Content uploaded by Chris Barry
Author content
All content in this area was uploaded by Chris Barry
Content may be subject to copyright.
A Study of First Click Behaviour and User
Interaction on the Google SERP
Chris Barry
1
and Mark Lardner
2
Abstract Firms use Search Engine Marketing (SEM) to drive users to their Web-
site. Some are prepared to pay for placement; others use Search Engine Optimiza-
tion (SEO) hoping their result percolates up the organic SERP. Despite extensive
SEM efforts, firms can only speculate over the first critical interaction between the
first SERP and a user’s first click. This study sheds some light on users’ first click
behaviour on Google and the early interaction thereafter. The research reveals that
users evaluate the SERP from the top downwards, deciding instantly whether to
click into each link, while first clicks are predominantly at the top of the SERP,
especially towards organic links. For certain queries top sponsored links received
almost as many clicks as organic links despite what users profess. Recommenda-
tions to firms include advice that strategies should be primarily SEO focused and
that paid search campaigns should maintain a position in the top sponsored links
section of the Google SERP.
1 Introduction
Firms use search engine marketing (SEM) to drive users to their Website. There
are two main strategies: paid search and search engine optimization (SEO). The
two strategies are not mutually exclusive and various arguments can be made for
pursuing each strategy (Barry and Charleton, 2008). Whichever SEM strategy is
chosen, competition for a high search engine results page (SERP) ranking is in-
tense. Firms can make critical decisions about SEM strategy if they have insight s
into early user interaction. This study seeks to provide such insights on the SERP
of the predominant search engine (Lewandowski, 2008), Google.
1
Chris Barry
National University of Ireland Galway, chris.barry@nuigalway.ie
2
Mark Lardner
National University of Ireland Galway, m.lardner1@nuigalway.ie
2 Chris Barry and Mark Lardner
2 The Nature of Search Online
Searching online refers to the input of a query (a list of one or more keywords)
into the text box of a search engine followed by a list of results returned on the
SERP (O’Brien & Keane, 2006). The abundance of content online makes firms
ever more reliant on the search engine optimally indexing and ranking their Web
pages so they are found in the earliest part of a user’s interaction. Search engines
determine these ranks through their own typically undisclosed ranking algorithms
(Bar-Ilan, 2007). Failing to find what they want after only the first click could cre-
ate negative perceptions of the search engine and the chosen Website and possible
reputational damage. Furthermore, search engine algorithms are increasingly more
impressed when users actually click on a link for any given query (Ciaramita,
Murdock and Plachouras, 2008). It is also becoming clearer how little time users
spend making judgments on the relevancy of links and how easily distracted or
impatient they can be (Browne, Pitts and Wetherbe, 2007; Szetela, 2008).
A SERP consists of two main areas: the visible area and the scrolling area
(Hochstotter and Lewandowski, 2009). Information that is immediately seen on a
screen is the visible area of the SERP while the scrolling area must be scrolled to.
In Hochstotter and Lewandowski’s study the SERP links that occur ‘below the
fold’ (i.e. links only visible when the user scrolls) are very rarely clicked upon.
This increases the pressure on firms to ensure their site appears above the fold. An
important characteristic of user behaviour is that very few users actually move on
to the second SERP after performing a search query (Bar-Ilan, Levene, Mat-
Hassan, 2006; Richardson, Dominowska and Ragno, 2007). Furthermore, a recent
report found that 68% of all users concentrate their clicks on the first SERP
(iProspect, 2008). This makes competition for a high ranking very intense and
vulnerable to ranking fluctuation at all times.
The results that are returned by a search engine with any given query are di-
vided into two sections; organic listings and sponsored listings (Jansen & Resnick,
2005; Jansen & Molina, 2006). Organic links are links that the search engine’s al-
gorithm deems most relevant to the query, whilst sponsored links are links that
appear as a result of a party purchasing the right to be displayed on the SERP for a
particular keyword (Jansen & Resnick, 2005).
