) …………. ………………. 43
Figure 13 – Case 2: debtexplorer.co.uk (larger imag e available in thesis appendix) ………………. ………………………. 44
Figure 14 – Case 2: shoppingday.co.uk (larger image available in thesis appendix) ……………….. ………………………. 44
Figure 15 – Case 3: jaknaweb.com (larger image avai lable in thesis appendix) ……………………. ………………………….. 48
Figure 16 – Case 3: schuti.cz (larger image availab le in thesis appendix) ………………………. ………………………………….. 49
Figure 17 – Case 3: vydelek-online.cz (larger image available in thesis appendix) ……………….. ………………………… 49
Figure 18 – Case 3: nakuptesi.cz (larger image avai lable in thesis appendix) ……………………. ……………………………… 50
Figure 19 – Case 1: Criteria for selecting affiliat e marketing provider in Lee’s model …………… ……………………….. 59
Figure 20 – Cross-case analysis: Monitoring the mar ket – Lee's model …………………………… ………………………………….. 76
Figure 21 – Role of content providers in affiliate marketing ………………………………….. …………………………………………… … 81
Figure 22 – Case 1: letenkar.cz ………………. …………………………………………… …………………………………………… …………………………… 90
Figure 23 – Case 1: eurovikendy.cz ……………. …………………………………………… …………………………………………… ……………………… 91
Figure 24 – Case 1: i-noviny.cz ………………. …………………………………………… …………………………………………… …………………………… 92
Figure 25 – Case 1: birmingham.pl …………….. …………………………………………… …………………………………………… ……………………… 93
Figure 26 – Case 1: manchester.cz …………….. …………………………………………… …………………………………………… ………………………. 94
Figure 27 – Case 2: loanexplorer.co.uk ………… …………………………………………… …………………………………………… …………………… 95
Figure 28 – Case 2: remortgageexplorer.co.uk……. …………………………………………… …………………………………………… ………….. 96
Figure 29 – Case 2: debtexplorer.co.uk ………… …………………………………………… …………………………………………… ………………….. 97
Figure 30 – Case 2: shoppingday.co.uk …………. …………………………………………… …………………………………………… ………………….. 98
Figure 31 – Case 3: jaknaweb.com ……………… …………………………………………… …………………………………………… ……………………… 99
Figure 32 – Case 3: schuti.cz ………………… …………………………………………… …………………………………………… …………………………… 100
Figure 33 – Case 3: vydelek-online.cz …………. …………………………………………… …………………………………………… …………………. 101
Figure 34 – Case 3: nakuptesi.cz ……………… …………………………………………… …………………………………………… ………………………. 102


ix
LIST OF TABLES
Table 1 – Characteristics of Buying Situations (Far is, Robinson, 1967) …………………………. …………………………………… 27
Table 2 – Case 1: Advantages and issues with affili ate marketing ………………………………. ……………………………………….. 54
Table 3 – Case 1: Business model compared to affili ate marketing revenue share ………………….. ………………………. 56
Table 4 – Case 1: Business model compared to target market size ……………………………….. …………………………………….. 56
Table 5 – Case 1: Advantages and disadvantages of r espective commission models ………………….. …………………… 57
Table 6 – Case 1: Selection process of an affiliate program …………………………………… …………………………………………… …. 58
Table 7 – Case 2: Advantages and disadvantages of a ffiliate marketing ………………………….. …………………………………. 60
Table 8 – Case 2: Business model compared to affili ate marketing revenue share ………………….. ………………………. 61
Table 9 – Case 2: Business model compared to target market size ……………………………….. …………………………………….. 62
Table 10 – Case 2: Advantages and disadvantages of respective commission models …………………. …………………. 62
Table 11 – Case 2: Selection process of an affiliat e program ………………………………….. …………………………………………… … 63
Table 12 – Case 3: Advantages and disadvantages of affiliate marketing …………………………. ………………………………… 65
Table 13 – Case 3: Business model compared to affil iate marketing revenue share …………………. …………………….. 67
Table 14 – Case 3: Business model compared to targe t market size ………………………………. …………………………………… 67
Table 15 – Case 3: Advantages and disadvantages of respective commission models …………………. …………………. 68
Table 16 – Case 3: One-to-one and one-to-many affil iate programs ………………………………. ……………………………………. 68
Table 17 – Case 3: Selection process of an affiliat e program ………………………………….. …………………………………………… … 69
Table 18 – Cross-case analysis: Advantages of affil iate marketing ……………………………… ……………………………………….. 70
Table 19 – Cross-case analysis: Target Market relat ed to affiliate marketing revenue share ……….. ……………….. 72
Table 20 – Cross-case analysis: Compensation models ………………………………………….. …………………………………………… ….. 74
Table 21 – Cross-case analysis: Advantages and disa dvantages of respective affiliate marketing program
types ……………………………………… …………………………………………… …………………………………………… …………………………………………… …….. 75

INTRODUCTION
– 1 –
1 INTRODUCTION
Although the Internet existed since the end of 1960 s, it was the introduction of World
Wide Web (WWW), a new media for publishing information, whic h started its rapid
growth (Chaffey, 2003:19, Hoffman et al., 1995). Si nce that, the Internet became the
fastest spreading media in human history. While it took long 38 respectively 13 years for
radio and television to reach 50 million users mile stone, the Internet reached the same in
just five years (Zeff, 1999:5).
The power of WWW was in the ability to bring easy-t o-use media, which allowed
consumers find information using a network of web sites (Chaffey, 2003:19). However,
creating a website is only the first step to establ ish company's presence online, which
needs to be followed by attracting Internet users t o visit the website (Rowley, 2004).
It did not take long since the World Wide Web was i ntroduced before the companies
realized the marketing possibilities of the Interne t (Zeff, 1999:1). Chaffey notes that from
the merchants' point of view the Internet brings ma ny opportunities, but also threats.
Companies can take advantage of new technologies an d improve their competitiveness,
but on the other hand, they also need to face new s tart-up companies, some of them with
purely virtual operations. Firms have to adjust the ir businesses to the new trends,
otherwise they will not be able to survive (Chaffey , 2003:xi). Sarkar et al. argue that one
such trend is emergence of new and diverse intermed iaries in the value chain of the
existing companies (Sarkar, Butler, Steinfield, 199 5).
1.1 Emergence of Cybermediaries and Content Providers
With the rapid development of the Internet starting in 1990s, many authors (such as
Chaffey, 2003:52; Hoffman, Novak, Chatterjee, 1995) argued that the introduction of the
Internet would lead to disintermediation – eliminat ing intermediaries in the value chain
between manufacturers and end customers. Jacobs cl aims that in the new economy, due
to the new communication opportunities, distributio n channel is moving from two or
three intermediaries to zero or one intermediary al lowing both cost savings and
customization (Jacobs, 2002).
However, it was demonstrated that even if the Inter net can decrease the number of
intermediaries in some fields, companies can still exploit their expertise and adjust their
activities to the new conditions in order to stay c ompetitive at the market (Giaglis, Klein,
O’Keefe, 2002).
Sarkar et al. remark that the Internet offers oppor tunities for cybermediation –
emergence of new types of intermediaries operating at electronic markets (Sarkar et al.,
1995). Cybermediaries , or sometimes also called infomediaries (Jacobs, 2002), are
“organizations that perform the mediating tasks in the world of electronic commerce”
and it is anticipated that their influence will inc rease (Sarkar et al., 1995).

INTRODUCTION
– 2 – Strauss, Adel and Frost (2006) define cybermediary as "an online organization that
aggregates and distributes information". Online cyb ermediaries act as the middle-man
between the buyer and the seller.
As Jacobs points out, cybermediaries introduced new possibilities of sales channel
management and have become a potential competitive advantage for the companies
cooperating with them, but on the other hand, they can also present a threat in the
means of introducing new level of complexity. Thus, managers need to understand new
methods in electronic channel management in order t o be able to produce adequate
decisions (Jacobs, 2002).
According to Sarkar et al., cybermediaries offer fo llowing services to the customers:
/square4 Search and evaluation – facilitating choice of reta iler, product or service.
/square4 Needs assessment and product matching – identifying customers’ needs and
selecting the right product or service for them.
/square4 Customer risk management – decreasing the level of perceived risk.
/square4 Product distribution – physical distribution of the products or services. (Sarkar et
al., 1995)
Also merchants benefit from the cybermediaries’ ope ration in different ways and they
purchase cybermediaries’ services such as:
/square4 Product Information Dissemination – informing custo mers about new products,
services, special features etc.
/square4 Purchase Influence – intermediaries can influence p urchase decisions to a great
extent.
/square4 Provision of Customer Information – providing detai led information about
customers and their buying behavior.
/square4 Producer Risk Management – reducing producers’ risk connected with selling
products or services at electronic market place. (S arkar et al., 1995)
In spite of growing importance of cybermediaries, t heir role in promoting products and
services on the Internet is not explored sufficient ly. OECD states in its report about
impacts of electronic commerce that the role of int ermediaries will be crucial in the
future development and it should be more researched (OECD, 1999). Jacobs also
mentions that little have been written in the area of marketing communication via
cybermediaries and further research in this field s hould be conducted (Jacobs, 2002).
Cybermediaries offering online advertising to merch ants (though links or product
listings) are referred to as content providers (Sarkar et al., 1995). They are defined as
people or companies that “distribute copyright cont ent via the Internet” such as owners
of websites, bloggers or forum members (Eisenmann & Brown, 2000). Content providers
are sometimes also named publishers . (Wikipedia Affiliate Marketing, 2007).
Content providers vary in size. They range from big companies such as magazine and
newspaper publishers, to physical persons, such as blog owners. According to Zeff, both
big and small content providers can make use of onl ine advertising as a revenue source.
Big websites have high-volume audience, whereas sma ll websites usually are able to
deliver niche audience (Zeff, 1999:4-5). Duffy note s that the number of small content

INTRODUCTION
– 3 – providers is very large. They comprise of small com panies and individuals that are
familiar with possibilities that online advertising offers. (Duffy, 2005)
1.2 Online Advertising
There have been made many definitions of the term Internet advertising . A discussion
had been held whether the Internet is another form of traditional advertising or its
nature lies more in direct marketing. Over the year s, the debate has resulted, according
to Zeff, to a compromise: "Internet advertising is the convergence of traditional
advertising and direct response marketing." (Zeff, 1999:12)
Chaffey refers to online promotion as traffic building campaigns – a method to increase
the audience of a website. According to him, advert ising takes place when "advertiser
pays to place advertising content on another websit e" (Chaffey, 2003:331).
As in each industry, companies can be divided into categories based on their role in
business transactions. Zeff mentions three parties that take part in on-line advertising.
They are:
/square4 Sellers
/square4 Buyers
/square4 Infrastructure
(Zeff, 1999:18-21)
Sellers
Sellers are content providers that produce websites with advertising space to sell. Many
of them today are media companies. Because the cont ent providers' inventory is harder
to sell alone, they quickly formed ad networks. Net works offer experienced sales force
and are easier to deal with for advertisers (ibid).
Buyers
Buyers are companies with products or services to p romote. Often, buyers are
represented by agencies, which manage their adverti sing campaigns (ibid).
Infrastructure
Infrastructure comprise of companies that run servi ces to measure and audit campaigns,
targeting and personalization companies, ad managem ent software developers and
providers of technological infrastructure (ibid).
Chaffey describes the simplest model of online adve rtising on figure 1. In the example, Ad
site displays thousand times an ad, on which 10 vis itors click (hence 1% clickthrough).
Those visitors are taken to the Destination site, w here 1 of them actually purchases the
advertised goods. In this case, spending of £20 on displaying the ad resulted in one sale,
thus the advertising cost per purchase was £20 (Cha ffey, 2003:332). Detailed
information about online campaigns efficiency and i ts calculation is described later in
this paper.

INTRODUCTION
– 4 –

Figure 1 – Basic model of online advertising (Adapt ed from Chaffey, 2003:332)
As Zeff points out, first web ads on the Internet c opied the world offline. A part of a
website was assigned to advertisements, which usual ly had rectangular format and did
not take more than about 10% of page space (Zeff, 1 999:1). Later on, website visitors got
used to the standard banners and it was necessary t o employ new types of online
advertising, so that the Internet users would be at tracted (Benway & Lane, 1998).
Hoffman and Novak note that with the growing share of advertising expenditures
allocated in the online advertising, companies star ted to examine, how many visitors,
attracted by the ads on other websites, are finally turned into customers. After counting
all costs, the sum needed to acquire one customer w as in many cases even higher than
his or her average lifetime value. As this situatio n was not sustainable in the long-run, it
was necessary to adopt more efficient forms of cust omer acquisition. Based on this
needs, the concept of affiliate marketing arose (Ho ffman & Novak, 2000).
1.3 Affiliate marketing
According to Gallaugher et al., affiliate marketing is classified as a type of online
advertising, where merchants share percentage of sa les revenue generated by each
customer, who arrived to the company’s website via a content provider. Content
provider, also referred to as affiliate , usually places an online ad (for example a banner or
a text link) at its website. When visitors click at the ad, they are redirected to merchant's
website and affiliation is tracked by a cookie stor ed on visitors’ computers. (Gallaugher,
Auger, Barnir, 2001)
Merchants, within online marketing called advertisers , pay for the content providers’
services only when a visitor coming from their webs ite executes a specified action. Such
action can be a purchase of a product, filling in a form with personal data, subscription to
a newsletter etc. (ibid.).
Duffy points out, that affiliates take the whole ri sk connected with marketing merchant's
products. The concept is simple. If affiliate's mar keting efforts work, affiliate makes

INTRODUCTION
– 5 – money. If they don't, affiliate does not make money and pays opportunity costs. There
are no limitations for an affiliate how much money it can spend or earn (Duffy, 2005).
An agreement between the content provider and merch ant is referred to as affiliate
program . Chatterjee (Chatterjee, 2002) defines affiliate p rogram as
an arrangement between a company and many affiliate firms characterized by
unidirectional linking with the purpose of generating traffic and transact ions similar
to 'instant access to salesforce of thousands.'
Chatterjee describes affiliate marketing as transac tion oriented, without any
commitment to joint future success or exclusivity r estrictions. Affiliate programs require
minimal effort and investment and may be terminated easily. (Chatterjee, 2002).
Employing affiliate marketing is advantageous for m erchants from many perspectives.
Hoffman and Novak claim that by employing affiliate marketing, merchants can let
thousands of independent websites, called also cont ent providers, to display ads for its
products and only pay them when the ad would actual ly lead to a sale. Advertising costs
move from fixed to variable costs, which can facili tate allocating money to advertising
(Hoffman & Novak, 2000):
In the revenue-sharing mode, the price of advertisi ng is a function of the desired
response by the market. Measurable market responses include key marketing
objectives like unit sales, software downloads, qua lified leads, product inquiries, and
so on. Thus, the results-oriented model is the answ er for marketing managers who
are being asked to justify the sums earmarked in th eir budgets for Internet
advertising.
As Hoffman and Novak demonstrated, affiliate market ing enables better targeting of
online advertising which improve their effectivenes s. Content providers have to choose
affiliate programs very carefully, because of the o pportunity cost connected with not
employing competing programs. Therefore, they targe t the advertising even more
precisely than merchants themselves, as otherwise t hey would not get optimal income.
(Hoffman & Novak, 2000).
Recommendation of a product or service on a partner website can create halo effect and
thus encourage the customers to purchase (Gallaughe r et al., 2001). Apart from
increasing the sales, employing the content provide rs for online promotion is also
beneficial for enhancing the reach and creating bro ader exposure (Chatterjee, 2002).
Moreover, through content providers, companies can gain customers that are usually
very difficult to reach and save on online campaign s planning (Hoffman & Novak, 2000).
Gallaugher et al. add that using affiliate marketin g is more cost-effective to the merchants
than other forms of online advertising, because it diminishes the administrative costs
connected with buying advertising. If the program i s managed well, it can enable
advertising on such a great amount of websites that would be otherwise impossible to
acquire (Gallaugher et al., 2001).
Nevertheless, in order to get the exposure of merch ants' ads at high number of websites,
it is crucial to persuade content providers to join the affiliate program and to offer them
sufficient incentives to remain using it (Chaffey, 2003). However, the literature does not
provide answers for further questions about this to pic – such as what are the

INTRODUCTION
– 6 – appropriate incentives or why should content provid ers prefer affiliate marketing to
other forms of revenue sources.
1.4 Problem Discussion
The growth of affiliate marketing has been rapid in recent years. The major online seller
Amazon, as a pioneer of affiliate marketing, has at tracted over 1,000,000 content
providers that have been cooperating with the compa ny since their affiliate program was
introduced in 1996 (Amazon, 2007). Rowley claims th at quarter of Amazon's revenue is
generated by affiliates (Rowley, 2004).
Nowadays, more and more companies have started thei r affiliate programs in order to
efficiently acquire new customers. Affiliate networks have emerged and position
themselves as intermediaries between content provid ers and merchants providing
technical solutions to manage affiliate programs. S uch networks are for example
Commission Junction, Zanox or TradeDoubler. Affilia te networks introduce several new
programs every day (Commission Junction, 2007; Zano x, 2007).
Even though merchants can select some content provi ders and offer them participation
in an affiliate program, it is the content provider s who decide what advertising they will
add at their websites. With the growing number of a ffiliate programs available at the
Internet, content providers have great variety to c hoose from. For example, only in their
Travel sections Commission Junction offered in Sept ember 2007 175 different affiliate
programs (Commission Junction, 2007) and Zanox 350 programs (Zanox, 2007).
These intermediary affiliate networks are valued by the merchants for having a good
access to content providers that can generate desir able traffic to the merchants’ websites
and increase revenues substantially. Importance of affiliate networks is also appreciated
in the market and grows constantly. In 2003 affilia te network Commission Junction was
valued and sold for only $58 million (ValueClick, 2 003). In January 2007, AOL bided $900
million to buy TradeDoubler – Sweden-based affiliat e network of a similar size. This offer
was not accepted by the shareholders who considered it as too low. (Lannin & Bolding,
2007)
1.4.1 Research Purpose
Several studies (such as Hoffman & Novak, 2000; Pap atla & Bhatnagar, 2002) had been
conducted focusing on the companies offering affili ate marketing programs, however
none of the studies was concerned with the perspect ive of content providers. As
demonstrated above, content providers can nowadays choose out of thousands of
affiliate programs and marketing these programs tow ards content providers is crucial. If
the merchants do not attract enough content provide rs into their affiliate programs, their
links and banners will be exposed to fewer customer s than their competitors’, resulting
in losing positions in the sales, brand management and product awareness (Hoffman &
Novak, 2000).
As this work is a pioneering research in this area, we aim to study basic information
regarding the perception of affiliate marketing by content providers. Therefore, the
overall purpose of the thesis is to characterize th e content providers’ usage of affiliate
marketing.

INTRODUCTION
– 7 – 1.4.2 Research Questions
Bruner (2004) claims, that since 2004 it has become more difficult for merchants to
advertise at the websites of attractive content pro viders. With the evolution of new types
of online advertising and growing interest in onlin e advertising, content providers have
more options to choose from (ibid.).
While the advantages of affiliate marketing for mer chants are theoretically defined (for
instance Gallaugher et al., 2001; Hoffman & Novak, 2000; Chatterjee, 2002), benefits for
content providers are not described properly. As Li bai et al. note, content providers
accept risk connected with employing affiliate mark eting (Libai, Biyalogorsky, Gerstner,
2003). However, the question, why and when they pre fer to take the risk instead of using
other types of online advertising, has not been ans wered so far.
In order to cover the research purpose and characte rize the usage of affiliate marketing
by content providers, we need to find out, what per suade content providers to use
affiliate marketing, how their usage starts and fin ally, when they prefer affiliate
marketing to other types of online advertising.
Thus, this work aims to clarify the key benefits th at persuade content providers to prefer
affiliate marketing. As such assessment can be very subjective, this study will focus at
content providers’ perception of these benefits. By perception we mean judgments that
people make when characterizing and evaluating thei r activities (Slovic, 1987).
Moreover, the situations, when affiliate marketing is more beneficial for content
providers, will be studied. The first and second re search questions are therefore
formulated as follows:
Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
According to Duffy, in order to ensure success of t heir affiliate programs, companies and
content providers should create a win-win relations hip that would lead to generating
enough sales at the merchants’ web sites and suffic ient earnings for content providers
(Duffy, 2005). Libai et al. claim, that setting an interesting commission rate and
commission model is crucial for attracting content providers to an affiliate program
(Libai et al., 2003).
On the contrary, Amazon succeeded to persuade a gre at number of content providers to
advertise its products in its affiliate program, ev en though its main competitors (such as
Barnes & Nobles) have implemented the programs as w ell and have been offering even
higher commissions than Amazon (Hoffman & Novak, 20 00). Amazon's basic commission
is 4% from the order amount (Amazon, 2007), whilst Barnes & Nobles provides 5%.
(Barnes & Nobles, 2007).
As we have shown in the example of Amazon, the comm ission rate is not necessarily the
most important factor that plays role in content pr oviders’ selection of affiliate programs.
However, the other factors influencing the selectio n process are not known and have

INTRODUCTION
– 8 – never been researched. Therefore, the thesis intend s to provide an insight about the
content providers’ selection of affiliate programs and answer the following research
question:
Research Question 3
How do content providers select affiliate marketing programs they intend to participate in?