Barry and Charleton (2008) report that SEM strategies are the predominant
methods firms use to promote their visibility and that the primary objective of an
SEM strategy is to ensure that one’s site is displayed at or near the top of the first
SERP. Paid search allows firms or organisations to place bids on certain keywords
and hence have their advertisements displayed along side or above the organic list-
ings as sponsored listings (Bar Ilan, 2007). The higher ranked sponsored listings
typically cost more. Payment may also be made for a click through to a Website.
SEO involves altering a Web page so that it appears to be more relevant to the
search engines. SEO aims to exploit search engine algorithms to attract users to
specific Websites presented as organic results (Barry and Charleton, 2008). SEO
3
is carried out by altering the aspects of the Website (such as relevant keywords in
the metadata of a page) that a search engine’s ‘spider’ is thought to explore in
judging relevancy (Jones, 2008). In a recent study of Irish small to medium enter-
prises (SMEs), it was found that the majority of SMEs show a preference for SEO
over paid search (Barry and Charleton, 2008).
One of the chief methods that Google uses to evaluate the full extent of the
relevancy of a page is through a metric called ‘PageRank’ which helps measure
the popularity of a page (Cho & Roy, 2004). The page popularity is determined by
how many other Web pages link to a specific page on the Internet. A simplified
example - Google will present a hyperlink at the top of its SERP if that hyperlink
contains the relevant keywords in its metadata and also if that hyperlink is linked
to by more external sites than other hyperlinks with similar keywords (Cho &
Roy, 2004). Google uses complex algorithms to rank the results organically on the
SERP. The highly optimised pages (via SEO) are those more likely to appear on
the first SERP (Jones, 2008).
3 Evaluation of the SERP
SERPs are presented in some form of text based summaries and based on the in-
formation contained within these summaries users make relevance judgments
based on what links best suit their information needs (Jansen & Resnick, 2005;
O’Brien and Keane, 2006). The question of how users evaluate the relevance of
the SERP can be answered through the investigation of users’ first click behav-
iour. First click behaviour describes “the process of when a user (a) poses a query
to a search engine to fulfill some information need. (b) evaluates the results list re-
turned to that query and (c) then chooses one of these results as a link to follow”
(O’Brien, Keane and Smyth, 2006, pp1)).
Klockner, Wirschum and Jameson contend users either adopt a depth-first or a
breadth-first approach (2004). A depth-fist approach refers to when users evaluate
each entry in the list in turn, starting from the top, and deciding systematically
whether to choose that link or to move onto the next consecutive link. A breadth-
first approach is where a user looks ahead at a number of list entries and then re-
visits the most promising ones in order to open them. In an eye tracking study
conducted by Klockner et al (2004), 65% of participants utilised the depth-first
approach with the remaining participants adopting a fully or sometimes partially
breadth-first approach.
Keane, O’Brien and Smyth (2008) showed that users do have a tendency to
choose results at the top rather than results down the bottom of a list, but this ten-
dency will not be as prevalent when the relevancy of the top result is weakened.
Joachims, Granka, Pan, Hembrooke and Gay (2005) contend that users’ trust in
the search engine mean links ranked first received many more clicks than the fol-
lowing link and also that the top two links received more attention than the rest of
4 Chris Barry and Mark Lardner
the SERP. Opinions are divided as to whether a click on a link is representative as
a vote of relevance. Joachims et al (2005) believe that a click on a hyperlink from
a user on the SERP is representative of an endorsement of that page. On the other
hand, Ciaramita et al (2008) believe that clicks on particular links are not an indi-
cation of absolute relevance, but only indicates that the clicked item is perceived
to be more relevant than other links that were ranked elsewhere but not actually
clicked.
While users have been shown to examine both organic and sponsored listings,
organic links are ultimately perceived as being more relevant (Jansen, Brown and
Resnick, 2007), even if sponsored links are reported to be just as relevant as or-
ganic links (Jansen and Resnick, 2006). Laffey also shows that with Google spe-
cifically, users are more likely to click on organic links, rather than sponsored
links (2007). Users have also expressed negative emotional reactions to sponsored
search listings (Marable, 2003). Most search engines including Google have a
vested interest in hoping that sponsored search has a future since their business
model depends significantly on it. Previous research also emphasises the impor-
tance of appearing high in the rankings for sponsored listings (Richardson et al,
2007).
Queries are often divided into transactional, informational and navigational
queries. Informational queries are queries seeking images, songs, videos or docu-
ments while transactional ones are queries with a commercial motive (purchasing
a specific product or service) and navigational-based queries are entered to take
users to a specific URL (Jansen, Booth and Spink, 2008). Previous studies have
shown that the majority of users use search engines to find out specific pieces of
information or to ask questions (Rose and Levinson, 2004; Zhang, Jansen and
Spink, 2009). Jansen et al provided some useful insights into user perception of
sponsored and organic links for transactional queries (2007): sponsored links are
likely to be more relevant than organic links for e-commerce queries; top listed
sponsored links are more relevant than side sponsored; and the rank of side-
sponsored links are not correlated with the relevance of sponsored links. A recent
study showed that 80% of Web search queries are predominantly informational
with the remaining 20% of queries split between navigational based and transac-
tional-based queries (Jansen et al, 2008).
4 Research Objectives and Methodology
Despite extensive SEM efforts, firms can only speculate about the first critical in-
teraction between the return of the first SERP and a user’s first click. Thus the
main objective of the study is an investigation of how Web searchers perceive and
interact with the Google SERP for informational and transactional queries from a
first click behavioural perspective. Secondary objectives are: to explore the nature
of user’s link assessment strategy for both informational and transactional queries;
5
to reveal how favourably users view organic versus sponsored links; and to iden-
tify where, within the Google SERP, are first clicks most concentrated.
In order to reveal a rich picture, both quantitative and qualitative research
methods were used. Three research techniques were used: verbal protocols, inter-
views and Web based questionnaires. Verbal or ‘think aloud’ protocols refer to the
practice of collecting data by getting participants to think aloud as they partake in
a set of specific tasks (Crutcher, 1994). Verbal protocols were used to examine
how users perceived and interacted with the Google SERP from a first click be-
haviour perspective, interviews to examine in more detail the issues and concerns
arising from the verbal protocol sessions, and Web based questionnaires were ad-
ministered in order to capture necessary participant demographics. Each partici-
pant, under controlled laboratory conditions, was asked to respond to the most
common types of Google searches - one using an informational query and another
using a transactional query (Jansen et al, 2008). These types of queries were used
since research shows that they are the two most common types of queries (Zhang,
Jansen and Spink, 2009). Both tasks were very specific, achievable and easy to
explain to a novice user. All verbalisations and on screen behaviour (e.g., clicks,
scrolling, mouse movements) were recorded by an open source computer pro-
gramme called ‘CamStudio’. The combination of verbal protocols with a simulta-
neous recording of on-screen activities has shown to be a valuable observational
method, previously used by Van Waes (1998). For the study, a convenience sam-
ple of twenty participants was used.
5 Research Findings
5.1 Link Assessment Strategy Analysis
A variant of the depth and breadth-first approaches of Klockner et al (2004) to
analyse first click behaviour was used. This more sophisticated analysis identifies
usage patterns that characterise each approach and combines verbalised thoughts
to more fully understand user behaviour. These patterns are identified in Table 1
below and represented in Figure 1.