LITERATURE OVERVIEW
– 9 –
2 LITERATURE OVERVIEW
The literature overview depicts available theories and models that will be used for
answering the research questions. First, we will lo ok into affiliate marketing in a broader
scope of online advertising and explain the most co mmon terms used in both online
advertising and affiliate marketing to have an insi ght into both first and second research
question. Furthermore, the advantages and issues co nnected with online advertising will
be described as well as benefits connected to affil iate marketing in order to understand
how content providers can view them and create fram ework for the first research
question.
Then, types of online advertising and their connect ion to affiliate marketing will be
depicted followed by available compensation models that give also an insight about
different ways how to measure effectiveness in onli ne advertising and affiliate
marketing. This will provide a theoretical basis fo r answering the second research
question.
Finally, as a theoretical basis for the third resea rch question, concepts connected with
supplier selection will be described, so that the p rocess of selecting a provider of an
affiliate program can be explicated.
2.1 On-line Advertising
Simply having a company website on the internet, al though it is a form of promotion, is
not considered as online advertising any more (Chaf fey, 2003:310). Instead, advertising
is the mean of attracting customers to that website – a form of traffic acquisition (Chaffey,
2003:324).
For content providers, online advertising is an imp ortant revenue stream. As we will
show later in chapter 2.1.6, affiliate marketing, b ased on result-oriented compensation
methods, is a special form of selling advertising s pace. To understand when it is
beneficial, we need to understand advantages and di sadvantages of other possibilities
that content providers have.
According to Zeff, content providers have following four options how to sell the
advertising space:

LITERATURE OVERVIEW
– 10 – /square4 Direct selling . Sell the advertising space directly through sales representatives.
/square4 Site representation firm . Website hires a company to sell the advertising s pace.
Such companies charge fees around 35-40 percent and often require a minimum
of pageviews, validation and demographics survey in order to represent a
website.
/square4 Ad networks . Ad networks group together several websites in or der to allow
them easier access to advertisers.
/square4 Auctions . Website can offer free advertising space in an au ction.
(Zeff, 1999:239-250)
Furthermore, Gallaugher et al. (2001) add one more option:
/square4 Affiliate marketing .
2.1.1 Affiliate marketing
As was already described in the introduction chapte r, affiliate marketing is an agreement
between a merchant and content providers to promote merchants’ products or services
at their websites. Content providers get paid only if the visitors from their website
execute a specified action (Chatterjee, 2002).
Thus, in affiliate marketing, the task of content p roviders is not only to deliver the
advertisement, but also to persuade users to conver t to customers. Affiliate marketing is
therefore sometimes renamed to performance marketing . (Wikipedia Affiliate marketing,
2007)
According to Libai et al., affiliate marketing can be divided into two groups – One-to-one
affiliate marketing and One-to-many affiliate marke ting. In one-to-one programs, a
merchant signs contract with a chosen affiliate. Th e terms of the contract are negotiable
and do not influence contracts with other affiliate s. This type of the affiliate program is
used for big players at the market that can attract many new buyers to the merchant’s
website. (Libai et al., 2003)
When using one-to-many affiliate programs, the merc hant sets the same conditions for
all affiliates, who can decide, whether they join t he program. One-to-many program is
usually used, when the merchant cooperates with man y affiliates and it would be too
costly to negotiate contract terms with all of them . (Ibid)
According to Chaffey, affiliate marketing is partic ularly beneficial to small websites, as
they would not have the chance of selling to major advertisers otherwise. These websites
can especially take advantage of affiliate networks . (Chaffey, 2003:344)
Companies can decide to build a private affiliate p rogram at their website or to join
third-party networks and use their technological so lutions. (Hoffman & Novak, 2000).
Affiliate network is “a value-added intermediary pr oviding services, including
aggregation, for affiliate merchants and affiliates .” (Marketing Terms, 2007)
Although individual programs offer merchants bigger control and possibility to adjust
the terms to different affiliates, third-party netw orks usually provide latest technological
solutions and reporting tools and have a wide base of content providers. (Oberndorf,
1999).

LITERATURE OVERVIEW
– 11 – 2.1.2 Advantages of Online Advertising
According to Zeff, online advertising has following four advantages:
/square4 Targetability
/square4 Tracking
/square4 Deliverability and Flexibility
/square4 Interactivity
(Zeff, 1999:13)
Targetability
Ability of precise targeting in online advertising is incomparable to the traditional media.
Ads can be targeted according to language, geograph ic region, time of the day and
platform to name few (Zeff, 1999:13).
Newly emerged advertising platforms, such as Google AdWords and AdSense, went far
beyond common targeting and advertise extensive tar geting possibilities as one of their
key features. In Google AdWords, ads can be targete d according to:
/square4 Language of the user
/square4 Location of the user up to the level of a city
/square4 Searched keywords on google.com and its partner web sites, which include AOL,
Netscape Netcenter, Shopping.com, Ask.com etc.
/square4 According to context and topic of other websites.
/square4 Website category or even particular websites using site targeted ads.
(Google AdWords Learning Center, 2007)
Google's AdWords service has received tremendous ac ceptance, boosting Google's
revenues to USD 10.6 billion in 2006 (Google Press Release, 2007) and its stock price
grew five times in three years since Google's IPO i n fall 2004. (Yahoo Finance, 2007)
Affiliate marketing provides even further ways of t argeting. According to Hoffman and
Novak, whereas in traditional online advertising it is the merchant, who decides how to
target the advertising, by employing affiliate mark eting, content providers themselves
asses which merchants and products best suit their audience (Hoffman & Novak, 2000).
Nevertheless, merchants can still decide, which con tent providers they want to cooperate
with. (TradeDoubler, 2007).
Tracking
Zeff states that whereas measurement of advertiseme nt response has always been
difficult in traditional media, with advertising on line it is relatively easy and yet precise.
Advertisers can measure number of times their ad wa s displayed, number of times it was
clicked on, down to factual number of sales leads i t generated (Zeff, 1999:13).
Today, there are software packages available that c an track ad performance. For
example, in Google Analytics, users can not only tr ack web ad performance, but also
relate it to a broad range of factors, including us er location, language or search query. By

LITERATURE OVERVIEW
– 12 – linking the performance directly to the product and its price, advertisers can optimize
their advertising campaigns for best performance (G oogle Analytics Product Tour, 2007).
Affiliate marketing not only provides the above sta ted information, but according to
Hoffman & Novak it also monitors the activity of th e customer after clicking on the ad
through a cookie, that is stored on his or her comp uter (Hoffman & Novak, 2000).
Merchants usually track the customers’ activity for several days or weeks and pay the
commission to the content providers when the sale i s executed during this period
(Zanox, 2007).
Deliverability and Flexibility
As Zeff remarks, on-line advertising is much more f lexible than traditional advertising.
Campaign results can be seen immediately and the ca mpaign can be changed in case the
results do not meet expectations. Furthermore, the campaign is delivered nonstop (Zeff,
1999:14).
Interactivity
Zeff further states that the goal of advertisers is to promote a brand or a product. Online,
consumer can interact with the product virtually, e .g. download a software demo and try
it out. Internet is the media where information see kers can be turned into customers
most easily. (Zeff, 1999:14)
Hoffman and Novak argue that affiliate marketing mo ves this concept even further, as
content providers are paid only when the purchase o r other specified action is executed.
Therefore, it is up to them to seek innovative ways to increase the interactivity and
persuade their visitors to buy an affiliated produc t or service. (Hoffman & Novak, 2000).
Advantages of Affiliate Marketing
There are two sides to which affiliate marketing is beneficiary: Merchants, who employ
affiliate marketing as their marketing strategy, an d content providers – affiliates, who
actually present on their websites merchants' produ cts and services (Hoffman & Novak,
2000). Benefits, that affiliate marketing offers to the merchants, were already discussed
in the introduction chapter, thus we will now only focus on content providers’
perspective.
Affiliate marketing gives content providers new opp ortunities for generating revenues
from their websites and provides them with the oppo rtunity to cooperate with large
number of merchants, to which they would not have a ccess otherwise (Duffy, 2004).
As Papatla and Bhatnagar proved, content providers benefit most from the participation
in the affiliate program, if there is close connect ion between the website and products or
services offered through the program. The connectio n does not only apply to product
types, which should match website orientation, but also to brand perceptions, consumer
loyalty etc. (Papatla & Bhatnagar, 2002).
Moreover, some content providers can participate in affiliate marketing programs,
because they perceive it as a good service for thei r visitors such as providing sale
coupons, updated information about new products etc . (Duffy, 2004).

LITERATURE OVERVIEW
– 13 – 2.1.3 Disadvantages of Online Advertising
Although online advertising is big step forward fro m traditional advertising and
addresses many issues advertisers had to deal with, it still has some shortcomings.
Banner Blindness
Banner blindness is a phenomenon, where users tend to ignore banners and cannot
recall the messages shown to them. According to Ben way and Lane, who first described
the problem, tendency to banner blindness seems to increase if the users are performing
a particular task, whereas is lower if they just su rf the web with no particular reason.
(Benway, Lane, 1998)
Banner blindness could be well addressed in affilia te marketing, as content providers are
those who know their customers best and also know, which advertising would be
effective (Hoffman & Novak, 2000). Moreover, becaus e affiliate marketing campaigns are
paid per action (PPA), it is in best interest of co ntent providers to avoid banner
blindness.
Pricing Models
Hoffman and Novak were already in 2000 writing abou t inadequate pricing models that
were used for online advertising. They argue that p ricing was derived from traditional
advertising media, such as magazines. Instead of th e price being dependant on the
number of readers, online content providers charged for number of visitors, i.e. for ad
impressions (PPI, pay per impression). Although thi s model seemed to look fair, big
advertisers soon realized that the costs they were paying was too high in comparison
with their margins and lifetime value of acquired c ustomers.
In the case study of CDNow.com (now part of Amazon. com), one of pioneers of affiliate
marketing, Hoffman and Novak state that with common prices for on-line advertising
$70 CPM, average click-through-rate of banners of 1 % and conversion rate of 1%, they
were paying about $700 for one acquired customer. N ot many online retailers can allow
such costs. With gross margin about 20%, a customer would have to spend more than
$3500 over lifetime just to pay acquisition costs. After the launch of affiliate program of
CDNow.com, called "Cosmic Music Network", affiliate marketing accounted for 15% of
their customers, which cost only 2% of marketing bu dget. (Hoffman & Novak, 2000)
Nevertheless, as Libai et al. remark, content provi ders can be concerned with the ability
of merchants to persuade visitors coming from the c ontent providers’ websites to
perform the desired action. If the merchants fail, content providers will lose their
commissions, even though they could not have influe nced the conditions (Libai et al.,
2003).
Rise of Prices
According to Bruner, whilst in early years of onlin e advertising the offer of advertising
space highly exceeded the demand for online adverti sing by merchants, the situation has
changed in 2004. Prior 2004, online advertising was considered to be relatively cheap. In
2004, large media buys, especially within some indu stries, caused shortage of advertising
space, which of course ended up in the rise of the online advertising price.

LITERATURE OVERVIEW
– 14 – This spin-up of prices made the advertisers thoroug hly inspect their advertising
spending and implement complex measures how to eval uate the outcomes. Merchants
are aware that some part of their advertising spend ing is generally wasted and they are
searching for advertising models with better effect iveness. On the other hand, the overall
growth of online advertising brings content provide rs wider variety of options how to
sell their free advertising space and it is more di fficult to attract them. (Bruner, 2004)
2.1.4 Types of Online Advertising
Overview of online advertising types
As shown later in chapter 2.2, online advertising i s sometimes virtually the only source of
revenue for content providers. In order to be able to understand their preferences in
online advertising, it is necessary to outline what specific types of advertising content
providers can use.
Zeff describes many means of advertising that were used on the internet since its
beginning and categorized them into three main cate gories:
/square4 Advertising via E-mail
/square4 Advertising via the Web
/square4 Ad Models outside the Box
(Zeff, 1999:26)
Further in this paper "ad models outside the box", such as cursors, screensavers, toolbars
etc. will not be discussed, as these are nowadays e ither obsolete or became a standard
media in advertising via the web.
Chyba! Nenalezen zdroj odkazů. gives a comprehensive view of types of online
advertising and their relation to affiliate marketi ng. Figure shows all forms of online
advertising that can be employed by merchants to ac quire customers. Colored fields take
advantage of a network of content providers, out of which dark green fields show, where
affiliate marketing is applicable.
As indicated below, affiliate marketing can be empl oyed in advertising via e-mail by
inserting links with affiliate reference to the tex t of e-mail or newsletter. In advertising
via the web, affiliate links can be connected to ba nners, textlinks, advertorials as well as
interstitials. Sponsorship and microsites are not s uitable, because visitors do not leave
content providers’ websites and performance based p ayment models are not applicable.
Affiliate marketing can be also applied in search e ngine marketing as described later
(Zeff, 1999:26-58; Chaffey, 2003; Janoschka, 2004:5 1-72).
Fiore remarks, that specific employment of affiliat e marketing are storefronts, where
content providers can sell other company’s products or services at their own branded
website using technological solution provided by th e advertiser. Website visitors do not
have to leave the content provider’s website before finalizing the purchase (Fiore,
2001:122). Storefronts are often referred to as white labels.

LITERATURE OVERVIEW
– 15 –
Figure 2 – Overview of online advertising and its r elationship to affiliate marketing. Based on catego rization by
Zeff, 1999:26-58, Chaffey, 2003 and Janoschka, 2004 :51-72
Advertising via E-mail
According to Janoschka, e-mail is usually the first thing that users check when they
connect to the Internet. In the USA, 93% of interne t users in 2004 used e-mail on a
regular basis. (Janoschka, 2004:45) Although nowada ys e-mail advertising is being
considered as a synonym for unsolicited e-mail, i.e . spam , there are legal and ethical ways
on how to use e-mail as an online advertising media . (Zeff, 1999:26-27)
Note: Some authors including Strauss et al. do not categorize E-mail in online advertising.
Strauss et al. argue that by definition, direct e-m ail messages, even if they contain
multimedia, are considered direct marketing and not advertising. (Strauss et al.,
2006:340-342) However, this polemic is out of the s cope of this paper and e-mail will be
further in the work referred to as a part of on-lin e advertising.
Discussion Lists and Newsletters
One way to use e-mail as an advertising media accor ding to Zeff is the sponsorship of
discussion lists and newsletters . Newsletters are created by websites in a form of online
publication and users opt in to receive them. Conte nt of discussion lists is created by
users by sending a message to the discussion list. Commonly, discussion lists are called
newsgroups . Discussion lists are hosted and moderated by a de dicated server. (Zeff,
1999:27-32)
In both examples, targeting is a key to success. Me mbers of such list are people that are
highly interested in the topic, as they had to take an action (subscribe) to become
members of such a list. Usually, the sponsor messag e can either by textual or in a form of
a banner (similar to website banners) (Ibid).
Direct E-mail
Zeff remarks that direct e-mail can also be used to send advertisements to users. Direct
e-mail must not have a form of spam. The effectiven ess of spam has proven very poor in
the past and because people hate spam, it hurts the brand image too. (Zeff, 1999:32) ONLINE
ADVERTISING
Advertising via E-mail Advertising vie the
Web Ad Models outside the
Box
Discussion Lists
Newsletters Direct E-mail Banners Textlinks
Sponsorships Advertorial
Interstitial Microsties Search Marketing
Search Engine
Marketing Search Engine
Optimization

LITERATURE OVERVIEW
– 16 – European Commission, in its directive "concerning p ersonal data and the protection of
privacy in the electronics communications sector" s tate recommendations on how
member states should act against spam. Article 13 s tates that " The use of … or electronic
mail for the purposes of direct marketing may only be allowed in respect of subscribers
who have given their prior consent." Among other ru les, the user has to be always given
the right to opt-out of the list and not receive an y other unsolicited e-mail from a
particular company. Moreover, it is up to national legislation, if opt-in principle is
applied. In such case, a prior consent of the recip ient is always required prior sending
any unsolicited e-mail. (EU Directive, 2002)
Advertising via the Web
Banners, buttons, interstitials and other common ad vertising media are referred to as
Web ads . According to Janoschka, web ad is a rectangular a rea, located on a page or in
between pages, which displays text and graphics bas ed advertising. When user clicks on
a web ad, he/she gets redirected to a linked target website. (Janoschka, 2004:51)
Banners
Banners are the oldest form of advertising on the I nternet. Banners have the advantage
to blend text and multimedia, such as graphics, ani mation or sound.
Banners come in a variety of sizes, which are not s tandardized, however due to
compatibility with existing website designs, there are some recommended sizes. The
most common size is a fullbanner , sized 468×60 pixels; second most popular is skyscraper
with dimensions 120×600. Other common sizes of bann ers are 234×60, 180×150, and
120×240. (Janoschka, 2004:51)
According to Strauss et al., web ads, especially tr aditional banner formats, are slowly
disappearing in favor of other forms of on-line adv ertising. (Strauss et al., 2006:324)
DoubleClick report states that market favors fewer large formats instead of large
amounts of smaller formats. (DoubleClick, 2005)
Text Links
Surprisingly, the simplest form of advertising, tex t links or text ads , belong to the media
that is most effective in online advertising. Moreo ver, text links are not intrusive and well
accepted by users. (Zeff, 1999:47-50)
Cho et al. write about the topic that text ads were mainly used by search engines, which
placed contextual text ads along the search results . The first successful implementation
appeared on GoTo.com, later renamed to Overture and acquired by portal Yahoo!.
Another search engine, Google, quickly incorporated similar model, named AdWords.
Both companies now control vast majority of the mar ket. (Cho et al., 2005)
Sponsorships
Zeff shows that a website sponsorship goes beyond j ust placing a banner on the website.
Such campaign is usually aimed at brand image rathe r than just to attract visitors to the
advertiser' website. By sponsoring a website, adver tisers choose to merge their brands
with the brands of sponsored websites. If the spons ored brand is trustworthy, customer
trust can be gained too. (Zeff, 1999:50-51)

LITERATURE OVERVIEW
– 17 – Advertorial
According to Zeff, advertorial is an editorial with a clear advertising purpose. Such
advertorials have high response rate, because users tend to trust editorials. This
approach can however backfire if the users would fe el deceived by editorial, resulting in
negative brand image. (Zeff, 1999:51-54)
Interstitials
Interstitials is a common name for pop-ups, e-merci als, over-the page, expandable,
superstitials, screen-interrupts and other modern f orms of multimedia advertising.
According to Zeff, interstitials often take a long portion of the screen, sometimes the
whole screen, disabling the user to view the page u ntil a specific action is taken or the
advertising pop-up times out. Advantage of intersti tials is that the user's attention is
guaranteed and not disturbed by other website conte nt. Relatively complex messages
can be broadcasted by interstitials, including soun d or even video. A button or link is
usually provided allowing the user to skip the adve rtisement. User, however, has no
means of predicting, when the interstitial will app ear again (Zeff, 1999:55-58).
According to some authors including Zeff and Straus s et al., disadvantage of interstitials
is the possibility that the users will be upset by interstitials and will choose not to visit
the server again. Because the format is very intrus ive compared to normal forms of
online advertising, interstitials may lead to negat ive brand image as well. (Zeff, 1999:55-
58; Strauss et al., 2006:333; SPIR, 2007)
Microsites
Microsites, as defined by Chaffey, are a special fo rm of advertising. When user clicks on
the ad, he or she is not taken to the advertiser’s website, but to a sub-page of current
website with some detailed information about the ad vertiser or its products. In this case,
user does not leave the content provider's website. (Chaffey, 2003:332)
2.1.5 Traffic Acquisition by Content Providers
According to Janoschka, there are three ways, how v isitor can access a website:
/square4 Type the address directly
/square4 Use a search engine to find the website
/square4 Use a hyperlink in a form of a text link or web ad
(Janoschka, 2004:44)
Typing directly an address is connected with loyalt y of visitors to the websites and is out
of the focus of this thesis. Using a hyperlink is a type of online advertising that was
already discussed in the previous chapter, while em ploying search engines for traffic
acquisition is slightly different from other forms of online advertising and is referred to
as search marketing.
Search Marketing
As Duffy claims, search marketing is a very special form of online advertising. It is a
technique of marketing a website using search engin es (Duffy, 2005). According to
DoubleClick, this marketing technique has recorded an astonishing growth in past years.

LITERATURE OVERVIEW
– 18 – In 2001, keyword search formed only 4% of total adv ertising budget, but just two years
later, in 2003, it already was 35% and the estimate for 2004 was already 40%
(DoubleClick, 2005). Search marketing comprises sea rch engine marketing and search
engine optimization. (Strauss et al., 2006:328-329) .
Search Engine Marketing
According to TradeDoubler, search engine marketing (SEM) is a term used for keyword
advertising , sometimes also called contextual advertising . In this case, advertisers buy
advertising space on page displayed to the user aft er he/she searched for a specific
terms. For example, a hotel chain can advertise on the word "hotel", or specifically a
hotel in Paris can advertise on a phrase "hotel Par is". (TradeDoubler Keyword
Marketing, 2007)
Although search engine marketing may not seem relat ed to affiliate marketing,
merchants can actually allow content providers to t ake over their own search engine
marketing activities. In this scheme, affiliates th emselves decide on which keywords to
bid, what price to pay and even what message to pro mote in search engines. Perhaps
even more importantly, content providers pay for se arch engine listings themselves and
only receive commission when a sale is realized. (I bid)
Duffy remarks, that when content providers advertis e on Google or Overture and attract
traffics to their websites through SEM, it is cruci al for them to precisely evaluate the
conversion rate and set costs on a pay per click ba sis. (Duffy, 2005)
Search Engine Optimization
Search engine optimization (SEO) is another techniq ue website owners undergo in order
to promote the websites. Unlike search engine marke ting, SEO is based on tuning organic
search results, i.e. altering the website in a way that increases the probability of the
website being displayed in top results for a partic ular keyword (Strauss et al., 2006:328-
330).
According to Duffy, when using SEO, content provide rs optimize the content of their
websites to be attractive for search agents like Go ogle or Yahoo! and rely on traffic
generated by natural search. This approach is slowe r, but does not require funding to
build cash flow and is less risky (Duffy, 2005).
2.1.6 Compensation Models of Online Advertising
Measuring Campaign Effectiveness
This chapter will explain basic terms that are used for measuring online advertising
campaigns and that are important to understand to b ecome familiar with online
advertising compensation models discussed further o n. As in the world of traditional
advertising, cost and effectiveness is an important criterion of online campaigns. Zeff
(Zeff, 1999:Chapter 4) put together a glossary of c ommon terms used in media buys and
effectiveness evaluation, which is in detail supple mented by Chaffey (Chaffey,
2003:333,382-391).

LITERATURE OVERVIEW
– 19 – Ad Views, Impressions, Page Views, Hits
All of the above terms are synonym for number of ti mes an ad is displayed. Exact
definition varies a little according to a particula r website's methodology. Some websites
measure at the point an ad is requested, some measu re at the point the ad is completely
downloaded by user. (Zeff, 1999:98)
Reach
Reach is similar to Ad views, but counting actual p eople instead of just hits. Reach should
therefore be lower than number of ad views, as some people may view the ad multiple
times (Chaffey, 2003:333).
Click-throughs
According to Zeff, click-throughs is the number of times the advertisement is clicked on.
(Zeff, 1999:100).
CTR, Response rate
CTR is the rate between click-throughs and impressi ons, i.e. the number of times an ad is
clicked on, divided by ad views.
CTR = Number of clicks / Number of impressions (Zeff, 1999:157-158).
CPM
Cost per mile, or cost per thousand is a price to p ay for thousand impressions of an
advertisement on a website. This pricing is commonl y used by larger content providers,
whereas smaller sites prefer fixed prices (Zeff, 19 99:157,217).
CPA, CPL, CPC
Cost per action (or cost per lead) is price to pay whenever user does some action. Most
commonly, such agreed action is a click on and ad, therefore CPC – cost per click. (Zeff,
1999) Affiliate marketing uses purely this type of pricing and defines action according to
merchants needs. (TradeDoubler Business Model, 2007 )
CPS
Cost per sale is used especially in affiliate marke ting. Cost for advertising is calculated as
a percentage of a sale value. (Wikipedia Compensati on methods, 2007)
Conversion Rate
Percentage of visitors to company website that actu ally turn out to customers (Chaffey,
2003:397).
Effective Frequency
Effective frequency is described by Chaffey as: "Th e number of exposures or ad
impressions (frequency) required for an advertiseme nt to become effective". The actual
number is believed to be lower than with traditiona l media. (Chaffey, 2003:333)

LITERATURE OVERVIEW
– 20 – Compensation models
This chapter will cover basic models that are used on the internet to price the advertising
as they are described by Chaffey (Chaffey, 2003:337 ) and supplemented from Wikipedia
(Wikipedia Compensation Methods, 2007).
Compensation models are categorized into three grou ps:
/square4 Per Time Period
/square4 Per Exposure / Per View
/square4 Result Based Models
o Per Response
o Per Action
Per Time Period
Paying for the advertisement on a time basis can be compared to hiring. Commonly, the
price is based on a cost per mile (CPM) basis. Adve rtiser can choose whether the ad will
run on the whole site or its section, or will be di splayed whenever a visitor searches for a
specific keyword. (Chaffey, 2003:337; Zeff, 1999:15 9) According to Hoffman and Novak,
this method is not used any more (Hoffman & Novak, 2000).
Per Exposure / Per View / Per Impression
In this model, advertiser and content provider agre e on a payment on a cost per mile
(CPM) basis (Chaffey, 2003:337). This method is als o known under the name pay per
impression (PPI) (Wikipedia Compensation Methods, 2007).
Result-based models
Although content providers can prefer selling adver tisements on a CPM basis, arguing
that because they do not have control over the crea tive, they act solely as suppliers of
viewers, there has been a significant rise of compe nsation models based on results
(Chaffey, 2003:337). According to Chaffey, these co mprise of:
/square4 Per Response
/square4 Per Action
(Chaffey, 2003:337)
Per Response
Response on an online advertisement is measured in clicks, thus the response is click.
This model is also known as Pay per click (PPC) The cost is calculated according to the
number of click-throughs – in this model, costs are measured in cost per click (CPC).
(Wikipedia Compensation methods, 2007)
The most known example of pay per click compensatio n model is Google AdWords. In
Google AdWords, advertiser only pays when the user actually clicks on the ad and lands
on advertiser's website. On the side of content pro viders, CPC pricing is used in Google's
twin service, AdSense. (Google AdWords Learning Cen ter, 2007) As reported by Strauss
et al., according to the keyword popularity, advert isers can spend between 10 to 500
thousand USD per month. (Strauss et al., 2006:328) Cho et al. state that the advertisers

LITERATURE OVERVIEW
– 21 – pay in between 0.01 to 100 USD per click to Google in AdWords program (Cho et al.,
2005).
To counteract unsuccessful ads, Google automaticall y chooses the best ad to display
between many advertisers based on their willingness to pay higher price, their media
click-through rate and allowed daily or monthly bud get. The higher is the amount
advertiser is willing to pay, the better is the ad placement and thus the probability being
clicked on. In other words, Google is selling adver tising in continuous auctions,
separately for each keyword. (Cho et al., 2005; Tra deDoubler Keyword Marketing, 2007)
Per Action
Duffy notes that although payment for clicks was a step forward from the merchant's
point of view, they still do not deliver satisfying results. Often, clicks were not related to
actual sales. Duffy further argues that the only wi n situation for merchant is actually the
sale. (Duffy, 2005)
According to Chaffey, per action based pricing mean s that the payment is calculated from
number of specific actions that were triggered by u sers that clicked a particular ad. The
action can consist of downloading a product sheet, software trial, sales lead or even
purchase. (Chaffey, 2003:337)
Wikipedia lists several methods, where per action c ompensation model is used:
/square4 Pay per sale (PPS) – also called Revenue share – is a method when advertiser pays
a percentage of a sale made by a customer coming fr om content provider's
website. Respective pricing is called Cost per sale (CPS).
/square4 Pay per action (PPA) and commonly used synonym Pay per lead (PPL) – are
terms used for general per action pricing, where ac tion can be anything from
filling out a form to sign up. Respective costs are measured in Cost per action
(CPA) or Cost per lead (CPL). On the contrary to pa y per sale, the cost is fixed in
this model (as sale cannot be measured). Common spe cial forms of pay per action
are pay per call and pay per install, where content provider is paid for customer's
call to advertiser or download and installation of computer software.
(Wikipedia Compensation methods, 2007)
Result-based models, both per response and per acti on, are tricky for content providers
when signing long term contracts. They make it is e asier for them to sell advertisement
space, but on the other hand, content providers lac ks control of actual media that is used
and product that is advertised. If a CTR (click-thr ough-rate) is low, content provider is
paid less for the same than with other compensation methods. (Zeff, 1999:157-159)
Compensation Models in Affiliate Marketing
According to Fiore, in affiliate marketing particul arly, advertisers can choose from three
above described compensation models in order to pay commissions to the content
providers:

LITERATURE OVERVIEW
– 22 – /square4 Pay per sale (PPS) – commission is paid for each sale made by a visitor from the
affiliate website.
/square4 Pay per lead (PPL) – the affiliate is paid after a website visi tor carries out a
specified action – subscribes to a newsletter, fill s in a questionnaire etc. Note: Pay
per lead is also known as Pay per action (PPA)
/square4 Pay per click (PPC) – content providers receive commission for e very click at the
link to advertiser’s website. (Fiore, 2001:125)
According to the AffStat 2007 report cited by Wikip edia, Pay per sale accounts to 80% of
affiliate programs, pay per action to 19% and all o ther models (PPC, PPI) account to only
1%. (Wikipedia, 2007)
The above mentioned three basic plans can be combin ed with other types. The most used
combinations are two-tier and residual programs. Two-tier programs operate at the same
basis as multilevel marketing, when affiliates are paid also for the activities of newly
recruited affiliate. In residual programs, affiliat es get commissions for all purchases
carried out by previously referred visitors for a s pecific time period. (Harris, 2007;
Wikipedia, 2007)
The commission, which merchants pay to affiliates, can be either set as a flat fee – the
affiliate will get the same amount of money for eac h sale, or as a percentage of the sale
value (Hoffman & Novak, 2000).
2.2 Business Models of Content Providers
In order to better understand how content providers position themselves in the online
value chain and profit from their websites, it is c rucial to comprehend their sources of
revenue. Insight into the available business models is needed in order to be able to
assess situations under which different forms of on line advertising are more or less
efficient, which will provide the framework for ans wering the second research question.
Gallaugher et al. list seven revenue streams for on line content providers:
1. Online advertising
2. Subscription fee for online content
3. Online ordering of print publication
4. Syndication / revenue sharing with online services
5. Per-unit charges for online content
6. Online sale of non-content merchandise and services
7. Affiliate marketing
(Gallaugher et al., 2001)
The cited article however only dealt with magazine and newspaper companies, thus
some of revenue streams are not applicable generall y. A better understanding may
provide model is presented by Strauss et al., who h ave categorized business models
applicable to cybermediaries into following categor ies:

LITERATURE OVERVIEW
– 23 – /square4 Brokerage model
/square4 Agent model
/square4 Online retailing
(Strauss et al., 2006:302-308)
Brokerage Model
According to Strauss, in brokerage model, cybermedi ary acts as a provider of
negotiations platform between buyers and sellers. E ither the seller or buyer or both are
charged a fee. Such companies operate on business t o business, business to customer or
customer to customer markets. Usually, they provide many value added services for both
buyers and sellers. Within brokerage model, two mod els are common:
/square4 Online exchange
/square4 Online auction
(Ibid)
Agent Model
As described by Strauss, online agents represent ei ther the buyer or the seller and act in
their interests. Within Agent models, a following s tructure can be recognized:
/square4 Models representing sellers
o Selling agents
o Metamediaries
o Virtual malls
/square4 Models representing buyers
o Purchasing agents
o Reverse auction
o Buyer cooperative
(Ibid)
Selling Agents
Strauss depicts selling agents as cybermediaries th at represent one or more companies
and help them to sell their products. A commission model is usually applied through an
affiliate program. In order to receive a commission , the agent has to refer a user that will
conduct a purchase order or other specified action. Whereas in traditional marketing an
agent usually represents one manufacturer, agents o nline often represent the whole
industry and can offer competing products. (Ibid)
This model is very common within travel industry, w here travel agents represent
airlines, hotels, car rental companies etc. For cus tomers, using such agents is beneficiary,
as they can negotiate better deals and buying proce ss is more convenient. (Ibid)
Another model represents catalog aggregators , which aggregate information about
products into online catalogs from many producers. Customers can then choose

LITERATURE OVERVIEW
– 24 – conveniently from many products and they have insta nt information about price and
availability. This model is used mainly within B2B solutions. In the most advanced level,
catalog aggregators can for example be directly lin ked to vendors' ERP (Enterprise
Resource Planning) systems. (Ibid)
Metamediaries
According to Strauss, metamediary is a website that connects stakeholders related to a
major event, such as purchase of asset or marriage. Metamediaries reduce customer's
information search, assure quality of suppliers and provide relevant information. They
can also facilitate transactions. Merchants, on the other hand, benefit from the traffic to
their website, that metamediaries generate, and by co-branding. Metamediaries are
based on commission models. Strauss et al. stress t hat trust is important in the relation
between the metamediary and customers and that some of them may even choose not to
use web ads in order to remain perceived unbiased. (Ibid)
Virtual Malls
Strauss describes virtual malls as online equivalen t to brick-and-mortar shopping malls.
Similar to the metamediaries model, stores get the advantage of virtual mall traffic and
co-branding. Customers can benefit from having the same shopping cart in all stores and
making on purchase at the end, loyalty programs, gi ft registries, search facilities and
shopping suggestions, recommendations and product r eviews. (Ibid)
Purchasing Agents
Purchasing agents gather orders from individual cus tomers, which remain anonymous in
many cases, and commit large volume orders in order to obtain better price from a
vendor. Basically, in this case the customer is the one who set the price for which he/she
wishes to buy a product. (Ibid)
Reverse Auction
Reverse auction is similar to purchasing agents, wi th the exception that the business is
kept on the individual basis. Customer sets the pri ce for which he/she undertakes to
make a purchase and vendors bid to meet the price o r get at least close enough to
interest the buyer. (Ibid)
Buyer Cooperative
Buyer cooperative is an alternation of purchasing a gents. Again, customers are grouped
in order to lower the price, but in the case of buy er cooperative, there is a set relation
between the amount of buyers and the price. Buyers, who already joined, know at any
time how many other buyers they need to find in ord er to set the price lower. In this
case, buyers themselves do the marketing to find ot her customers. Strauss et al. note
however, that none of the big companies that used t his model were able to make profit so
far. (Ibid)
Online Retailing
Online retailers are synonym for online stores. The y buy goods from vendors and sell
them again to end customers. This area of study is very large and is not covered by this
paper, because it is related neither to affiliate m arketing nor content providers. (Ibid)

LITERATURE OVERVIEW
– 25 – 2.3 Affiliate program selection
Choosing of an affiliate program by a content provi der is not covered at all in the
literature, and there are also no models that are c lose to this problem. Nevertheless, it
can be assumed, that the selection process is very similar to the process of choosing a
supplier of a service.
Even though the content providers do not pay money for the service, they devote part of
their advertising space to the affiliate program an d they lose possibility to sell the space
to other parties. Therefore, there is an opportunit y cost connected with every affiliate
marketing provider that has to be taken into accoun t (Hoffman & Novak, 2000).
Thus, the theory explaining the selection process o f the affiliate marketing programs will
be based on the literature dealing with process of choosing suppliers. First, the buying
process and buying situation will be depicted, in o rder to understand the background of
the supplier selection process. Afterwards, two mod els describing this process will be
discussed, in order to build the framework for answ ering the research question three.
2.3.1 Buying process
Before the purchase decision is made, the decision maker goes through several steps of
buying process (Palmer, 2001).

Figure 3 – Buying process (adapted from Palmer, 200 1)
The process is started by identifying need, which i s a motivator of an action that would
satisfy the need. Then, the information are sought for in order to find the best solution.
Subsequently, the alternatives are assessed and the final decision is made. After the
purchase, the buyer evaluates it.
Fisk (1981) further developed a model of buying pro cess for using in the service sector.
The model divides buying process into three stages – pre-consumption, consumption and
post-consumption stage. During the first stage, a d ecision-maker identifies the solution,
which he or she expects to be the best. In the next stage, the service is consumed and the
actual experience is matched with expectations rais ed in the previous stage. If the Need recognition
Information search
Evaluation
Decision
Post -purchase evaluation

LITERATURE OVERVIEW
– 26 – expectations are not met, it leads to dissatisfacti on. In the last stage, the service is
evaluated and the satisfaction with the product or service increases the motivation to
buy it again in the future or continue using it, wh ilst dissatisfaction leads to opposite
results.

Figure 4 – Evaluation process model for services, F isk, 1981
For the purpose of this work, the evaluation of alt ernatives and decision to buy stages
will be researched, as they are crucial for answeri ng the above mentioned research
questions.
2.3.2 Buying situations
The length and difficultness of the process is depe ndent on the buying situation, in which
it takes place. Faris and Robinson (1967) recognize three types of buying situations, in
which the supplier is selected consequently:
1. New task
2. Modified rebuy
3. Straight rebuy
These buying situations differ in three characteris tics – newness of the problem,
information requirements and consideration of new a lternatives.

LITERATURE OVERVIEW
– 27 – Distinguishing Characteristics of Buying Situations
Table 1 – Characteristics of Buying Situations (Far is, Robinson, 1967)
Type of buying
situation Newness of the
problem Information
requirements Consideration of
new alternatives
New task High Maximum Important
Modified rebuy Medium Moderate Limited
Straight rebuy Low Minimal None
When decision makers encounter the new task, they h ave to deal with situation they
have no experience with and therefore they will nee d to gather a lot of information. They
will also look for and consider many different alte rnatives before making their final
decision. (Ibid)
Buyers dealing with the modified rebuy already have relevant buying experience, but
they need to gain new information about the changed task and some new alternatives
are sought for, in case there is a possibility to i ncrease return by that. (Ibid)
In the straight rebuy, the task is routinized, deci sion makers do not usually seek for new
information nor for new suppliers, even though the specification of the purchase can
differ. These variations however do not cause the c hange of supplier. (Ibid)
Although this concept was identified in 1967, it is still accepted in the literature and
authors (such as De Bore, 2001) consider it as vali d, because of its applicability to wide
variety of problems connected with purchasing and s upplier selection.
De Bore (2001) adds one additional characteristic, which significantly influences the
attention paid to the purchase decision – the level of perceived risk. If the level of
perceived risk is low, then the search for informat ion is not as demanding as in the high-
perceived risk situation. Also the number of consid ered alternatives will be not so high
for low-perceived risk purchases and vice versa.
We suppose, the described buying situations are app licable to selecting a provider of an
affiliate program, because content providers can ei ther decide to employ new programs
for their websites, replace current ones or continu e using them.
2.3.3 Supplier Selection Process (De Boer)
As demonstrated previously, we assume that the proc ess of selection an affiliate program
by content providers is similar to the supplier sel ection process. De Boer states, that
there are four phases of this process:
Problem definition – organization should decide, what they want to ac hieve, whether
they need to employ suppliers to reach the goal and how many of them will be
potentially necessary.
Criteria definition – criteria important for the selecting the supplie r has to be chosen
and defined. The company can use the historical dat a about suppliers, if it is available.
Pre-qualification of suppliers – as there can be many available suppliers, their number
should be reduced to a smaller set of acceptable su ppliers.
Choice of the supplier – the final decision is made and the supplier is s elected from the
previously chosen set of suppliers.