In the informational query sessions, the variant depth-first approach was sig-
nificantly more evident with 17 out of the 20 participants characteristically using
it, without any scrolling action before the first click. Each first click was on the
first organic link and verbalisations made in tandem with the first clicks included:
“…guess I should go for the first one” and “First up is Wikipedia which I usually
use. I’m going to try that one first”. There was little hesitation and behaviour was
hurried. The action appeared a somewhat automated response, entrusting the first
6 Chris Barry and Mark Lardner
result on the SERP with a confident assertion of relevancy. Depth-first was also
more common for the transactional session with 13 participants utilising the ap-
proach. Depth-first approaches came in a mix of sequential and non-sequential
click-through patterns on the SERP. Some participants evaluated the list sequen-
tially, e.g. “…I’m going to go down to the next ones…there’s Amazon (scroll),
Play (scroll), there’s a Dublin one so I’m going to go for that one”. Other depth-
first instances were more instantaneous “…Amazon is the first one I’m going to”.
While search sessions did take longer for participants to complete than for the in-
formational query, the hurried depth-first approach was still more regular.
Table 1 Usage Patterns of Link Assessment Approaches
Overall, instances of breadth-first search were far fewer for both the informa-
tional and transactional sessions. Only three participants exhibited the breadth-first
approach for the informational search. One participant performed a partial scroll
of the SERP before returning to the top link, stating: “…Wikipedia is your best
bet”. The lack of a breadth-first strategy when informational searching online
could be indicative of impatience, consistent with suggestions of Browne et al
(2007) and Szetela (2008) above. There were notably more breadth-first usage for
the transactional query. Breadth-first search is exhibited through the following il-
lustrative verbalisations: (a) “I’m scrolling down to see if there is any link that
immediately catches my eye…”; (b) “I’m going to scroll down to see if there is
anything else that looks good before I click into that. There’s lots of reviews and
stuff, not looking for that. I’m going to go for Amazon.com”; and (c) “OK, any of
these I recognise? Don’t know if they’re good…hmmm… (Scrolls to bottom). See
do I recognise any first before I move on…”. Another participant displayed signs
of breadth-first search by musing “E-bay, Amazon (third top sponsored and first
Link Assessment Approach
Variant Depth-First
Variant Breadth-First
Characterised by:
An immediate first click without any par-
tial/full scroll
Sequential/non-sequential click-through
pattern starting from top to bottom of
SERP
Verbalisations (e.g. “First thing I’ll do is
click the top one”)
Behaviour appears more impatient and
rushed
Characterised by:
First click being preceded by partial or full
scroll through SERP (Then a retreat up the
SERP)
Participant placing cursor over one or more
links momentarily before retreating up the
SERP for first click
Verbalisations (e.g. “I’ll just look through
the list first to see what else there is...”)
Behaviour appears more thorough and
aware
7
organic links respectively) kind of stick out to me anyway…sure I’ll look at the
first one…”.
Fig. 1 First Click Behaviour: Incidence of Depth-First and Breadth-First Link Assessment
Strategies
= One Participant
Informational
(Query)
Transactional
(Query)
Depth-First
(Link Assessment)
Partial/Full Breadth-First
(Link Assessment)
These combined results show that most participants (75%) are not prepared to
pour over the entire SERP in detail, instead preferring to adopt a more hurried
depth-first approach whereby the user makes more spontaneous and rushed judg-
ments when deciding where to click first. This finding also correlates with the
study by Klockner et al (2004) who found that the depth-first approach was used
65% of the time, with the remaining users adopting a breadth-first search.
5.2 User Perceptions of Organic and Sponsored Links
The click data collected from the verbal protocol sessions produced some interest-
ing results. As mentioned above all participants chose the top organic link as their
first click for the informational query. In the absence of sponsored links, partici-
pants tend to gravitate towards the first organic link, irrespective of the link as-
sessment strategy. Of all the first clicks carried out on the transactional session 11
(55%) of the first clicks were on organic links, confirming a disposition towards
the use of organic links (Laffey, 2007). In this case, the first organic link received
the most clicks.