LITERATURE OVERVIEW
– 28 – (De Boer, 2001)
Methods for supplier selection
De Boer (2001) reviewed methods described in the li terature for each of the process
stages and observed, that the literature does not d eal with the problem definition and
criteria definition in a new task situation. Theref ore, the criteria need to be researched
empirically from the content providers in this stud y.
Pre-qualification
Several methods used for pre-qualification of suppl iers are enumerated by De Bore
(2001):
Categorical methods – supplier’s performance in each criterion is asse ssed as positive,
neutral or negative. Subsequently, suppliers are ca tegorized into three groups according
to the previous evaluation.
Data envelopment analysis – compares the cost of costs of using suppliers an d benefits
delivered by suppliers, which enables the buyer to identify, which suppliers would be
efficient to use and which would be inefficient.
Cluster analysis – statistical method, which sorts the suppliers in to clusters, so that the
differences between suppliers in the cluster are mi nimal, whilst differences between
suppliers from different clusters are maximal.
Case-based reasoning systems – use software that provides information and
experiences from previous cases similar to the solv ed one using database of such
situations.
Final Choice
For the final choice of the supplier, De Boer (2001 ) states that there are five groups of
methods:
/square4 Linear weighting models
/square4 Total cost of ownership models
/square4 Mathematical programming models
/square4 Statistical models
/square4 Artificial intelligence models
The last three models are based on historical quant itative data to a great extent and
therefore, they will be not described into details, as they are not applicable to choosing a
provider of affiliate program, when there is no (or little) historical data available.
Linear weighting models assign weight to each criterion. The rating of the criterion is
then multiplied by the specific weight. This way, s ingle number is calculated for each
supplier and the one having obtained the highest nu mber should be selected. Several
modifications of the model have been proposed, such as non-compensatory models,
which set a minimum level for each criterion that h as to be achieved; analytical hierarchy
process, which enables using verbal qualitative eva luation of the supplier’s performance
in each criterion; statistical techniques and the f uzzy sets theory to model uncertainty
and imprecision. (Ibid)

LITERATURE OVERVIEW
– 29 – Total cost of ownership models seek to assess all the costs connected with the pur chase
from a specific supplier and include them into the price offered by the supplier (ibid).
Some authors such as Smytka and Clemens (1993) prop ose to combine this method with
rating system for criteria that are difficult to as sess quantitatively.
2.3.4 Supplier Selection and Supplier Management System ( Lee)
More structured model which integrates supplier sel ection with supplier management
and purchasing strategy was developed by Lee, Ha an d Kim (2001). The authors state
that the supplier selection and supplier management system comprises three parts:
/square4 Purchasing strategy system
/square4 Supplier selection system
/square4 Supplier management system
Purchasing strategy includes four criteria – qualit y, delivery, cost and service, which
were identified as crucial in manufacturing. Purcha sing strategy system influences the
choice of criteria for supplier selection. Supplier management system should identify the
managerial criteria using the information from the supplier selection system. Finally,
supplier selection system should calculate the weig hts of each criteria and rank suppliers
according to them. Offers of alternative suppliers are compared to the performance of
current ones and if surpassing the existing supplie rs, they will replace them. Thus, the
model is applicable especially for modified rebuy b uying situations, when there are new
suppliers considered, but the task is already known and well defined.
For the purpose of the study we will focus only on the depicted supplier selection system,
as the supplier management and purchasing strategy systems are out of the focus of the
thesis.

LITERATURE OVERVIEW
– 30 –

Figure 5 – Model of supplier selection and supplier management system, Lee et al., 2001
2.4 Summary
For content providers, online advertising is an imp ortant source of revenue, and in some
cases even the only source of revenue. Content prov iders today have wide variety of
options how to use online advertising on their webs ites and towards their users (chapter
2.1.4), as well as they have several options how to sell advertising (chapter 2.1) and a
variety of compensation models to set the price for it (chapter 2.1.6). Chapter 2.1.5
further discusses search marketing, the main techni que used to attract visitors to content
providers' websites.
We have drawn a framework to understand possible ad vantages and disadvantages of
online advertising (chapters 2.1.2, 2.1.3), as well as we have covered several forms of
affiliate marketing usage (chapter 2.1.1). Combined with knowledge of applicable
business models and revenue sources (chapter 2.2), we have assessed the situations in
which the usage affiliate marketing is beneficial a nd in which not.

LITERATURE OVERVIEW
– 31 – Furthermore, we have shown two models of buying pro cess (chapter 2.3.1), which can be
used to study how content providers choose affiliat e programs. The process is different
for several buying situations, as shown in chapter 2.3.2. Supplier selection models
presented in chapter 2.3.3 and 2.3.4 further evolve the evaluation and decision stage of
buying process.
The above described models were used in order to fo rm the questions for interview
guide that enabled answering the research questions of this thesis.

METHODOLOGY
– 32 –
3 METHODOLOGY
In this chapter, the research methodology used in t he work is discussed and explained.
First, the choice of research purpose, approach, me thods and strategy are motivated.
Then, used sampling methods are clarified and data collection and analysis are depicted.
Finally, ways to ensure validity and reliability in the research are reviewed.
3.1 Research purpose
This study is a pioneering research, therefore, exp loratory research purpose was applied.
This type of research purpose should be employed wh en the researched problem has not
been precisely defined so far and needs to be clari fied (Ghauri & Gronhaug, 2005). The
work seeks to determine and explain the content pro viders’ usage of affiliate marketing
and provide new information about the problem area.
As the work does not try to “portray an accurate pr ofile of persons, events or situations”
(Saunders, Lewis, Thornhill, 2000:97), which is the goal of descriptive research, nor find
cause-effect relationship among variables and expla in it as in explanatory research (Yin,
1994, 3-9), other research purposes were not applic able.
3.2 Research strategy
Different strategies can be used when carrying out the research in business area. Yin
enumerates them as case study, experiment, survey, history and archival records (Yin,
1994:3). For this work only case study and survey c ould be taken into consideration, as
all other strategies did not enable collecting appr opriate data for answering the research
questions.
The choice of the research strategy is influenced b y the formulation of the research
problem and the research questions. Case study is p referred for research questions
starting with “how” or “why” aiming to provide expl anations about the problem area,
while survey is relevant for research questions sta rting with “who”, “what”, “where”,
“how many” or “how much” focusing on describing the examined phenomenon (Yin,
1994:6). Formulation of research questions in this thesis (starting with “how”)
supported the choice of case study instead of surve y.
Survey enables researching a large sample size in a n economic way, but the data are
usually not as wide-ranging as if they were collect ed using another research strategy. On
the other hand, case study does not allow collectin g data from a large sample size, but
offers to gain deep understanding of the problem an d researching many variables
(Saunders, Lewis, Thornhill, 2000:92-93), which is the goal of this thesis. Because of the
above stated reasons, case study research strategy was employed.
The scope of this paper did not allow collecting da ta from a large sample of content-
providers. Still, conducting only one case study wa s also not suitable, as single case study

METHODOLOGY
– 33 – should be employed, when it represents either extre me or unique case, critical case that
can test a well-formulated theory, or revelatory ca se that provides an opportunity to
research so far inaccessible problem area (Yin, 1994 :38-39).
None of the options applied for this thesis, theref ore multiple case studies were carried
out and compared, as their findings are usually con sidered to be more credible and solid.
Moreover, multiple case studies enable comparison a mong the cases and increase the
research validity (Yin, 1994:45).
3.3 Selection of cases
For selecting cases, probability methods, which are “based on the concept of random
selection – a controlled procedure that assures tha t each element of total population is
given a known nonzero chance of selection” (Cooper & Schindler, 1998), could not be
applied in this work, because it was not feasible t o embrace all the content-providers
using affiliate marketing at their websites. This g roup is very large and it is changing
constantly.
Instead non-probability methods were employed. Thre e content providers were chosen
by the researchers for conducting the case studies. They were chosen by convenience
selection method, which means selecting cases accor ding to their availability and
accessibility for the researchers (Saunders, Lewis, Thornhill, 2000:177), using personal
contacts. The reason for applying this type of sele ction methods was mainly the
availability of such contacts and expected willingn ess to provide the data.
In order to answer the research questions of this w ork, content providers that were
selected for case studies needed to meet several cr iteria. First, they had to own or run
several different websites and use affiliate market ing at least at part of them for one year
or longer. This condition should have ensured that the content providers had experience
with different types of websites, they were knowled geable about affiliate marketing and
were able to assess its advantages and disadvantage s.
Second, they had to employ at least one other type of online advertising at their websites,
so that they were able to compare affiliate marketi ng with other advertising options.
Furthermore, running websites should have been the primary job for content providers,
so that they were highly involved in performance of their websites and generated
revenue and tried to achieve the best results for t hem.
Websites of the selected content providers should h ave been focused at different topics
and should have not used identical affiliate progra ms. Focusing only at one business area
or few affiliate programs could distort results of this work, as they could be dependent
on performance of merchants providing the programs, or having dominant position at
the market.
3.4 Data collection
For the exploratory studies such as this one, the q ualitative methods for data collection
and analysis are appropriate, as they frequently wo rk with a smaller sample size that
enables gaining deeper insight into the problem (Gh auri & Gronhaug, 2005). Due to the
small amount of researched subjects, generalization is usually not possible (Ghauri &
Gronhaug, 2005), however that was not the aim of th is thesis.

METHODOLOGY
– 34 – Moreover, data collected in the research were not s uitable for statistical processing, but
they required categorization and interpretation, wh ich is also characteristic for
qualitative methods (Saunders, Lewis, Thornhill, 20 00:378). Therefore, this study
employed qualitative research methods.
Data were collected by interviews with people respo nsible for choosing, implementing
and evaluation of affiliate marketing in the chosen subjects and by observation at their
websites. Interviews were chosen, as they enable ac quiring in-depth data and are
suitable for qualitative analysis (Saunders, Lewis, Thornhill, 2000:245).
In the study, interviews were conducted by telephon e or during personal meeting. All
interviews were recorded and were semi-structured, which means that list of themes
and questions needed to be covered was prepared, bu t it was possible to make variations
in every interview according to the actual needs (S aunders, Lewis, Thornhill, 2000:246-
247). This way it was possible to compare the inter views, but also adapt their flow to the
new information learnt during the interview.
Questions were asked according to an interview guid e (available in appendix), which
was prepared in advance in order to cover all topic s connected to research questions of
this work. In the beginning, interviewees were enco uraged to describe their websites
and introductory questions were asked. Even though these questions were not directly
connected with any research questions, they were im portant for understanding
website’s focus, visitors, business model etc. If n ecessary, sequence of questions from the
interview guide was changed to keep continuous flow of the interview.
Apart from interviews, observation at the content-p roviders’ websites was applied. Due
to the observation, it was possible to verify data collected at the interviews and find out
additional information about employing affiliate ma rketing and other forms of online
advertising at content providers’ websites.
3.5 Data analysis
Data in multiple-case studies can be investigated t hrough within-case and cross-case
analysis. In the within-case analysis, collected da ta are compared with the theories in the
frame of reference (Huberman & Miles, 1994:143-173) . Within-case analysis was used to
find out similarities and differences between the f indings from the case study and
theoretical background described in the literature overview chapter of this work.
Potential discrepancy between findings and theory w as analyzed and explanation was
sought after. The same structure of within-case ana lysis was used for all three case
studies.
After within-case analysis, cross-case analysis, th at enables confronting cases and
comparing findings from all of them (Huberman & Mil es, 1994:143-173), was used. In
the cross-case analysis, findings from the cases we re compared and discussed.
3.6 Validity and reliability
Two measures are usually taken into account, when t he research quality is assessed –
validity and reliability. Validity is “concerned wi th whether the findings are really about
what they appear to be about” (Saunders, Lewis, Tho rnhill, 2000:101).

METHODOLOGY
– 35 – In order to assure validity, the research methodolo gy and frame of reference for this
study were adhered to. Multiple sources of data wer e used – three content-providers
were interviewed and also their websites were obser ved in order to verify the data
provided by the informants.
External validity – specifying the area, in which t he conclusions from the study can be
generalized (Yin, 1994:35-36) – had to be also taken into consideration because of the
small number of informants. Three content providers , that were interviewed, cannot
represent precisely the whole group of content prov iders on the Internet, which is very
variable and diverse.
The chosen informants were affected by their backgr ounds, type of business models they
had, or opportunities offered at the market(s) they were active at. These factors had also
impact on research reliability – assuring that the same results would be achieved when
conducting another research (Saunders, Lewis, Thorn hill, 2000:101).
For strengthening the reliability of the research, the same interview guide was used for
all interviews, which were recorded and analyzed by two researchers in order not to
miss out important information and to reduce interv iewer bias and errors. To assure that
informants did not omit important information for t he research or altered the facts,
observations were made at their websites before the interviews. Due to this, data gained
by the interviews could be compared immediately wit h data from observations and
potential differences were explained. Content provi ders were also not asked about
sensitive data (such as their income) that they cou ld be unwilling to share.
Still, it is not possible to generalize the finding s and conclusions of this research to all
content providers at the Internet. However, in its settings the research provides in-depth
insight into the problem area and offers new perspe ctive to the companies considering
employing affiliate marketing. The work can be also used as a basis for further studies
about the topic.

DATA PRESENTATION
– 36 –
4 DATA PRESENTATION
4.1 Case no. 1
Interviewee P ETR SRŠEŇ is a Czech content provider, residing in Prague. H e owns around
40 different websites, targeted to Czech, Polish, S lovak and international visitors.
4.1.1 Examples of websites:
Letenkar.cz ( translated “air tickets guide” ) is a Czech website dedicated to air travel.
Website presents new routes of airlines, new bases of low cost airlines, special
promotions and general air travel news.

Figure 6 – Case 1: letenkar.cz (larger image availa ble in thesis appendix)
Following revenue sources are used at letenkar.cz:
/square4 Ad networks (Google AdSense)
/square4 Affiliate marketing (Potenza.cz 1)
Eurovikendy.cz ( translated “Euroweekends” ) is a Czech tourist website with information
about weekend getaways in major European cities. A separate page exists for each
destination (e.g. Paris, Rome etc.) with detailed i nformation about the city, tips to main
sites to see, accommodation, transportation etc. Ac cording to the season, some special
pages are promoted – e.g. St.Valentine's getaways, New Year Eve's getaways etc.

1 Potenza (www.potenza.cz) is the biggest affiliate network in the Czech Republic

DATA PRESENTATION
– 37 –

Figure 7 – Case 1: eurovikendy.cz (larger image ava ilable in thesis appendix)
The website Eurovikendy.cz uses following revenue s ources:
/square4 Ad networks (Google AdSense)
/square4 Affiliate marketing (Invia.cz 2, Potenza.cz)
/square4 Linkbuilding text links (direct selling)
According to the interviewee, affiliate program fro m travel agency Invia.cz is the best
source of revenue for this website.
I-noviny.cz ( translated “i-news” ) is a Czech regional online newspaper. It contains daily
news from region of Česká Lípa, information about t he region, cultural events, discussion
forum and flea market. Website has audience of abou t 1000 visitors per day on average.

2 Invia (www.invia.cz) is the biggest online travel agency in the Czech Republic

DATA PRESENTATION
– 38 –

Figure 8 – Case 1: i-noviny.cz (larger image availa ble in thesis appendix)
I-noviny.cz use following sources of revenue:
/square4 Ad networks (Google AdSense)
/square4 Direct selling of advertising space (banner ads, mo stly from regional companies)
/square4 Affiliate marketing (Invia.cz)
/square4 Linkbuilding text links (direct selling)
According to the interviewee, direct advertising is the main revenue source for this
website.
Birmingham.pl ( http://www.birmingham.pl/ ) is a website targeted to people from
Poland willing to work in Birmingham, United Kingdo m. This website is an example of an
individual affiliate program, agreed between the ag ency that provides jobs and
accommodation and content provider. Website contain s detailed information about the
city of Birmingham, accommodation possibilities, jo b market, administration needed to
move there and start working, transportation, spare time possibilities and online forms
to ask for quote for accommodation and job.

Figure 9 – Case 1: birmingham.pl (larger image avai lable in thesis appendix)
Birmingham.pl solely relies on a custom made affili ate program, with compensation
model pay per sale (PPS).

DATA PRESENTATION
– 39 – Accommodation in Manchester ( http://www.manchester.cz/ ) – this website follows the
same model as accommodation in Birmingham above. We bsite is available in 7
languages.

Figure 10 – Case 1: manchester.cz
Individual affiliate program is used at the website , with compensation model pay per sale
(PPS).
For selling the advertising space at his websites, the interviewee uses direct selling,
advertising networks such as Google AdSense or Skli k 3, affiliate networks such as
Commission Junction, direct affiliate programs of A mazon.com, Invia and Kralovna 4 and
custom made pay per lead agreements.
4.1.2 Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
The interviewee perceives the biggest advantage of affiliate marketing compared to the
other forms of online advertising in the financial gains, which, when implemented well,
can be much higher than gains from advertising netw orks.
On the other hand, according to the interviewee, a strong disadvantage is the difficulties
connected with the implementation. He observed that he needs to devote great amount
of time in order to find the most efficient way, ho w to use an affiliate program at a
specific website, as he needs to assess visitors’ b ehavior and adjust the ads to it.
He has to overcome banner blindness and make the ad vertising more appealing to the
specific group of visitors. Moreover, the behavior of visitors differs according to the
website’s topic. Therefore, it is not possible to a pply the same patterns at the websites,
which focus on diverse groups of visitors.
Except for the described issues with the implementa tion, he thinks that affiliate
programs are easier to use than most of the other t ypes of online advertising. The
interviewee notes that even though direct selling o f advertising can bring higher income

3 Sklik (www.sklik.cz) is an advertising network own ed by the biggest Czech online portal Seznam.cz
4 Kralovna (www.kralovna.cz) is the biggest online s eller of air tickets in the Czech Republic

DATA PRESENTATION
– 40 – than affiliate marketing, it is also more time-cons uming as it requires approaching
possible merchants, persuading them, setting up per sonalized advertising, preparing
reports and handling all paperwork. In affiliate ma rketing, the access to merchants is
straightforward and simple.
The interviewee adds that the simplest way to earn money on content websites is to use
advertising networks, such as Google AdSense. Advan tages are that the whole process is
fully automated and it is not necessary to issue an y invoices, communicate with
advertisers etc. However, these networks do not ena ble gaining such high income as
other forms of online advertising. Because of these reasons, he considers affiliate
marketing as the best compromise between financial gains and overhead costs
connected with selling the advertising.
4.1.3 Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
The interviewee offers several types of online adve rtising to the merchants – sending e-
mails with advertising content to his visitors, dis playing banner ads and text links at his
websites. Only banner ads and text links are used f or affiliate marketing.
He has two main sources of income from his websites – online advertising and affiliate
marketing. He does not consider other sources of re venue as beneficial as these two.
Affiliate marketing accounts for about 50% of his i ncome from the websites, other half is
income from direct advertising and advertising netw orks.
The content provider sets the pricing for online ca mpaigns using three compensation
models – per time period, pay per click and pay per sale. Compensation model per time
period is applied for the banner advertising, when he charges the given amount of money
for 3 or 6 months. He does not offer banner adverti sing for shorter periods of time,
because he does not want to devote time to sending new quotes to the advertisers, issue
more invoices etc.
The interviewee uses pay per click model only with ads provided by advertising
networks Google AdSense and Sklik. These networks d ecrease the amount of time
necessary for administrating the ads, however the f inancial gains are lower than with
other compensation models. The interviewee does not offer pay per click model for
advertising space sold directly, as the price per o ne click would have to be relatively high
in order to pay the administrative costs connected with selling and it would not be
competitive. Pay per click advertising provided by advertising networks is only used at
websites, at which the interviewee does not have be tter options how to sell advertising
space.
In affiliate programs, pay per sale model is applie d. He accepts both flat commission
rates as well as percentage of the sale amount. He prefers percentage commission,
because he considers it fairer than the flat rate. However, he does not perceive this
criteria being important in his decision making abo ut using a specific affiliate program.
From all the above described compensation models, h e prefers pay per time period,
because he knows, how much income it will generate and it also brings him the highest
earnings. On the other hand, it is more difficult t o sell advertising based on this model
and it is more demanding as far as administration i s concerned.

DATA PRESENTATION
– 41 – The interviewee applies agent business models (sell ing agent and metamediary) at his
websites and he thinks that affiliate marketing can be applied to all of his websites.
Nevertheless, he uses it only at part of his websit es, because he considers combination of
banner advertising, text links and affiliate market ing as the most efficient and it makes
him more independent on specific revenue sources.
The interviewee assesses effectiveness of the adver tising campaigns at his websites
using two main measures. First, he measures the cli ck through rate as a ratio between
the number of visitors that click at the banner and the total number of visitors at the
webpage. If click through rate is not satisfactory, he changes placement of the ad, its text,
image etc.
Second, the number of visitors that are turned into customers at the merchant’s website
is measured by conversion rate. The interviewee con siders the conversion rate as the
most important performance indicator, as it determi nes the income gained from a
specific ad. If the conversion rate is low, he acti vely works with the merchant to learn the
reasons and eliminate them.
Search engine optimization and search engine market ing are used to drive traffic to the
interviewee’s websites. Natural search results acco unt for about 80% of the visitors,
which is achieved by carefully optimizing the websi te for certain search phrases and
placing text links used for link building 5 (either reciprocally or for a fee) at other
websites. Approximately 15% of visitors come from p aid advertising using search engine
marketing, 5% are other forms, such as when user di rectly types the URL in the browser.
When employing search engine marketing, the intervi ewee buys text links in advertising
networks Google AdWords and Sklik. He carefully con trols that the price paid for buying
the advertising is lower than the one earned for cl ick.
The interviewee uses both one-to-one and one-to-man y affiliate programs. One-to-one
programs are more financially beneficial for him, b ut their administration is demanding.
Moreover, one-to-one programs are usually not autom ated and trust between both
parties is needed. Commissions are assigned to the content provider based on for
example orders received through e-mails, which the interviewee has no control of. Thus,
it can also happen that the commission is not recog nized because of merchant’s mistake
or even deliberate deception. All one-to-one progra ms, that the interviewee participates
in, are agreed directly with the merchants.
When using one-to-many affiliate programs, the whol e process is automated and it is not
necessary to communicate with anybody. Even though the commissions are usually
lower, the overhead connected with the program are much lower as well. Because of the
same reason, the interviewee valuates the affiliate networks, which provide all the
necessary technology to manage the program and assu re that the commission is assigned
every time.

5 Link building is a search engine optimization tech nique, which capitalizes on the fact that search en gines
consider websites with more external links better a nd place them higher in the search results. (Wikipe dia
Search Engine Optimization, 2007)

DATA PRESENTATION
– 42 – 4.1.4 Research question 3
How do content providers select affiliate marketing programs they intend to participate in?
When choosing a new provider of an affiliate progra m, the interviewee applies two
different approaches. First, he continuously seeks information about affiliate programs,
which seem financially interesting for him. If he f inds such affiliate program, he tries to
design a website that could exploit the program and bring profit. In this case, he searches
for programs, for which he is capable to attract su fficient amount of visitors, but that are
not in extremely competitive markets.
If he wants to add a new affiliate program to an ex isting website or replace the current
one, he searches for programs that would provide be tter financial conditions than the
previous ones, or than other types of online advert ising he could use at the website.
For both situations, the most important criteria fo r choosing a program are financial
benefits. Nevertheless, the interviewee puts emphas is not only on the amount of
commission rate itself, but also on the merchant’s capability to convert visitors into
customers. If the merchant fails and is not able to sell the advertised product or service,
interviewee loses his commission.
When these conversion rates are not satisfactory fo r a longer period of time, the
interviewee decides to change the affiliate program provider, as well as when he finds a
program with significantly better commission rates. However, he always considers
changing the provider carefully, as it requires dev oting additional time and results are
not guaranteed.
In order to attract more visitors to click at the b anners or text links at his websites, he
values very much the possibility to customize the b anners and text links according to the
needs of his visitors. Therefore, if the affiliate program has strict rules concerning
changing the ads, it can discourage the interviewee from using the pr ogram, because he
fears that he would not be able to persuade suffici ent number of his websites’ visitors to
click at such banners or text links.
The interviewee selects only affiliate programs tha t are connected with the topics he
focuses on and he is knowledgeable about – travelin g, working abroad, technology and
web hosting. Even though he is aware of potential o f affiliate programs in other fields, he
does not intend to use them at the moment. He also wants the program to bring some
additional value to his websites, he is not willing to advertise a product or service that he
does not consider useful.
When choosing an affiliate program provider, the in terviewee does not make any pre-
selection of the program, before choosing the final provider.
Remark:
The interviewee excludes certain form of text links from advertising. These links are
used mainly for link building and their primary fun ction is not to attract visitors directly
to other websites. Still, these links are also sour ce of revenue for the interviewee.

DATA PRESENTATION
– 43 –
4.2 Case no. 2
THOMAS WILKINSON is the owner of Second Sighting Ltd., a company ba sed in Lancaster,
UK. Second Sighting Ltd. runs 4 websites on which a ffiliate marketing is used. All
websites are targeted to the UK market.
The interviewee started with affiliate marketing on a recommendation from a friend,
who was already in the business, in late 2005. He c onsidered it an interesting way how to
start business and he valued that he did not need a ny financial capital to start.
4.2.1 Examples of websites:
Loanexplorer (http://www.loanexplorer.co.uk/) is a simple website allowing visitors to
apply for loans for various purposes. The applicati on itself is loaded from merchant
(storefront, white label) and commission is made ba sed on leads (PPL model).

Figure 11 – Case 2: loanexplorer.co.uk (larger imag e available in thesis appendix)
Remortgageexplorer (http://www.remortgageexplorer.c o.uk/) is a variation on previous
website and promotes a special type of loan – remor tgage. Affiliate marketing is used for
sending applications to merchants. Commissions are paid based on leads (PPL).

Figure 12 – Case 2: remortgageexplorer.co.uk (large r image available in thesis appendix)

DATA PRESENTATION
– 44 – Debtexplorer ( http://www.debtexplorer.co.uk/ ) is another variation of the same
website model, this time offering debt management s olutions. Again, PPL model is used
for receiving commissions from affiliate programs.

Figure 13 – Case 2: debtexplorer.co.uk (larger imag e available in thesis appendix)
Shoppingday ( http://www.shoppingday.co.uk/ ) is a shopping website, which offers
products from hundreds of online stores. Website us es XML technology to download data
feeds from vendors, including product descriptions, image links and prices and blend
them into one shopping place. Pay per sale model is used to receive commission on
intermediated sales.

Figure 14 – Case 2: shoppingday.co.uk (larger image available in thesis appendix)
4.2.2 Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
Interviewee considers easiness of use as a main adv antage of affiliate marketing. He
claims that one only needs a computer and internet connection – there is no capital
needed to start with one’s own business.
Affiliate marketing also enables earning higher rev enue than displaying ads of
advertising networks. The difference is so signific ant that it is beneficial for content
provider to pay for contextual ads at such networks in order to attract visitors to his
websites and display affiliate marketing ads there.

DATA PRESENTATION
– 45 – Further advantage is good tracking, which enables d etermining very precisely visitors’
behavior at the website and after leaving the websi te. Thanks to this, the interviewee can
modify his website to achieve higher conversion rat e.
He also values low switching costs and no obligatio ns – when a program does not
perform as expected, switching is very easy especia lly due to the large amount of
available affiliate programs in the financial secto r, which is his main domain.
On the other hand, there are several disadvantages of affiliate marketing. Deception is
claimed to be the worst one. According to the inter viewee, it is very easy to cheat on
content providers, as they don't have any control o ver the fact whether a sale was made
or not. Interviewee should be paid for every lead t hat he generates through his financial
websites. However, in some cases the leads are not accepted by merchants, who claim
that the customer, who filled in application for a specific product, was not interested in it.
Some affiliate networks unfortunately do not help c ontent providers with this issue,
because they are too afraid to lose the merchant. M oreover, affiliate networks do not
have more control over the merchants that content p roviders themselves do. When leads
are rejected by the merchant, it results in financi al loses, because content provider still
has to pay for visitors coming to his websites from advertising networks.
Especially in the financial sector, interviewee cla ims that payment certainty is very low.
Apart from possible deception or lead rejections, i t is also not uncommon that a
merchant goes out of business.
4.2.3 Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
Base on the categorization by Strauss et al. (2006) , interviewee uses selling agent as a
business model for the websites. Within financial w ebsites, the model is based on
attracting customers by contextual advertising thro ugh advertising networks such as
Google AdWords, Yahoo! or Live.com. Shopping website , on the other hand, relies more
on traffic from natural search result on search eng ines. Its business model can best fit in
subcategory catalog aggregators.
Affiliate marketing is the main source of revenue o n all websites, the only other revenue
source is income from ads displayed through ad netw ork Google AdSense. Affiliate
marketing income is based on sending applications f or financial products to merchants
(e.g. banks). Every application is considered a lea d and commission is paid.
Advertising is displayed in the form of banners and text-links, storefront (while label) is
used to display some of the content. Website owner also keeps a shopping blog, where he
uses advertorials to promote special deals on some products.
Generally, content provider accepts the commission set by the affiliate program provider,
as his position is not strong enough to set the pri ce himself. He prefers to use pay per
lead (PPL) instead of PPS (pay per sale) model. The reason is that he does not have any
control over the conversion rate on the merchant's website and thus PPS is less certain.
He thinks that flat commission rate is more suitabl e for PPL compensation model – the
merchant pays for sending an application regardless of the final amount of purchase.

DATA PRESENTATION
– 46 – Effectiveness is measured carefully, with the cruci al ratio claimed to be the conversion
rate – number of leads divided by number of website visitors. Other ratios are not very
important. If the conversion rate is low, interview ee investigates the reason and makes
appropriate changes on the website to increase it. In a broader view, the financial criteria
(thus return on investment) are the most important criteria.
Interviewee claims that most people come to his web sites through the ads he places on
advertising networks from major search engines (sea rch engine marketing). Currently,
he also educates himself in search engine optimizat ion in order to leverage its potential.
Another important source are social networking serv ices 6, like digg.com etc.
Interviewee only uses one-to-many affiliate program s offered by affiliate networks as he
claims to be too small to be able to negotiate spec ial deals. He however considers one-to-
one programs better, mainly because affiliate netwo rk does not take its part of the
commissions in such situation. In some cases, affil iate networks actually receive up to
50% of commission paid by the merchant. In case of one-to-one programs, the earnings
could therefore be much higher. On the other hand, however, some reputable affiliate
networks actually increase the perception of credib ility and reliability of the programs,
as they can persuade merchants to pay for more lead s, or to pay in time.
Observation:
Financial products website business model is little blurred, because it is something
between selling agent and a metamediary. All three websites hide the brand of the actual
vendor of the service, but also do not offer almost any information about the products
being offered. Websites actually do not add any val ue to the customer's information
search, their purpose is only to make the customer fill in his/her data in order to receive
a proposal (and website owner a commission).
4.2.4 Research question 3
How do content providers select affiliate marketing programs they intend to participate in?
Instead of selecting affiliate programs for a parti cular website, the interviewee first
chooses a program that he thinks could be advantage ous and then he thinks how to use it
and create websites for it.
As the main criteria for choosing an affiliate prog ram, content provider names
commission rate. He is only interested in programs that offer high commission rates, his
preferable compensation model is pay per lead. He r equires high commission rate in
order to be able to cover his expenses connected wi th attracting visitors to his websites
through advertising networks.
Furthermore, he assesses reliability of both mercha nt and the intermediary affiliate
network. A rule of thumb is that better known merch ants are more reliable. He claims to
have negative experience with some networks and doe s not trust them any more.

6 Social network is defined by Kempe et al. (2003) a s „graph of relationships and interactions within a
group of individuals“. In online marketing context, social network services are websites, which offer tools
for users to interact with each other and share the ir interests and activities. (Wikipedia Social netw ork
service, 2007)

DATA PRESENTATION
– 47 – Very important criteria are also other conditions o f the program, especially how lead
rejection is defined. When the definitions are vagu e or it is easy for the merchant to
reject a lead, it is a sign of unreliability. The i nterviewee does not accept such programs,
because he could easily lose his money.
Pre-selection of the programs is made only by choos ing a category of the programs,
based on the knowledge that some financial products generally have high commission
rates and some not. Particularly, in financial sect or mortgages and debts are known to
have high commission, whereas credit cards low.
The final selection is made from available programs on favorite affiliate networks. For
each program taken into consideration, he conducts information search, especially by
searching for user experiences on several affiliate forums. This way, he can find out, if
somebody else had a problem with that merchant and why.
The selection process was claimed to be the same if a program was going to be replaced.
The interviewee decides to replace the current affi liate program, when the lead rejection
rate grows higher than 20%. Then, he starts to sear ch if there is a better product, which
would have either higher commission or that would g ive fewer possibilities to the
merchant to reject the lead. He notes that switchin g affiliate program at his websites is
easy and he is willing to do that every time, when he is not satisfied with the program
performance.

DATA PRESENTATION
– 48 – 4.3 Case no. 3
JAN PETR is a Czech entrepreneur, who owns several successf ul websites. He started with
a website www.jaknaweb.com as a hobby when he was s till a high school student. This
website provided tutorials and discussions about HT ML, JavaScript and other web-
building related technologies. This website soon gr ew very popular and advertisers
started to approach. Today, Jan Petr manages severa l other websites, some of them with
very different business models. All websites are ta rgeted to the Czech market.
4.3.1 Examples of websites:
Jaknaweb.com ( translated “how to make a web” ) is an information portal, which allows
visitors to find information about HTML, JavaScript , cascading style sheets and other
website building technologies. Advertising mainly comes from companies sharing the
same target market, such as hosting companies, doma in registrars, advertising networks
and statistics providers.
The main source of revenue at this website is direc t selling of online advertising, minor
revenue source is affiliate marketing.

Figure 15 – Case 3: jaknaweb.com (larger image avai lable in thesis appendix)
Schuti.cz ( translated “with appetite” ) is a newly developed website, which will contain a
list of all restaurants, bars and pubs in the Czech Republic. The target market for this
website is ordinary people who enjoy eating out and social drinking. Users are able to
find, evaluate and share their favorite restaurants , pubs and bars. Merchants can present
special promotions, news in cuisine, concerts and p arties. Affiliate marketing, together
with pay per inclusion model are planned to be the most important revenue sources for
this website.

DATA PRESENTATION
– 49 –

Figure 16 – Case 3: schuti.cz (larger image availab le in thesis appendix)
Vydelek-online.cz ( translated “earnings-online” ) is a website designed to leverage Google
AdSense affiliate program. Website contains inform ation about Google AdSense, what
information is necessary for application, where to apply, and how the system works. An
identical project promoting Google AdWords is used at a different domain.

Figure 17 – Case 3: vydelek-online.cz (larger image available in thesis appendix)
Nakuptesi.cz ( translated “do shopping!” ) is an online grocery store start-up. Website
offers virtually all grocery shopping with same day home delivery service. Service is
currently only available in one city with approxima tely 400 000 inhabitants, launch in
other cities is planned in mid 2008. Interviewee is responsible for online marketing of
this website and actually uses affiliate marketing from both sides. He places paid links
into search engines and is paid by commissions from received orders and at the same
time he is responsible for planning and managing it s own affiliate program.

DATA PRESENTATION
– 50 –

Figure 18 – Case 3: nakuptesi.cz (larger image avai lable in thesis appendix)
4.3.2 Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
The interviewee claims that affiliate marketing is significantly more complicated to use
than direct advertising or using advertising networ ks.
In the affiliate marketing, the website owner carri es the risk and has to devote his time to
implement the affiliate program well in order to ge t the best results. When displaying
direct advertising and ads from advertising network s, it is the advertiser, who has to
assess, whether the website’s target market is comp atible and the price for the campaign
complies with their requirements. The website owner will get paid, even if the results are
not as good as expected and he only bears opportuni ty costs.
On the other hand, interviewee adds that he is able to earn significantly more by
employing affiliate marketing than by using adverti sing networks. Moreover, after
finding the right way how to implement it, affiliat e marketing is not very time consuming
for the interviewee and he claims that it is a very convenient way how to make money
with little effort. Therefore, he is willing to emp loy affiliate marketing at his websites, if
the difference between revenues from affiliate mark eting and the advertising networks is
high enough to compensate the risk and devoted time in the beginning.
Possibility of deception was named as another disad vantage of affiliate marketing.
However, interviewee found a way to fight against i t. From time to time, he creates a test
order and tracks how the order is being treated by the merchant. He has an agreement
with all merchants allowing him to make such tests. Also, as the competition among
affiliate programs increases, deception was claimed to be lesser problem than it used to
be.

DATA PRESENTATION
– 51 – 4.3.3 Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
Banners and text links are employed for displaying the advertising and affiliate
marketing at the content provider’s websites. The i nterviewee uses following sources of
revenue:
/square4 Direct advertising
/square4 Affiliate marketing
/square4 Advertising networks
Direct advertising is the biggest source of revenue in total, but it is used only at larger
websites with substantial number of visitors. Adver tisers usually pay for a certain time
period, which can be as long as one year in advance . Interviewee notes that on well-run
and known websites, advertisers themselves approach him and he does not need to
make any special effort to sell advertising space.
Compensation model pay per time period is valued by the interviewee, because it gives
him financial security – he knows, how much money h e will get in the specified period.
Moreover, because he sells most of the advertising for longer time period, the
administrative costs are relatively low and he does not have to search for new
advertisers too often. Interviewee does not sell di rectly any advertising based on pay per
click models (PPC).
Affiliate marketing is only used on some websites a nd is highly valued by the
interviewee. Being a well-known publisher, he can t ake advantage of private agreements
or public programs with special-tailored conditions . Again interviewee is approached by
merchants offering him affiliate programs on one-to -one basis and he only actively
searches for affiliate programs when a new project is planned. Interviewee has tried
affiliate networks but was not satisfied with offer ed commission rates and program
availability and he has stopped using them.
Apart from one-to-one affiliate programs, the inter viewee also participates in several
one-to-many programs. One-to-one programs offer him higher revenue, however, one-
to-many programs are automated and require only sma ll amount of time to administrate.
They are a good option, if the interviewee does not have any other advertisers offering
interesting prices for the ad campaign.
Most of the affiliate programs are based on PPS (pa y per sale) compensation model. With
some merchants, the sale is counted by a percentage of a sale value, sometimes the price
is flat. Some merchants also offer PPC (pay per cli ck) in addition to PPS as a special form
of motivation. He values most PPS model with price set as a percentage of sale value,
because it enables him earning the highest revenue.
From the advertising networks, Google AdSense is cu rrently the only one used on
interviewees’ websites. Advertisements are paid per click and he values it for offering
reasonable revenue and not colliding with other rev enue sources, such as affiliate
marketing. Interviewee has tried several other adve rtising networks but he was
disappointed with the output.
Within the Czech market, it is not easy to find goo d products with well managed affiliate
programs. In such situations, Google AdSense still provides some revenue. According to

DATA PRESENTATION
– 52 – interviewee, it is very useful to check available a dvertisers on Google AdSense prior to
creating a new project, as it gives an idea how muc h advertisers need to pay for
advertising and how tense the margins will be.
The interviewee tried to offer e-mail and newslette r advertising to main advertisers but
they did not seem to be too enthusiastic about it a nd finally they abandoned such form of
advertising. Interviewee was not able to think of a ny reason for this behavior.
Price for advertising / affiliate commission is dep endent on contract duration,
advertising space size and position and on advertis ed product. He notes that he is able to
charge more from bigger companies than from smaller firms. Price level is set by
comparing pricelists of similar websites.
The interviewee assesses advertising efficiency by the revenue that it brings to him. If
the revenue is lower than income generated by Googl e AdSense (which the interviewee
considers as a benchmark for all ads), he examines the reasons and replaces it by other
ads. In affiliate marketing, the only measure that is taken into consideration, apart from
generated income, is the conversion rate at the mer chants’ side (number of sales divided
by number of visitors coming from the content provi der’s website).
Most important traffic sources for interviewees’ we bsites are:
/square4 Natural search results
/square4 Paid search engine advertising
/square4 Banner advertising
/square4 PR articles
/square4 Paid links
Interviewee uses both search engine optimization an d search engine marketing
techniques to drive traffic to his websites. For so me websites, banner campaigns and
paid links on another websites are used. He also wr ites public relations articles. He
claims that social networking is going to be used o n currently developed projects.
4.3.4 Research question 3
How do content providers select affiliate marketing programs they intend to participate in?
The interviewee does not usually have possibility t o choose among several affiliate
programs, as there are not so many of them availabl e in the Czech Republic. In most
cases, he does not choose affiliate programs for a website that has already been created,
but he monitors the available affiliate programs co nstantly and creates a website for
leveraging a particularly interesting program readi ly available.
When thinking about a new project, the interviewee thoroughly examines the market.
The most important factor is presence of products t hat match the audience of his
website. For example, people learning to create web sites are likely to be interested in
hosting and domain registration.
When a new program is found or offered to the inter viewee, he assesses its possibilities
and in some cases, he performs a one month test to see what revenue can be expected. If
the results are not satisfactory, he does not use t he program any more in the future.

DATA PRESENTATION
– 53 – He chooses programs according to following criteria :
/square4 Merchant reputation
/square4 Product reputation
/square4 Program tracking and monitoring
/square4 Financial criteria
/square4 Availability of a market segment or niche
/square4 Possibility to create cheap and efficient traffic s ource for the program
Merchant and product reputation are important for t he interviewee, because well-known
merchants and products are more likely to be accept ed well by the website visitors.
However, as he does not have bad experience with sm aller merchants, he is willing to try
their programs, if they provide interesting conditi ons. In such case, it is necessary that
the affiliate program has good tracking and monitor ing for verifying the reliability of the
merchant.
The financial criteria are the most important when evaluating the efficiency of the
affiliate program. In case a program does not perfo rm well, it is abandoned and if
possible, Google AdSense is used as a substitute. A program is considered poor when
monthly revenue is lower than average earnings from AdSense.