Anomalously, despite nine participants (45%) choosing a sponsored link as a
first click, verbalisations and post-test feedback indicated that sponsored links are
not deemed relevant. One participant suggested that the sponsored links are auto-
matically ignored: “…your train of thought is to automatically go to the first link
up here below the sponsored links". Other verbalisations also show an explicit dis-
regard for sponsored links: “…they’re sponsored links. I never go for them, so I’m
going to go down to the next ones”. Sponsored links were also being viewed as
having “…catches in them.” and not being “…your best bet for getting a deal”,
almost a hindrance. This view concurs with the finding of Jansen et al (2007) that
8 Chris Barry and Mark Lardner
stated that there were no strong relevance correlations with sponsored links. Con-
ceivably, the anomaly might be explained by some confusing the organic and
sponsored areas. A number of participants did not seem to distinguish the top
sponsored links section as being separate to the organic links with one participant
referring to the first organic link as the “…fourth one on the list”. Another par-
ticipant expressed difficulty differentiating between organic and sponsored links.
In contrast, organic links are viewed far more favourably, described by one as
the “…normal area…”, another commenting “I’d prefer to use the non sponsored
links because they’re more impartial…”. Consistent with other studies participants
appeared to find organic links more relevant than sponsored links (Hotchkiss et al,
2004; Jansen and Resnick, 2005; Jansen and Molina, 2006; Jansen et al, 2007).
The actual search engine behaviour of users and the comments in the post-test
interview, that somewhat contradict their previous action, could be explained by
the concepts of espoused theory and theory-in-use (Argyris, 1980). It may be the
case that users overtly subscribe to some notion of the purity of organic search
(espoused theory), but in the hurried world of online search, they are often pre-
pared to compromise by utilising all available links (theory-in-use).
All first clicks were concentrated entirely on top sponsored links rather than
right-hand sponsored links. This correlates with research from Jansen et al (2007)
that found no real relevance correlation with right-hand sponsored links. This find-
ing is also backed up with a remark from one participant who noticed that one link
in the right-hand sponsored section was a dating site as opposed to an e-commerce
based site for DVDs. Right-hand sponsored links received far more negative atten-
tion than top sponsored links throughout the transactional session. They were de-
scribed as “…just advertisements that I don’t want to be wasting my time check-
ing” and “…untrustworthy…”. Another participant remarked dismissively that
“…if they were any good they would be in the main search”.
5.3 First Click Distribution on the SERP
For the informational query all participants’ first clicks were carried out on the
first organic link. This finding is consistent with previous research where it was
found that links ranked first in the SERP, received many more clicks than the fol-
lowing link on the SERP (Joachims et al, 2005). The same research also found that
users pay most attention to the first and second links on the SERP.
In respect of transactional queries, 95% of all first clicks occurred between the
top sponsored link and the third organic link (within the visible area of the SERP),
again consistent with Joachims et al (2005). The pattern between these links is that
the first top sponsored link received four more first clicks than the next link and
the first organic link received five more first clicks than the following organic
link. This emphasises the importance of ensuring that SEO secures a place in the
‘visible’ area of the screen (Hochstotter and Lewandowski, 2009).
9
Verbalisations from the interviews also reveal a tendency to gravitate towards
the top links on the SERP. “I usually just go to the one on the top of the page…if it
doesn’t suit me I just go next one down”. Some participants display a lack of pa-
tience when describing their online search habits: “If I don’t get it, at a push, in
the first five links, more often than not I just give up, I just go and type in some-
thing else”, consistent with the notion that it only takes one second for a user to
make a relevancy judgment (Szetela, 2008). The apparent urge shown by partici-
pants here to act quickly on the SERP and then refine the query if no relevancy is
found at the top demonstrates that users are becoming increasingly dissatisfied
with the search engine, the longer the time they spend on it (Browne et al, 2007).