DATA ANALYSIS
– 54 –
5 DATA ANALYSIS
In this section, data from all three cases will be analyzed. First, each case is discussed
separately in within-case analysis and subsequently cases are compared in the cross-
case analysis.
5.1 Case no. 1
5.1.1 Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
In the following table, advantages and issues conne cted with affiliate marketing
described in the literature overview are compared w ith content provider’s perception of
them.
Table 2 – Case 1: Advantages and issues with affili ate marketing
Advantages of affiliate marketing Perceived as
Targetability Both advantage and disadvantage
Tracking Minor advantage
Deliverability Not different from other forms of on line
advertising
Interactivity Disadvantage
New revenue opportunities Very important advantage
Access to new merchants Advantage
Added value to the website Neither advantage nor di sadvantage
Issues with affiliate marketing
Banner blindness Disadvantage, but the same as for other
forms of web advertising
Problems with conversion rate Very important disadv antage
Other advantages not described in the literature
Higher income than in some other forms
of online advertising Very important advantage
Little overhead costs Very important advantage

DATA ANALYSIS
– 55 – In accordance with the theory, content provider val ues affiliate marketing, because it
brings him an opportunity how to earn additional in come from his websites and easy
access to new merchants. However, he considers targ etability, which is described in the
literature as one of the main advantages of online advertising, questionable. He notes
that affiliate marketing can be beneficial, but it is necessary to target it very precisely.
Therefore, it is necessary to devote large amount o f time to find out the most efficient
way of employing affiliate program at a particular website.
Possibilities of extended tracking offered by affil iate marketing are perceived as
advantageous, however the content provider sees the m only as a minor benefit.
Deliverability is not considered as more advantageo us than other types of online
advertising. However, interactivity is perceived as less beneficial compared to other
types of online advertising. As content provider is only paid when a merchant actually
succeed to convert a visitor into a customer, he lo ses control over his visitors’ actions
after leaving the website and have to rely on merch ants' abilities. Therefore, if the
merchant’s performance is not sufficient, content p rovider has only very limited options,
how to improve it.
Added value, which can affiliate marketing bring to websites, is not perceived as an
advantage by the content provider. He considers it as an important factor for selecting
the affiliate program, he is going to participate i n, but not as a benefit, that would
persuade him to prefer affiliate marketing to other forms of online advertising.
Two more advantages, that have not been described i n the literature so far, are perceived
as very important by the content provider. First, i t is the opportunity to make higher
revenue than when employing other forms of online a dvertising. Second, only little
overhead costs are connected with administrating af filiate marketing campaigns, which
makes its usage even more beneficial.
When the issues connected with affiliate marketing are concerned, the content provider
does not think that banner blindness is more challe nging in connection with affiliate
marketing than in other forms of online advertising .
However, he perceives conversion rate issues as a p ossible problem that can discourage
him from using affiliate marketing at some websites . As was already mentioned above,
content provider has only a little control over vis itors’ actions after leaving his websites
and he has limited (or none) possibilities how to i nfluence the merchants’ performance.
Therefore, he has to assess the conversion rate car efully in order to keep his income and
not being harmed by problems at the merchants’ side .
5.1.2 Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
Except of one, all interviewee’s websites can best be described as metamediaries or
selling agents. The only exception is the website i -noviny.cz, which does not comply with
definition of a cybermediary and cannot therefore b e categorized using the model
developed by Strauss et al.
Although affiliate marketing is used on all website s, only in some cases it is an important
source of revenue. Moreover, categorization based o n business models developed by
Strauss et al. does not seem to have strong correla tion with revenue sources.

DATA ANALYSIS
– 56 – Table 3 – Case 1: Business model compared to affili ate marketing revenue share
Website Business model Affiliate marketing
revenue share Notes
Letenkar.cz metamediary Minor Ad networks are
major source
Eurovikendy.cz metamediary Major
I-noviny.cz Online
newspaper
(not categorized
by Strauss et al.) Marginal Direct selling is
major source
Birmingham.pl metamediary /
selling agent Exclusive One-to-one program
Manchester.cz metamediary /
selling agent Exclusive One-to-one program
Affiliate marketing revenue share in table 3 was as sessed by the interviewee and by
observation to which extent affiliate marketing is used on respective websites.
With a closer look at the websites, we can however observe that affiliate marketing is the
main source of revenue on websites designed for wel l defined and rather small market
segments (niches):
Table 4 – Case 1: Business model compared to target market size
Website Market segment / niche Affiliate marketing revenue share
Letenkar.cz broad (travel market / air travel) Minor
Eurovikendy.cz narrow (leisure travel market /
weekend getaways) Major
I-noviny.cz narrow (defined by region) Marginal
Birmingham.pl niche Exclusive
Manchester.cz niche Exclusive
The only exception in this table is the example of a regional newspaper i-noviny.cz. On
the other hand, major revenue source on this websit e is direct advertising from regional
companies. The target segment is here defined by ra ther a small region in the Czech
Republic and, as it was additionally confirmed by t he interviewee, there are no local
companies with affiliate programs.
Content provider uses search engine optimization te chniques to drive traffic to his
websites, thus the main source are natural search r esults on major search engines.
Search engine marketing usage depends on its costs and is used only at those websites
where affiliate marketing is employed. In some mark ets, the costs are too high and
acquiring visitors by SEM does not pay off.
According to the data collected, financial gains se em to be the most decisive factor
whether to employ affiliate marketing or another so urce of revenue on a website. In the
case of advertising networks, the income depends ma inly on click through rate, in case of

DATA ANALYSIS
– 57 – affiliate marketing, there is also second ratio, wh ich is the conversion rate measured by
the merchant.
In both cases, content provider actively searches f or new ways how to improve these
ratios. It can therefore be assumed that affiliate marketing is only more beneficial if the
content provider is able to drive to the merchants the type of visitors that actually
convert into customers in high ratios. In order to achieve high conversion rates, the
affiliate program must precisely meet the needs of the visitors.
In situations, where the needs are not met precisel y, or there simply is not any suitable
affiliate program available, other forms of online advertising seem to bring higher
revenues. This is the case of weblog letenkar.cz, w here visitors are not actively searching
for air tickets but rather come regularly to stay i nformed about news in the industry.
Although visitors still have the opportunity to buy air tickets there from affiliated
merchants, the main and preferred source of revenue is an advertising network.
Content provider uses three commission models at hi s websites – pay per click, pay per
sale and pay per time period. Some of the programs, he uses, combine pay per sale (PPS)
model with pay per click (PPC). The main decisive c riterion in selecting the model is the
financial benefits in the long run. According to th e content provider, each of the
commission models has its advantages and disadvanta ges that are summarized in the
following table.
Table 5 – Case 1: Advantages and disadvantages of r espective commission models
Commission model Advantages Disadvantages
Pay per click Revenue is determined
only by content provider’s
performance
Easy to control Lower revenue
Pay per sale Higher revenue Dependent also on
merchants’ performance
Possibility of deception
Difficult to control
More difficult to administer
Pay per time period High revenue
Revenue is dependent on
content provider’s
performance
Secure revenue for a
specified time period Difficult to administer
Difficult to sell advertising
space
The table suggests that pay per sale compensation m odel used in affiliate marketing
offers higher revenue, but on the other hand brings lot of disadvantages too. Content
provider rather prefers to be paid per time period, which is less risky and the revenue is
even higher.
Setting up affiliate marketing efficiently requires great amount of time, thus content
provider has to compare expected outcome to expecte d costs. Although the performance
of affiliate programs can be significantly better t han ads provided by advertising
networks, content provider needs time to find out, how to set up the website to leverage

DATA ANALYSIS
– 58 – it. This overhead is lower with advertising network s, which decide on media
performance themselves.
According to the interviewee, both one-to-one and o ne-to-many programs have
advantages and disadvantages. He claims that one-to -one programs bring higher
commission, whereas one-to-many take advantage of a dvanced automation. Instead of
one-to-many programs, he values affiliate networks more, because they use better
technologies to manage the programs and reduce the risk of deception. Risk of deception
can be considered an important issue in affiliate m arketing.
5.1.3 Research question 3
How do content providers select affiliate marketing programs they intend to participate in?
Affiliate marketing programs are selected by the co ntent provider in two different ways.
He either already has a website, where he uses affi liate marketing, or he creates a
website to leverage a specific affiliate program. I n both cases he goes through a process
that is applicable to models that De Boer et al. (2 001) and Lee (2001) proposed (models
were already discussed in the literature overview).
Table 6 – Case 1: Selection process of an affiliate program
De Boer’s model
Problem definition Replacing current affiliate program:
Unsatisfactory results
Better offer
Costs connected to change of provider

Selecting new program for new website:
Exploit new business opportunities
Diversify risk
Criteria definition Connected to the chosen topics
Adds value to websites
Commission rate
Expected conversion rate
Time length of tracking
Pre-qualification of suppliers None
Choice of the supplier Linear weighting methods
The problem definition phase is different, if the c ontent provider already has a website
or if he decide to create one in order to exploit b enefits of an interesting affiliate
program. In the second case, the content provider i dentifies during the problem
definition phase his need to exploit new business o pportunities in order to diversify a
risk of unsatisfactory results of one website.
In the first situation, the problem definition phas e comprises dissatisfaction with the
current merchant. It can be caused either by bad pe rformance at the merchant’s side or
by better offer from a different affiliate program provider. Potential costs connected with

DATA ANALYSIS
– 59 – the change (such as opportunity cost, time devoted to the change) are also taken into
account. If the expected future benefits do not exc eed expected costs, the process is
finished and content provider will continue using t he current affiliate program.
When future benefits are higher than the costs, con tent provider defines criteria that are
important for choosing the affiliate program. In th is case, criteria are divided into two
groups. First group involves more general criteria – connection to the website topic and
ability to bring additional value to the website. S election criterion determining the
necessity of close connection of the program with t he topic of the website complies with
Papatla and Bhatnagar’s theory, that affiliate prog ram is most beneficial, if there is such a
connection. Moreover, the other criterion– added va lue – corresponds with the reasons,
why content providers can participate in affiliate marketing enumerated by Duffy
(2004).
Second group of criteria deals with the actual inco me that the program can generate. It
includes commission rate, expected conversion rate and length of time, when the
customer is tracked after clicking at a banner or a text link.
No pre-qualification of affiliate program providers is done by the content provider,
because he does not consider it necessary.
In the final phase, the provider is chosen by linea r weighting methods, where the first
group of criteria is non-compensatory – there is a minimum level, that has to be achieved
and that cannot be offset by anything else.
Criteria from the second group are assigned differe nt weights, which play crucial role in
the final decision-making process. The highest weig ht is assigned to the commission rate
and expected conversion rate (which is also connect ed with possibility to customize
freely the ads in the program). Minor weights are a ssigned to length of time, at which the
visitor’s activity is tracked after clicking at the banner or link.
The affiliate program with the best results will be selected by the content provider.
Linear weighting methods are used because of their simplicity and applicability to
various types of situations.
As the content provider continuously searches for n ew affiliate programs and evaluates
them using the above described criteria, Lee’s mode l can be employed as indicated at the
figure below. The process of selecting a new affili ate program continues, until a program
with significantly better conditions is found.
Available
affiliate
programs Criteria for
choosing the
program
Ranking system
for assessing the
affiliate programs
Figure 19 – Case 1: Criteria for selecting affiliat e marketing provider in Lee’s model

DATA ANALYSIS
– 60 –
5.2 Case no. 2
5.2.1 Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
As in the first case, the following table summarize s the advantages and issues of affiliate
marketing that are described in the theory and conf ront them with content provider’s
perception.
Table 7 – Case 2: Advantages and disadvantages of a ffiliate marketing
Advantages of affiliate marketing Perceived as
Targetability Advantage
Tracking Advantage
Deliverability Advantage
Interactivity Advantage
New revenue opportunities Very important advantage
Access to new merchants Advantage
Added value to the website Neither advantage nor di sadvantage
Issues with affiliate marketing
Banner blindness Not considered
Problems with conversion rate Very important disadv antage
Other advantages not described in the literature
Higher income than in some other forms
of online advertising Very important advantage
Easiness of use Very important advantage
Content provider perceives targetability, tracking, deliverability and interactivity as
advantages of affiliate marketing, which complies w ith the theoretical background
described in the literature overview. When targetab ility is concerned, he values the
possibility to change the merchants or ads very eas ily and fast, if he thinks, it would suit
better his target audience. Because he attracts the audience using search engine
marketing, he can also influence what group of visi tors will come to his website
according to merchants’ needs.
Tracking is perceived as beneficial, because he can assess visitors’ behavior at the
website and after leaving the website and adjust th e website according to it. For the same
reasons, deliverability is considered as an advanta ge, because if he finds out, the
conversion rate is not sufficient at the merchant’s side, he can make necessary changes.

DATA ANALYSIS
– 61 – Interactivity is advantageous for the content provi der, because visitors, that come to his
website, can only fill in a form and they will be s ubsequently contacted by a company
offering them products suitable for their needs and requirements.
In accordance with theory, also new revenue opportu nities and access to new merchants
are perceived as affiliate marketing advantages. Th ese advantages are connected with
two other benefits, which have not yet been describ ed in theory – higher income from
affiliate marketing when compared to other forms of online advertising and easiness of
use.
The content provider sees affiliate marketing as a very convenient and simple way how
to start business on the internet. The necessary st arting capital is low and everything is
automated, so it does not require devoting too much time to administrating the ads.
Moreover, affiliate marketing offers earning higher revenue than when employing other
forms of online advertising such as ads from advert ising networks.
Added value that would affiliate marketing bring to the websites, is not considered as an
advantage to the contrary to theoretical background .
When the issues connected with affiliate marketing are concerned, the content provider
does not consider banner blindness as a disadvantag e, because he uses banners only to a
very limited extent. However, he perceives problems with conversion rate as the main
issue. Because he has no control over the visitors’ actions after leaving his websites,
merchants can deceive him and do not grant him comm issions. This issue can be reduced
by using reputable affiliate networks and participa te only in programs, that have strict
conditions about assigning commission, but it canno t be eliminated totally.
5.2.2 Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
All websites presented by the interviewee use busin ess model of selling agent. In case of
shoppingday.co.uk, the business model is catalog ag gregator, a subtype of selling agent.
Affiliate marketing is the main source of revenue o n all websites:
Table 8 – Case 2: Business model compared to affili ate marketing revenue share
Website Business model Affiliate marketing
revenue share Notes
Loanexplorer.co.uk selling agent Exclusive
Remortgageexplorer.co.uk selling agent Exclusive
Debtexplorer.co.uk selling agent Exclusive
Shoppingday.co.uk selling agent Major AdSense is a
supplementary
income source
Similar to case 1, affiliate marketing is exclusive revenue source on websites that are
targeted to a very narrow and well defined group of customers – market niches.

DATA ANALYSIS
– 62 – Table 9 – Case 2: Business model compared to target market size
Website Market segment / niche Affiliate marketing revenue share
Loanexplorer.co.uk Niche Exclusive
Remortgageexplorer.co.uk Niche Exclusive
Debtexplorer.co.uk Niche Exclusive
Shoppingday.co.uk Broad Major
Search engine marketing is used to generate traffic to the websites and the website
owner is actively trying to reduce the dependency o n search engines by learning other
techniques to direct people to his websites. One su ch way is optimizing the website for
search engines, which has been described in the lit erature (Strauss et al., 2006:328-330;
Duffy, 2005). Another way how to attract more visit ors is to leverage the power of social
network websites – a newly emerged category of webs ites.
Interviewee distinguishes conversion rates on his w ebsite (number of leads divided by
number of website visitors) and on merchant’s websi te (number of sales divided by
number of leads). He avoids the problem of not bein g able to influence merchant’s
conversion rate by preferring programs paid by a le ad instead of sale. This way,
merchant is strongly motivated to continuously impr ove his conversion rate. Return on
investment is directly linked to the conversion rat e according to the interviewee and all
other common parameters were rated as not important .
Content provider assessed two commission models tha t he uses or used at his websites –
pay per sale and pay per lead. Following advantages and disadvantages have been found:
Table 10 – Case 2: Advantages and disadvantages of respective commission models
Commission model Advantages Disadvantages
Pay per lead Merchant’s conversion rate
is not a problem Uncertainty – declined
leads
Pay per sale Higher commission Low conversion rate on
merchant’s website lowers
the income, no control over
it.
Moreover, content provider prefers flat commission rather than commission based on
sale volume. This makes calculations easier and mor e reliable and revenue predictable.
Advantages and disadvantages mentioned by the inter viewee have not been found in the
literature. A logical reason for use of pay per lea d (PPL) compensation model could be
the easiness of administration and control. Long pr ocess takes place before an
application to a financial product is approved and volume of sale is not clear. PPL seems
to be more secure for content providers in this cas e.
Even though the content provider uses only one-to-m any affiliate programs offered by
affiliate networks, he considers one-to-one program more beneficial because merchants
do not have to pay part of the commission to the ne twork. As a disadvantage of one-to-
one program, he sees the possibility of deception o n the merchant’s side, which can be
normally decreased by using programs offered by rep utable affiliate networks.

DATA ANALYSIS
– 63 – 5.2.3 Research question 3
How do content providers select affiliate marketing programs they intend to participate in?
Selection process of affiliate marketing programs d iffers, when content provider wants
to replace the existent programs or when he conside rs creating a new website. In case
the content provider already has a website and he i s not satisfied with the results of
current affiliate programs, he chooses new programs according to selecting process
proposed by De Boer et al. (2001) and Lee (2001).
When the content provider wants to participate in n ew affiliate programs and creates
website for it, he goes through four stages describ ed in De Boer’s model as well, the first
phase is, however, different from the first situati on.
Table 11 – Case 2: Selection process of an affiliat e program
De Boer’s model
Problem definition Replacing current affiliate program:
Lead rejection rate is higher than 20%
Selecting new affiliate program:
New business opportunity
Extending website portfolio

Criteria definition Commission rate
Pay per lead compensation model is
preferred
Other conditions of program – possibility
to reject lead
Reliability of merchant
Reliability of affiliate network

Pre-qualification of suppliers Selecting from progr am categories that
can generate high commission
Choice of the supplier Linear weighting methods
In the first phase, the content provider assesses t he performance of current affiliate
programs. If he finds out, that the rate of rejecte d leads by a merchant is higher than
20%, he decides to replace the program by another o ne, which would either offer higher
commission, or would give fewer possibilities to re ject a lead. In case, the revenue is
satisfactory and the number of rejected leads is lo w, the selecting process does not
continue and the existing program is used.
If he considers using new affiliate programs, the p roblem definition phase comprises the
need to exploit new business opportunity and expand ing the portfolio of websites, the
content provider owns.
When the content provider decides to select a new a ffiliate program, he takes into
consideration several criteria. First, he puts emph asis the on commission rate, which is
the most important criterion. He prefers pay per le ad compensation model, but he
considers also other compensation models, if they a re suitable for the websites he owns.

DATA ANALYSIS
– 64 – When the pay per lead model is applied, selection c riteria involve also the possibility to
reject a lead by merchants. If the program conditio ns enable rejecting lead easily, content
provider does not participate in the program, becau se he could easily lose his
commissions.
Furthermore, reliability of merchants and affiliate network is assessed by the content
provider. When he does not consider affiliate netwo rk, which offers a program, reliable
or he has negative experience with it, he will not apply for the program. He prefers co-
operating with merchants and affiliate networks, wh ich are well-known and reputable.
The content provider selects the affiliate programs from categories of program that
enable generating high commissions such as programs offering mortgages or loans.
Finally, the provider of an affiliate program is ch osen using linear weighting methods.
The highest weights are assigned to the commission rate (in connection with offered
compensation model) and little possibility to rejec t leads. Lower weights apply to
reliability of merchant and affiliate network. Merc hant offering the best conditions will
be selected. The content provider uses linear weigh ting methods, because he thinks they
are easy to use and can cover all important criteri a.
Because the content provider constantly assesses th e performance of the affiliate
programs that he uses, Lee’s model can be applied i n the same way as was indicated in
the analysis of case 1.

DATA ANALYSIS
– 65 –
5.3 Case no. 3
5.3.1 Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
The following table summarizes advantages and disad vantages of affiliate marketing
described in the literature and content provider’s perception of them.
Table 12 – Case 3: Advantages and disadvantages of affiliate marketing
Advantages of affiliate marketing Perceived as
Targetability Both advantage and disadvantage
Tracking Neither advantage nor disadvantage
Deliverability Not different from other forms of on line
advertising
Interactivity Minor disadvantage
New revenue opportunities Very important advantage
Access to new merchants Advantage
Added value to the website Not considered
Issues with affiliate marketing
Banner blindness Disadvantage, but the same as for other
forms of web advertising
Problems with conversion rate Important disadvantag e
Other advantages not described in the literature
Higher income than in some other forms
of online advertising Important advantage
Little overhead costs Advantage
From the advantages described in the literature, th e content provider considers as an
evident advantage only new possibilities of revenue and access to new merchants.
Because he offers both direct advertising and affil iate marketing at his websites, the
merchants have more options to choose from and can easily find the solution that suits
their needs.
Targetability is perceived as both advantage and di sadvantage, because thanks to it,
affiliate marketing can generate higher income than other forms of online advertising,
however, it requires devoting time to find out the best way how to present the programs
to the right audience.

DATA ANALYSIS
– 66 – Tracking is not viewed as advantage or disadvantage , the content provider only sees it as
a tool to decrease the risk of deception from the m erchant’s side. Deliverability is not
considered as different for affiliate marketing and other forms of online advertising.
Interactivity is seen as a minor disadvantage, beca use the content provider does not have
control over the visitors’ action after leaving his websites. Therefore, his income is
affected by the merchants’ performance. If the merc hant cannot utilize the interactivity,
the content provider will lose his money as well.
Added value to the website is not taken into consid eration, when assessing the benefits
of affiliate marketing and it is not perceived as a dvantage or disadvantage.
The interviewee mentioned two advantages of affilia te marketing that have not been
described in the literature – possibility to earn h igher income through affiliate marketing
than through other forms of online advertising and little overhead costs.
Problems with conversion rate are perceived as the most important disadvantage of
affiliate marketing. As mentioned above, the conten t provider cannot influence visitors’
behavior during the whole process and his revenues can decrease because of mistakes on
the merchants’ side. Therefore, affiliate marketing can be seen as more risky than other
forms of online advertising.
5.3.2 Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
Interviewee presented three very different websites , targeted to different market
segments. In case of the fourth website – nakuptesi .cz, the interviewee supplies search
engine marketing management and is paid on pay per sale basis. He does not use any
intermediate page but rather places the ads on sear ch engines to link directly to the
merchant. This way, the content provider actually d oes not need any content to use an
affiliate program and acts purely as a form of onli ne promotion outsourcing. This form of
affiliate marketing has not been addressed in liter ature yet.

DATA ANALYSIS
– 67 – Table 13 – Case 3: Business model compared to affil iate marketing revenue share
Website Business model Affiliate
marketing
revenue share Notes
Jaknaweb.com Metamediary Partial One-to-one program s
Schuti.cz Metamediary Partial The website is still under
construction, affiliate
marketing is going to be
one of two sources of
income.
Vydelek-online.cz Metamediary /
Selling agent Exclusive
Nakuptesi.cz Interviewee:
selling agent 7
Website: Online
retailing Not applicable
Again, there is no dependence of affiliate marketin g usage on business model. Instead, as
in previous cases, we can see that affiliate market ing is more used on websites targeted
to niches:
Table 14 – Case 3: Business model compared to targe t market size
Website Market segment / niche Affiliate marketing revenue
share
Jaknaweb.com Medium sized Partial
Schuti.cz Broad Partial
Vydelek-online.cz Small Exclusive
Nakuptesi.cz Broad Not applicable
Target market for websites jaknaweb.com and vydelek -online.cz is very similar, the
website jaknaweb.com is aimed at webmasters and web site designers, vydelek-online.cz
is targeted to website owners and blog owners. Jakn aweb.com however was not
designed to take advantage of affiliate marketing a nd interviewee is able to sell direct
advertising on this website, which is more benefici al for him.
Both search engine optimization and search engine m arketing are used to acquire traffic
for the website. Added to that, interviewee also us es online advertising by placing
banners or text links on other websites. Occasional ly, he also uses public relations by
supplying articles to online magazines. Again, soci al networking as a non-researched
form of traffic acquisition is planned to be used i n the future.
The main criterion for assessing effectiveness is t he financial income. Google AdSense is
used as a measure of minimum income from an adverti sing space and other advertising

7 Selling agent is probably the most appropriate mod el based on the classification by Strauss et al., 2 006. It
is however questionable, if such activity can be cl assified as a provision of content.

DATA ANALYSIS
– 68 – must be able to produce higher revenue. Similar to other two cases, conversion rate is
the only measured non-financial criteria.
Interviewee uses several compensation models on his websites, for which advantages
and disadvantages have been studied:
Table 15 – Case 3: Advantages and disadvantages of respective commission models
Commission model Advantages Disadvantages
Pay per time period Certainty of income, high price , low
administrative overhead Difficult to obtain on
smaller projects
Pay per sale (flat) Lower revenue than PPS
(percentage)
Pay per sale
(percentage) Higher revenue than PPS (flat)
Pay per click
(AdSense) Reasonable revenue, can be used in
addition to other revenue sources.
Except direct advertising, neither of the above men tioned advantages and disadvantages
was described as connected to a respective compensa tion model in the literature.
Interviewee uses both one-to-one and one-to-many af filiate programs. Following
advantages and disadvantages of both categories hav e been found:
Table 16 – Case 3: One-to-one and one-to-many affil iate programs
Program type Advantages Disadvantages
One-to-one Higher revenue Difficult to obtain on
smaller projects,
administration overhead
One-to-many Easy to apply to, easy to administer Lo wer revenue
Based on the findings, it does not seem to be impor tant whether a program is one-to-one
or one-to-many, but one-to-one programs offer bette r conditions than one-to-many
programs. Overhead costs emerge in one-to-one progr ams due to lack of complex
information system for managing the affiliate progr am.
5.3.3 Research Question 3
How do content providers select affiliate marketing programs they intend to participate in?
The content provider usually selects affiliate prog rams before creating a website, where
he implements the program. De Boer’s (2001) model o f supplier selection can be applied
to the process as illustrated below:

DATA ANALYSIS
– 69 – Table 17 – Case 3: Selection process of an affiliat e program
De Boer’s model
Problem definition New business opportunity
Leveraging an interesting program
Criteria definition Financial conditions
Merchant reputation
Product reputation
Program tracking and monitoring
Possibility to access market segment
Possibility to create traffic
Pre-qualification of suppliers None
Choice of the supplier Linear weighting methods
During the problem definition phase, the content pr ovider assesses the available
programs, tries to spot new business opportunities and exploit them. When selecting
affiliate programs for his websites, he takes into consideration several criteria. The most
important criterion is the commission rate and mode l. If they are acceptable, he
evaluates reputation of the advertised product and of the merchant offering the program.
Afterwards, program tracking and monitoring systems are assessed in order to minimize
the possibility of deception from the merchants’ si de. Finally, if all the above described
criteria seem beneficial, the content provider trie s to find out, what market segment can
respond in the best way to the program and how cost ly it would be to attract visitors
from such segment.
As there is only a limited amount of affiliate prog rams at the Czech market, the content
provider does not pre-select them in any way, becau se the small number enables him to
consider all of them. For this process, Lee’s (2001 ) model is applicable in the same way,
as was already described in the analysis of Case 1.
The affiliate program supplier is chosen using line ar weighting methods. The highest
weights are assigned to the financial conditions, o ther criteria can be compensated
among each other.
The content provider decides to stop using the prog ram, when the income generated by
the program is lower than the income from Google Ad Sense. In such a case, the program
is first replaced by Google AdSense and the content provider searches for new affiliate
programs going through the process described above.

DATA ANALYSIS
– 70 –
5.4 Cross-case Analysis
In the cross-case analysis, all three cases are com pared in order to show similarities and
differences between them and find common patterns t hat will enable answering the
research questions.
5.4.1 Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
The advantages and disadvantages of affiliate marke ting perceived by the interviewed
content providers are compared in the following tab le.
Table 18 – Cross-case analysis: Advantages of affil iate marketing
Advantages of affiliate marketing Case 1 Case 2 Case 3
Targetability +, – + +, –
Tracking + + neither adv.
nor disadv.
Deliverability = + =
Interactivity – + –
New revenue opportunities ++ ++ ++
Access to new merchants + + +
Added value to the website neither adv.
nor disadv. neither adv.
nor disadv. not
considered
Issues with affiliate marketing
Banner blindness – not
considered –
Problems with conversion rate – – –
Other advantages not described in the literature
Higher income than in some other forms of
online advertising ++ ++ ++
Little overhead costs ++ +
Easiness of use ++
Table legend:
+, – both advantage and disadvantage
+ advantage
++ very important advantage
– disadvantage
– very important disadvantage
= it is not different from other forms of online adv ertising

DATA ANALYSIS
– 71 – All three content providers consider new revenue op portunities and access to new
merchants as advantages of affiliate marketing. Mor eover, they all added, that affiliate
marketing not only offers them new revenue opportun ities, but that the income
generated by affiliate marketing can be significant ly higher than when using other forms
of online advertising.
This can be explained by the risk content providers have to bear, when employing
affiliate programs. As they are paid only for a spe cified actions, their income is never
secure and therefore, the programs have to offer co ntent providers sufficient incentives
to take this risk.
On the other hand, affiliate marketing decreases th e risk for merchants, because they will
not waste their money on advertising that is not ef ficient. Thus, they are willing to
advertise even at websites that would not be taken in consideration otherwise. Thanks to
this effect, content providers’ access to merchants is wider.
Little overhead costs were also considered as an ad vantage of affiliate marketing by two
content providers, whilst the third one emphasized the easiness of use connected with
affiliate marketing. As affiliate programs are most ly automated, content providers do not
have to devote their time to administrate them. Mor eover, affiliate programs are often
operated on one-to-many basis and therefore, compan ies try to make the system as easy
as possible to attract great amount of content prov iders and minimize the time they
would need to spend on answering their questions an d administrating the program.
Other advantages described in the literature were n ot perceived as benefits
unanimously. Targetability was perceived as benefic ial, however content providers were
also aware of the issues connected with it. They no ted that affiliate program, when
targeted well, can increase their income, but preci se targeting requires devoting time to
it and the results are not secure.
Tracking was considered as a minor advantage or nei ther advantage nor disadvantage.
This can be explained by the fact, that even though tracking offers content providers
good information about their visitors’ actions and behavior, it does not affect their
income directly. Nevertheless, content providers no te that good tracking can minimize
the possibility of deception and give them higher s ecurity.
Deliverability was not considered as an important b enefit of affiliate marketing, mostly
because content providers did not see any differenc e between affiliate marketing and
other forms of online advertising in this aspect.
Neither of content providers thought that affiliate marketing would add their websites
any new value. Affiliate marketing was seen as a fo rm of online advertising, which can be
usually replaced by another form. Content providers do not derive value from a specific
affiliate program, because it would make their webs ites dependent on the program and it
would be very difficult to replace it, when its res ults are not satisfactory.
Interactivity, even though it is described as advan tage of affiliate marketing in the
literature, was perceived as beneficial only by one content provider, who was paid per
lead (he could affect number of leads to a full ext ent). The other two content providers
saw interactivity more as and disadvantage and conn ected it with issues with conversion
rate. They stated that interactivity is problematic , because they lose control over their
visitors’ activity after leaving the website and th erefore cannot impact the actions of
visitors that would lead to assigning the commissio n to their account.

DATA ANALYSIS
– 72 – Because conversion rate is to a great extent influe nced by merchants’ ability to persuade
visitors to buy a product and exploit the interacti vity, content providers see it as a threat.
Moreover, as the content providers have only limite d possibilities how to control
merchants, they can also lose some of their commiss ions because of a mistake or
deception.
In all three cases, problems with deception have be en thoroughly discussed and
possibility of deception seems to be one of the mos t important issues in affiliate
marketing. Content providers do not actually have a ny proof of being deceived unless
they try it themselves with test orders. Submitting test orders is however not always
possible without actually paying for a product or s ervice or without prior consent from
the merchant. Merchants that want to avoid being ac cused from deception should
implement systems to allow content providers to hav e real time information about leads
and sales and have strictly defined rules stating u nder which conditions only a lead or a
sale can be refused.
Finally, banner blindness was perceived as a minor issue, however content providers
considered it as basically the same for all forms o f online advertising, because they all
use the same means to attract visitors.
5.4.2 Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
Most of the websites presented in the three cases u sed business models of a selling agent
or a metamediary. Affiliate marketing was used acro ss the websites with no relevance to
a particular business model. We have however found a pattern that binds share of
affiliate marketing revenue share to the size of th e target market for a particular website.
The table bellow summarizes this observation from a ll three cases:
Table 19 – Cross-case analysis: Target Market relat ed to affiliate marketing revenue share
Target market /
Affiliate marketing
revenue share Case I Case II Case III
Broad Minor (1) Major (1) Partial (1)
Medium Partial (1)
Narrow Major (1)
Marginal (1) Exclusive (1)
Niche Exclusive (2) Exclusive (3)
In all three cases, the narrower the target segment was, the higher was the share of
affiliate marketing. The number of websites include d in the segment for each case is
shown in parenthesis. Exceptions from the pattern a re described with individual cases.
All three content providers used advertising networ ks, particularly Google AdSense.
Income from Google AdSense is perceived as a measur e of a minimum income from a
website. Affiliate marketing was used only when the earnings generated from affiliated
ads were better.

DATA ANALYSIS
– 73 – Traffic acquisition summarized
All three interviewees rely on both forms of search marketing – search engine
optimization and search engine marketing, but also try to use other sources for traffic to
their websites. In all three cases, search engine m arketing, i.e. advertising in search
engines, is used mainly for websites that use affil iate marketing and that are targeted to a
market niche. A likely reason for this is that cont ent providers are able to track down
relevant ads on search engines based on affiliate m arketing outcomes and the
mathematics for the balance between traffic acquisi tion costs and affiliate marketing
revenue is straightforward.
All content providers are trying to leverage search engine optimization in order to better
promote their websites in natural search results. T he main advantage is that after one-
time investments to optimize their websites and pro mote them, the number of visitors
coming from natural search results (i.e. for free) should significantly increase. This
behavior was already described by Duffy (2005).
Other forms of traffic acquisition varied in cases. Whereas in case II and III, social
networking was used or planned to be used, direct b anner advertising and public
relations were also used in case III.
Efficiency measurement
All three content providers consider as main criter ion whether to employ affiliate
marketing financial indicators. As the main perform ance indicator, all three content
providers named conversion rate. None of them consi dered any other performance
indicator important for affiliate marketing, except the ration of declined leads in case II,
which however is not an indicator discussed in the literature overview. Conversion rate
as the main indicator confirmed the theory mentione d by Duffy (2005).
Compensation models
In the three cases, content providers generally sha red the same attitudes towards
different compensation models:

DATA ANALYSIS
– 74 – Table 20 – Cross-case analysis: Compensation models
Commission model Advantages Disadvantages
Pay per time period Certainty of income, high price , low
administrative overhead Difficult to obtain on
smaller projects
Pay per lead Dependent solely on content
provider’s performance
Commissions lower than in PPS Uncertainty, lead rejection
Possibility of deception
Pay per sale (flat) High revenue Dependent on merch ant’s
performance
Difficult to control
Possibility of deception
Lower revenue than PPS
(percentage)
Pay per sale
(percentage) Higher revenue than PPS (flat) Dependent on merchan t’s
performance
Difficult to control
Possibility of deception
Pay per click Dependent solely on content
provider’s performance
Can be used in addition to other
revenue sources.
Easy to control Low revenue compared to
others
Most valued model is advertising paid directly by m erchant and compensated on a time
period basis. According to content providers, this form of advertising brings the highest
revenue and income is secure. On the other hand, it is hard to find such advertisers and
they are only willing to cooperate with big website s.
Affiliate marketing brings highest revenue after di rect advertising, but only if it is used
right. Although it is easy to find advertisers, it is not easy to set up the website to achieve
satisfactory conversion rates. Website must be very well targeted and promoted
products and services must be attractive for majori ty of website visitors. Within
available compensation models in affiliate marketin g, content providers do not seem to
have strong preference on a particular compensation model.
The findings correspond with the literature focusin g on commission models only in some
aspects. In accordance with Zeff (1999), it was fou nd out, that result-based models (PPC
and PPS) offer better opportunities for selling adv ertising space. However, the lack of
control over advertising media and advertised produ ct was only an issue in PPS model,
not in all result-based models as Zeff (1999) claim ed. This discrepancy can be explained
by development of new advertising networks providin g contextual advertising
corresponding with the website topic that enables i ncreasing content providers’ control.
Furthermore, compensation model pay per lead solves the issue of merchant’s
conversion rate, but on the hand, commissions are b elieved to be lower and the fact that
some percentage of the leads is rejected may be uns atisfying for content providers.

DATA ANALYSIS
– 75 – Literature does not describe any other advantages a nd disadvantages connected with
specific commission models such as possible revenue , that can be generated though each
of them, or possibility of deception by merchants. Moreover, it was found out, that
preference of a specified commission model can be i nfluenced by the difficulties with its
administration.
Advertising networks (which offer PPC compensation) are used in cases where direct
advertising or affiliate marketing could not be use d. It is considered a secure but rather
little possible source of revenue for all websites. Moreover, ads from advertising
networks are very easy to set up and do not incur a ny costs related to further
management.
These findings correspond with the theory proposed by Chaffey (2003:344) that affiliate
marketing is beneficial especially to smaller websi tes, which are unable to sell
advertising directly.
Affiliate marketing program types
Content providers share similar point of view on on e-to-one and one-to-many affiliate
programs and on affiliate networks. Their perceptio ns are summarized in the following
table:
Table 21 – Cross-case analysis: Advantages and disa dvantages of respective affiliate marketing program types
Affiliate marketing
type / case Case I Case II Case III
One-to-one Higher commission
Low automation Higher commission
Risk of deception Better conditions
Low automation
One-to-many Lower commission
Advanced
automation

Affiliate networks High level of
automation
Lower risk of
deception Low commission
Lower risk of
deception Low commission
Content providers value one-to-one programs, becaus e these are able to bring highest
revenues. They are nevertheless harder to control a nd administer and also only available
to bigger websites. On the other end, affiliate net works provide necessary infrastructure
and easy management, but also take part of the earn ings so that commission from such
programs is lower. These findings correspond with t he theory (Oberndorf, 1999),
although in case of the interviewees, the main moti vation to prefer one-to-one programs
was price.

DATA ANALYSIS
– 76 – 5.4.3 Research Question 3
How do content providers select affiliate marketing programs they intend to participate in?
In all three cases content providers were usually s earching for new affiliate programs to
use at already existing websites, when they were no t satisfied with performance of the
used programs or when they believed they were cheat ed by the merchants.
Otherwise, they were constantly monitoring offer of affiliate programs and if they
encounter interesting affiliate program, they decid ed to create a new website for it. The
process of monitoring the market can be depicted us ing Lee’s (2001) model.

De Boer’s (2001) model is applicable for both above described situations, however the
first phase of the process differs. If the content provider wants to replace an existing
program, the problem definition involves dissatisfa ction with current affiliate program
provider or getting a better offer from another pro vider.
When content providers searched for affiliate progr ams for their new websites, they
defined the problem as the need to find new busines s opportunities, expand their
portfolio of websites in order to decrease dependen ce on one source of income and
chance to leverage interesting affiliate program.
Criteria, the content providers considered as the m ost important for selecting an affiliate
program, were different for every content provider. Nevertheless, all of them placed
emphasis on financial conditions and other factors connected to them (such as
conversion rate). Then, merchants’ reputation and m inimizing potential deception were
also taken in consideration, as well as tracking an d monitoring systems and connection
to the website topic.
None of the content providers did any pre-qualifica tion of the merchants. For the final
choice they all use linear weighting methods becaus e of their simplicity and possibility to
decide fast. The content providers did not consider using any other selection methods,
because they believed linear methods give them suff icient means to decide well. Available
affiliate
programs Criteria for
choosing the
program
Ranking system
for assessing the
affiliate programs
Figure 20 – Cross-case analysis: Monitoring the mar ket – Lee's model

DATA ANALYSIS
– 77 – The highest weights were assigned to the financial conditions. Usually some criteria were
not compensatory (such as connection to the website topic), some could be
counterbalanced with better performance in other fi elds.

CONCLUSIONS
– 78 –
6 CONCLUSIONS
6.1.1 Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
The work has identified three main advantages of af filiate marketing for content
providers. First, it offers new revenue opportuniti es for them. Second, generated income
from affiliate marketing can be higher than when us ing some other forms of online
advertising. Finally, using affiliate marketing pro gram is connected with low overhead
costs on the side of content providers.
Affiliate marketing was seen as a good combination between relatively low overhead
costs (when compared for example to direct advertis ing) and reasonable income.
Moreover, in some cases it was also a possible way how to attract new merchants, which
would normally not cooperate with smaller websites.
To exploit these perceived benefits, merchants with affiliate programs or those planning
to start new affiliate programs, should make them v ery easy to use (for example making
the invoicing process automated etc.). This would e ven lower the overhead costs. The
commission rate offered in the program should be co mpetitive to other programs in the
field. The program should also enable earning reven ue that is higher than average
income from Google AdSense in the same business are a.
The biggest disadvantage of affiliate marketing wer e considered to be problems with
conversion rate, because content providers do not h ave possibility to influence in full
extent the actions that would lead to assigning com mission to them. Therefore, they are
afraid of unsatisfactory performance of merchants, their mistakes or deception from
their side.
Problems with conversion rate are not so important in other forms of online advertising.
As content providers are paid only for actions, the y can influence, conversion rate at the
merchant’s side does not concern them. Performance indicators, that influence income
from these forms of advertising, can be measured by content providers themselves. In
such case the threat of deception is also minimized .
In order to minimize this disadvantage, companies, that have their affiliate programs,
should offer content providers a good monitoring sy stem that would give them accurate
statistics and enable to follow their visitors’ act ions after leaving their websites.
Moreover, visitors tracking should be fully automat ed to avoid mistakes.
In case merchants are not able to provide such syst em, they could consider offering
affiliate program through an affiliate network that supplies necessary solutions.
Reputable affiliate network can also add credibilit y to the program.