Participants did seem to possess an explicit trust in the top half of the SERP:
“…even if you’re not sure what you’re looking for, you assume that the first result
will be relevant”; “The first one that comes up is usually the most relevant”; “gen-
erally I will click on the top few”. Verbalisations like these suggest that SEO
strategies must be focused on a high SERP ranking if they are to have any chance
of winning the race for that first click. These findings and verbalisations also tie in
with previous research on this topic, that conclude that users have a tendency to
concentrate mainly on the top part of the SERP whilst paying significantly less at-
tention to the bottom part (O’Brien and Keane, 2006; Ciaramita et al, 2008; Keane
et al, 2008).
6 Conclusions
In light of the above findings and from an SEM perspective, companies would be
advised to make the achievement of a high-ranking position on the Google SERP
the primary objective of their SEM strategy. The findings demonstrate that com-
panies are at a distinct disadvantage if their link does not appear in the visible area
of the Google SERP. This distribution towards the top is also noted in other stud-
ies (Richardson et al, 2007; Keane et al, 2008). If it is accepted that users place
more trust in the top two links (Joachims et al 2005), users who utilise a depth-
first approach will execute that vital first click on the top part of the SERP. There-
fore, SEM strategies should work under the assumption that users assess the SERP
using the depth-first approach. SEM strategists should therefore take all steps nec-
essary (on site optimization and paid search campaigns) to ensure that their site
appears and stays in a high-ranking position on the Google SERP in order to in-
crease traffic to their site. Exploiting PageRank is key to achieving this strategy.
In this regard, Google could assist firms by offering more clarity on the reasons
why a link is on top of the SERP.
SEO specialists should also ensure that users who click through to their link on
the SERP are greeted with a landing page that meets the needs of the user and cor-
relates sufficiently with the apparent relevancy of the SERP link. This is critical so
that users do not feel compelled to navigate back to the SERP and click into a rival
10 Chris Barry and Mark Lardner
site (Todd, 2006; Szetela, 2008). Sites not ranked highly on the Google SERP that
are effectively less ‘popular’ need to design their SEO strategy that views the first
Google SERP as a highly competitive market where the onus is on the lower
ranked sites to justify and demonstrate to Google why their sites should be placed
at a higher ranking. This can be achieved if an SEO-focused SEM strategy is de-
signed to encompass as many of the reported factors Google use to determine
relevancy with special attention given to the PageRank metric. To achieve this, a
company needs skilled individuals to ethically implement continuous SEO
throughout the Web site.
Table 2 Behavioural Observations from Query Sessions and Verbalisations
Informational
(Query)
Transactional
(Query)
Organic
(Links)
• Heavily distributed on the
top of the organic SERP
• Deliberate depth-first ap-
proach recurrent
• SERP was quickly processed
• Participants more relaxed
• Heavily distributed on the 1
st
and 2
nd
link
• A less deliberate depth-first
strategy recurrent
• SERP processing was much
slower
• Greater awareness evident
Sponsored
(Links)
N/A
• Heavily distributed on the 1
st
and 2
nd
link
• Participant cynicism was
evident
• Greater anger/frustration
exhibited
• Less confidence shown in
sponsored links
Table 2 summarises the general behavioural observations that were exhibited
throughout the query sessions in respect of: link assessment strategy; user percep-
tions of organic and sponsored links; and the distribution of first click behaviour.
It shows that participants utilised a depth-first approach for each query, however
the depth-first approach appeared less deliberate for the transactional query than
the informational query as users exhibited heightened awareness when perusing
the SERP and processed it more slowly. Sponsored links elicited a far more cyni-
cal and negative reaction from participants overall.
Again, from an SEM perspective, companies should be aware of the implica-
tions emerging from this research. This study found some preference for clicking
on organic over sponsored links. Furthermore, from verbalisations, participants
expressed less faith, at times outright distrust, in sponsored links. In considering a
paid search campaign, this study has shown that first click behaviour is carried out
wholly on the top sponsored links. Choosing right-hand side sponsored links
would appear a near-redundant investment. The anomaly between verbalisations
11
about sponsored links and actual usage patterns is of concern. Some participants
appeared confused or uncertain regarding the actual authenticity and purpose of
the top sponsored links. While the findings put forward a possible benign explana-
tion about how people theorise they will behave in an espoused theory may differ
from how they behave in practice, their theory-in-use, the issue is primarily one
for Google to resolve by making more explicit the distinction for users.