CONCLUSIONS
– 79 – 6.1.2 Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
We have confirmed the hypothesis that the main deci sion factor for content providers
whether to use affiliate marketing on their website s seems to be the profit. They are
generally considering three income sources: direct advertising, affiliate marketing and
advertising networks. Generally, they expect the h ighest profit from direct advertising
and the lowest from advertising networks, while aff iliate marketing stands in the middle.
It seems that all three sources have their place on the market. It is not always possible to
use direct advertising and the use affiliate market ing is not always better than
advertising networks.
Although it is not possible to generalize from the findings in a qualitative study, it seems
that affiliate marketing is beneficial especially o n websites that are precisely targeted to
a well defined and narrow market segment. Usually, these are small websites that
advertise few products of one merchant. Content pro viders create websites specifically
designed to a specific affiliate program or product type rather than use it on existing
websites.
A possible explanation, why is advertising network in some cases better than affiliate
marketing and vice versa, is targeting. In case of an advertising network the supplier of
the service decides which ads to present to a parti cular visitor. Displayed ads are based
on page content, visitor’s native language and poss ibly on broader behavioral patterns of
a particular visitor. These ads are therefore highl y personalized and successful even on
websites with broad target segments.
With affiliate marketing, it is the website owner, who decides on advertised products or
services, advertisement message, language and form. Level of personalization is very
limited in this case, because it is very difficult to achieve it. Affiliate marketing can
therefore be only successful, if the campaign hits substantial percentage of visitors. A
small website specifically designed for a market ni che can attract these visitors and serve
them ads that perfectly match their interests.
We also found a relation between traffic sources an d affiliate marketing usage. Whereas
natural search results are main source of traffic f or most websites, paid search engine
advertising (search engine marketing, SEM) is also used on websites where affiliate
marketing is present. This relation probably result s from a fact that with affiliate
marketing it is easy to decide whether such adverti sements pay off. We therefore cannot
conclude that affiliate marketing should be used on websites that use search engine
marketing.
Although content providers have their preference on types of affiliate programs or usage
of intermediaries in a form of affiliate networks, the ultimate decision factor is always
the revenue. As deception was named as main issue o f affiliate marketing, we would
suggest that one-to-one programs are more suitable for well known merchants with
good brand image, while affiliate networks are a go od start for all other companies.

CONCLUSIONS
– 80 – 6.1.3 Research Question 3
How do content providers select affiliate marketing programs they intend to participate in?
When selecting an affiliate marketing program, cont ent providers go through three out of
four stages described in De Boer’s (2001) model. Th ey first define a problem. In this
phase, they are either not satisfied with a previou sly used program, or they want to
exploit new business opportunities offered by new a ffiliate programs at the market.
In the second stage of De Boer’s model, content pro viders define criteria for selecting the
affiliate program. The most important criteria were financial. However, the emphasis
was placed at expected income from the program in a specified time. The commission
rate itself played a substantial role, but other fa ctors influencing the final revenue were
also taken into consideration. These factors could discourage content providers from
participating in the program.
Content providers did not make any pre-selection of affiliate programs apart from
choosing a category of programs suitable for their websites. As the program category
could be also placed into the selection criteria, w e can assume that the third stage can be
eliminated from the De Boer’s model for choosing an affiliate program.
Finally, content providers select the affiliate pro gram using linear weighting methods.
They use these methods because they are simple, eas y-to-use and applicable to various
problems.
As content providers constantly monitor the market for new affiliate programs, they go
through stages described in Lee’s (2001) model. We can therefore conclude that both De
Boer’s and Lee’s models are applicable for the proc ess of selecting an affiliate program
provider both for a new or existing website.
When a company wants to start an affiliate program and attract content providers to
participate in it, it should carefully set the cond itions of a program to be competitive.
Commission rate and length of tracking time should be at least at the same level as
competitors’. Conditions should be very clearly def ined so that content providers would
not be afraid of deception or mistakes at merchant’ s side. The threat of deception can be
also minimized by offering good monitoring and trac king system. Finally, the company
should persuade content providers about its trustwo rthiness. It should for example
follow the most important affiliate forums in order to find content providers’ opinions
about the program. If any negative comments appear, the company should give
explanation or retrieve the error.
6.1.4 Other conclusions
Although the following was not the aim of our resea rch, the knowledge gained by
thoroughly studying the topic allowed us to make fe w more points:
Role of content providers in affiliate marketing
We have drawn a complex diagram of the role of cont ent providers in affiliate marketing:

CONCLUSIONS
– 81 –

The diagram above suggests that affiliate marketing is a form of promotion outsourcing,
where content providers take care about merchant’s promotion and receive
compensation based on their performance. But why me rchants actually need content
providers and why they do not take care about the p romotion themselves?
A probable answer would be that affiliate marketing is much more efficient than keeping
the promotion in the house. With affiliate marketin g, merchants can take advantage of
dozens or hundreds of individuals with lot of exper ience and lot of new ideas and they do
not need to care if they succeed or not. Merchants only spend money on promotion,
which proved to work. Moreover, affiliate marketing does not mean that the merchant
cannot do its own promotion too.
Content provider as obsolete terminology
Although the term content provider is extensively u sed in the literature, it does not
precisely describe the merit. Firstly, content prov iders, especially in some cases, actually
do not provide much content. Some of the websites p resented in the cases actually
contained very little own content, most of it was c ontent loaded from merchant’s servers.
In such cases, content providers only designed a we bsite and most part of their work was
its promotion.
Secondly, we have found cases where content provide rs did not use a website at all –
they simply placed ads to the search engines, which pointed directly to the merchant’s
website and received commission if visitors coming from such ads actually converted
into customers. In this case, content providers are solely providers of knowledge where
to advertise and for which price. Customers Content Provider Merchant Search Engines
Natural search
results
Paid advertising
Other websites
Ads Links Flow of customers
Flow of money
Figure 21 – Role of content providers in affiliate marketi ng

CONCLUSIONS
– 82 – Both content providers and affiliate networks gener ally use terms affiliates or publishers
to describe content providers. The term affiliates definitely is more appropriate to
describe content in the context of affiliate market ing, but cannot be applied on content
providers that use other income sources. The term p ublishers, on the other hand, is more
used in the sense of a newspaper or magazine publis her in the literature.
Suggestions for further research
The role of content providers in affiliate marketin g has been outlined in this study and
should be further researched, especially with conne ction to the factors that make the use
of intermediaries more effective than in house prom otion. Based on observation,
merchants strongly vary in to which extent they rel y on content providers.
While it seems to be clear, why advertising network s perform better on websites
targeted to broad target segments, it is unclear wh y affiliate marketing is better in
narrow segments. If advertising networks have techn ology that can serve perfectly
targeted ads to a wide variety of visitors, why do they have problems offering ads in the
same quality on niche websites?
Traffic acquisition is a primary concern for conten t providers as inevitably need website
traffic for their business. Yet, this area has not b een researched properly and some of the
traffic sources mentioned by content providers were not covered by the literature.

REFERENCES
– 83 –
7 REFERENCES
Affiliate Marketing . Wikipedia. Cited 2007-10-01. (Available online at :
http://en.wikipedia.org/wiki/Affiliate_marketing).
Barnes & Noble.com Affiliate Program. Barnes & Noble. c2007. Cited 2007-09-12. (Availabl e online at:
http://www.barnesandnoble.com/affiliate/questions.a sp?#reporting).
BENWAY, J.P., LANE, D.M. Banner Blindness: Web Sear chers Often Miss "Obvious" Links. Internetworking,
Dec 1998. (Available online at:
http://internettg.org/newsletter/dec98/banner_blind ness.html).
BRUNER, R.E., The Decade in Online Advertising 1994-2004 . DoubleClick. c2005. Cited 2007-09-16.
(Available online at:
http://www.doubleclick.com/insight/pdfs/dc_decaderi nonline_0504.pdf).
CHAFFEY, D. Internet Marketing: Strategy, Implementation and Pr actice . 2 nd ed. New York: Financial Times,
2003. 484 p. ISBN 0-273-65883-2.
CHATTERJEE, P. Interfirm alliances in online retail ing. Journal of Business Research , 2002, vol. 57, p. 714-
723.
CHO, M., et al. Search based Online Advertising. 2005, Cited 2007-09-16 (Available online at:
http://www.ischool.berkeley.edu/~hal/Courses/StratT ech07/Tech/Tech05/C-report.doc).
Commission Junction. Cited 2007-09-20. (Available o nline at: http://www.cj.com).
Compensation methods . Wikipedia. Cited 2007-10-01. (Available online at :
http://en.wikipedia.org/wiki/Compensation_methods).
DE BOER, L., LABRO, E., MORLACCHI, P. A review of m ethods supporting supplier selection. European
Journal of Purchasing & Supply Management , 2001, vol. 7, p. 75-89.
DUFFY, D.L. Affiliate marketing and its impact on e -commerce. Journal of Consumer Marketing , 2005, 22/3,
pp. 161-163.
EISENMANN, T., BROWN, A. Online Content Providers. Harvard Business School , Nov 2000.
EU Directive . European Commission. c2002, Cited 2007-09-07 (Ava ilable online at:
http://www.europarl.eu.int/commonpositions/2002/pdf /c5-0035-02_en.pdf).
GALLAUGHER, J.M., AUGER, P., BARNIR, A. Revenue Str eams and Digital Content Providers: An Empirical
Investigation. Information & Management , August 2001, vol. 38, no. 7, p. 473-485.
GIAGLIS, G. M., KLEIN, S., O’KEEFE, R. M. The role of intermediaries in electronic marketplaces: devel oping
a contingency model. Info Systems Journal , 2002, vol. 12, p. 231-246.
Google AdWords Learning Center. Google.com. c2007. Cited 2007-09-08 (Available onl ine at:
http://services.google.com/awp/en_us/breeze/2976196 /index.html).
Google Analytics Product Tour. Google.com. c2007, Cited 2007-09-10 (Available onli ne at:
http://64.233.179.110/analytics/tour/index_en-US.ht ml).
Google Press Release. Google.com. c2007. Cited 2007-09-08 (Available onli ne at:
http://investor.google.com/releases/2006Q4.html).
FIORE, F. e-Marketing strategies: plans for the new economy . 1 st ed. Indianapolis, Ind.: Prentice Hall, 2000.
300p. ISBN 0-7897-2475-8.
HARRIS, T. How Affiliate Programs Work. HowStuffWorks, c2007. Cited 2007-09-10. (Available online at:
http://money.howstuffworks.com/affiliate-program.ht m ).

REFERENCES
– 84 – HOFFMAN, D. L., NOVAK, T. P., CHATTERJEE, P. Commer cial Scenarios for the Web: Opportunities and
Challenges. 1995 (Available online at: http://jcmc. indiana.edu/vol1/issue3/hoffman.html).
HOFFMAN, D. L., NOVAK, T. P. How to Acquire Custome rs on the Web. Harvard Business Review , May – June
2000, vol. 78, no. 3, p. 179-183.
Internet Marketing Dictionary . Marketingterms.com. c2007. Cited 2007-05-10. (Ava ilable online at:
http://www.marketingterms.com/dictionary/affiliate_ network ).
JACOBS, J. The Rebirth of Intermediation: E-marketing and Infomedi aries as Distribution Channels . 2002.
Cited 2007-09-06. (Available online at:
http://www.joannejacobs.net/pubs/2002/anzmac02_man. doc ).
Join Associates . Amazon.com. c2007. Cited 2007-09-12 (Available on line at: http://affiliate-
program.amazon.com:80/gp/associates/join/102-577985 8-3403355 ).
JANOSCHKA, A. Web Advertising . Amsterdam: John Benjamins Pub., 2004. 227p. ISBN 90-272-5374-9.
KEMPE, D., KLEINBERG, J., TARDOS, E. Maximizing the Spread of Influence through a Social Network . 2003.
Cited 2007-12-17. (Available online at:
http://media.eurekalert.org/aaasnewsroom/2004/Klein berg-Influence-Paper.pdf)
LANNIN, P., BOLDING, P. Swedish saver group rejects AOL TradeDoubler bid . Reuters. 2007-02-21. Cited
2007-09-06. (Available online at: http://www.reuter s.com/article/technology-media-telco-
SP/idUSL2173259420070221 ).
LEE, E.K., HA, S., KIM, S.K. Supplier Selection and Management System Considering Relationships in Sup ply
Chain Management. IEEE Transactions on Engineering Management , August 2001, vol. 48, no.
3, p. 307-318.
LIBAI, B., BIYALOGORSKY, E. GERSTNER, E. Setting Re ferral Fees in Affiliate Marketing. Journal of Service
Research , May 2003, vol. 5, no. 4, p. 303-315.
OBERNDORF, S. Web Marketing: Get yourself affiliated . Multichannel Merchant. August 1999. Cited 2007-
08-30. (Available online at:
http://multichannelmerchant.com/news/marketing_web_ marketing_yourself/ ).
OECD. The Economic and Social Impacts of Electronic Comme rce . 1999. Cited 2007-09-02. (Available online
at: http://www.oecd.org/dataoecd/3/12/1944883.pdf ) .
PALMER, A. Principles of services marketing . 3 rd ed. London: McGraw-Hill, 2001. 516p. ISBN 0-07-709 748-3.
PAPATLA, P., BHATNAGAR, A. Choosing the Right Mix o f On-line Affiliates: How Do You Select the Best?
Journal of Advertising , fall 2002, vol. 31, no. 3, p. 69-81.
ROWLEY, J. Just another channel? Marketing communic ations in e-business. Marketing Intelligence &
Planning , 2004, vol. 22, no. 1, p. 24-41.
SARKAR, M., BUTLER, B., STEINFIELD, C. Intermediari es and Cybermediaries: A Continuing Role for
Mediating Players in the Electronic Marketplace. Journal of Computer Mediated
Communication, 1995, vol. 1, no. 3.
Search engine optimization. Wikipedia. Cited 2007-12-17. (Available online at:
http://en.wikipedia.org/wiki/Link_building).
SLOVIC, P. Perception of Risk. Science , 1987, vol. 236, p. 280-285.
SMYTKA, D.L., CLEMENS, M.W. Total Cost Supplier Sel ection Model: A Case Study. International Journal of
Purchasing and Materials Management , winter 1993, p. 42-49.
Social network service. Wikipedia . Cited 2007-12-17. (Available online at:
http://en.wikipedia.org/wiki/Social_network_service ).
SPIR přijal nové standardy reklamních formátů . SPIR. c2007. Cited 2007-09-07. (Available online at
http://www.spir.cz/spir/pubCZ/home/news/Main$pagele tManager$ArticleView/1571.jet?r
ef=1).
STRAUSS, J., ADEL, E.A., FROST, R. E-marketing. 4th edition. Upper Saddle River, N.J.: Pearson Prentic e Hall,
2006. 456p. ISBN 0-13-196902-1.

REFERENCES
– 85 – TIMMERS, P. Business Models for Electronic Markets. European Commission, Directorate-General III,
September 9, 1998 Research Note #98-21. Cited 2007- 09-05. (Available online at:
http://www.imse.hku.hk/imse2016/Readings/Essential/ BusinessModel-
Geoffrion/Timmers/98_21_n.html).
TradeDoubler Help Centre – Keyword Marketing . TradeDoubler. Cited 2007-08-30. (Available online at:
http://img.tradedoubler.com/images/uk/help_centre/k eywordmarketing.htm).
TradeDoubler Investor Relations – Business Model . TradeDoubler. Cited 2007-08-30. (Available online at:
http://ir.tradedoubler.com/ir/en/omtradedoubler/aff arsmodell/affarsmodell.html).
TradeDoubler – td Pull – make money online . TradeDoubler. Cited 2007-08-19. (Available online at:
http://www.tradedoubler.co.uk/pan/cms/products/prod ucts_publishers/prod_publ_td_pull)
.
ValueClick to Acquire Commission Junction; Acquisit ion Would Form Industry's Leading Affiliate Marketi ng
Solutions Provider. ValueClick . Cited 2003-10-10. (Available online at: http://ph x.corporate-
ir.net/phoenix.zhtml?c=84375&p=irol-newsArticle&ID= 457609 ).
Yahoo Finance. YAHOO!., c2007. Cited 2007-09-08. (Available online at:
http://finance.yahoo.com/q/bc?s=GOOG&t=5y).
Zanox . Cited 2007-09-20. (Available online at: http://www .zanox.com).
ZEFF, R. L., ARONSON, B. Advertising on the Internet . 2 nd ed. New York: Wiley Computer Publishing, 1999.
436p. ISBN 0-471-34404-4.


– 86 –

APPENDICES

APPENDIX : INTERVIEW GUIDE
– 87 –
Interview Guide
Introduction
What is the purpose of your website(s)?
Do you use on-line advertising on your website? Is it main or complementary source
of revenue?
How do you sell online advertising? Why?
Hints: Direct, Representation firm, Ad networks, Auc tions, Affiliate marketing
Research Question 1
How do content providers perceive advantages and di sadvantages of affiliate marketing
compared to other forms of online advertising?
1.1 What type of online advertising do you offer at your websites?
Hints: E-mail: discussion lists, newsletters, Web: banners, text links, sponsorships,
advertorials, interstitials, microsites
1.2 What do you perceive as main advantage and disa dvantages of affiliate marketing
compared to other types of online advertising that you use? Why?

APPENDIX : INTERVIEW GUIDE
– 88 – /square4 Advantages in:
o Targetability
o Tracking
o Deliverability and Flexibility
o Interactivity
/square4 Issues in:
o Banner blindness
o Pricing models
/square4 Other:
Research Question 2
How can the situations, under which affiliate marketin g is efficient for content providers, be
characterized?
2.1 How do you measure effectiveness of online camp aigns at your websites? Why
these?
Hints: Ad views, reach, click-throughs, CTR, CPM, CP A, CPS, conversion rate, effective
frequency
2.2 Which revenue sources do you use at your websit es? Why?
Hints: online advertising, subscription fee, syndic ation, per-unit charges, on-line sale of
merchandise, affiliate marketing
2.3 How do you set pricing for advertising at your websites (e.x. PPI, PPC, PPA)? Is
affiliate marketing the main source of revenue at t hese websites? Which type of
pricing do you prefer?
Hints: time period, exposure, per response, per act ion (PPS, PPL, PPC), other…
2.4 What business models (main source of revenue) d o you use at your websites? At
which of them affiliate marketing is beneficiary? W hy?
Hints: Brokerage (exchange, auction), Agent (selling , metamediaries, malls, purchasing,
reverse auction, buyer cooperative), Online retaili ng
2.5 How do you attract visitors to your website? Ho w do visitors enter your website?
Do you use search engine marketing or search engine optimization?
Hints: Traffic acquisition, SEM, SEO
2.6 Do you use one-to-one or one-to-many affiliate programs? Which one do you
prefer?

APPENDIX : INTERVIEW GUIDE
– 89 – 2.7 Do you use services of affiliate networks? Do y ou prefer them to programs offered
directly by a merchant? Why?
Research Question 3
How do content providers select affiliate marketing programs they intend to participate in?
3.1 Why did you decide to use affiliate marketing?
3.2 What criteria do you take into consideration? W hy these?
3.4 Do you make any pre-selection of the programs b efore making the final choice?
How?
3.5 How do you choose the particular program, you a re going to use?
3.6 Is the selection different, if you choose a pro gram for a specific website (place at
the website) that you have already used affiliate p rogram at?
3.7 Under what circumstances do you decide to chang e the affiliate program?

APPENDIX : SCREENSHOTS
– 90 –
Screenshots

Figure 22 – Case 1: letenkar.cz

APPENDIX : SCREENSHOTS
– 91 –

Figure 23 – Case 1: eurovikendy.cz

APPENDIX : SCREENSHOTS
– 92 –

Figure 24 – Case 1: i-noviny.cz

APPENDIX : SCREENSHOTS
– 93 –

Figure 25 – Case 1: birmingham.pl

APPENDIX : SCREENSHOTS
– 94 –

Figure 26 – Case 1: manchester.cz

APPENDIX : SCREENSHOTS
– 95 –

Figure 27 – Case 2: loanexplorer.co.uk

APPENDIX : SCREENSHOTS
– 96 –

Figure 28 – Case 2: remortgageexplorer.co.uk

APPENDIX : SCREENSHOTS
– 97 –

Figure 29 – Case 2: debtexplorer.co.uk

APPENDIX : SCREENSHOTS
– 98 –

Figure 30 – Case 2: shoppingday.co.uk

APPENDIX : SCREENSHOTS
– 99 –

Figure 31 – Case 3: jaknaweb.com

APPENDIX : SCREENSHOTS
– 100 –

Figure 32 – Case 3: schuti.cz

APPENDIX : SCREENSHOTS
– 101 –

Figure 33 – Case 3: vydelek-online.cz

APPENDIX : SCREENSHOTS
– 102 –

Figure 34 – Case 3: nakuptesi.cz