References
Argyris, C. (1980). Inner Contradictions of Rigorous Research. Academic Press, New
York, NY.
Bar-Ilan, J., M. Levene & M. Mat-Hassan. (2006). Methods for Evaluating Dynamic Chan-
ges in Search Engine Rankings: a Case Study. Journal of Documentation, 62(6), 708-729.
Bar-Ilan, J. (2007). Manipulating Search Engine Algorithms: the case of Google. Journal of
Information, Communication and Ethics in Society, 5(2/3), 155-166.
Barry, C. & D. Charleton. (2008). Researching Search – A Study into Search Engine Mar-
keting Practices in Ireland. In Proceedings of the International Conference on e-Business,
Porto, Portugal. 26-29 July 2008, (CD-ROM).
Browne, G. J., M. G. Pitts & J. C. Wetherbe. (2007). Cognitive Stopping Rules for Termi-
nating Information Search in Online Tasks. MIS Quarterly, (31), 89-104.
Cho, J. & S. Roy. (2004). Impact of Search Engines on Page Popularity. In Proceedings of
the 13
th
International World Wide Web Conference, New York, USA, May 17-20, 2004.
ACM Press, 20-29.
Ciaramita, M., V. Murdock & V. Plachouras. (2008). Online Learning from Click Data for
Sponsored Search. In Proceedings of World Wide Web Conference. Beijing, China, 2008,
17-22 May, New York, USA, 227-236.
Crutcher, R. (1994). Telling What We Know: The Use of Verbal Report Methodologies in
Psychological Research. Psychological Science, 5(5), 241-244.
Hochstotter, N. & D. Lewandowski. (2009). What Users see – Structures in Search Engine
Results Pages. Information Sciences, 179(12), 1796-1812.
Hotchkiss, G., M. Garrison, & S. Jensen. (2004). Search Engine Usage in North America.
[Internet] Accessed: 1
st
December 2008 on the World Wide Web at
http://www.enquiro.com/research.asp
iProspect (2008). “Search Engine Marketing Research: iProspect Blended Search Results
Study” [Internet] Accessed: 1
st
June 2009 at
http://www.iprospect.com/about/researchstudy_2008_blendedsearchresults.htm.
Jansen, B.J. & M. Resnick. (2005). Examining Searcher Perceptions of and Interactions
with Sponsored Results. In the Workshop on Sponsored Search Auctions at ACM Confer-
ence on Electronic Commerce (EC’05), 5-8 June, Vancouver, BC, Canada, 1-8.
Jansen, B.J. & A. Spink. (2005). An Analysis of Web Searching by European
AlltheWeb.com Users. Information Processing & Management, 41, 361-381.
12 Chris Barry and Mark Lardner
Jansen, B.J. & P. R. Molina. (2006). The Effectiveness of Web Search Engines for Retriev-
ing Relevant Ecommerce Links. Information Processing & Management, 42, 1075-1098.
Jansen, B. (2007). The Comparative Effectiveness of Sponsored and Non-sponsored Links
for Web E-commerce Queries. ACM Transactions on the Web, (1)1, 1-25.
Jansen, B.J., A. Brown & M. Resnick. (2007). Factors relating to the Decision to Click-on a
Sponsored Link. Decision Support Systems, 44(1), 46–59.
Jansen, B. J., D.L Booth & A. Spink. (2008). Determining the Informational, Navigational,
and Transactional intent of Web Queries. Information Processing & Management, 44,
1251-1266.
Joachims, T., L.Granka, B.Pan, H. Hembrooke, & G. Gay. (2005). Accurately Interpreting
Clickthrough Data as Implicit Feedback. In Proceedings of the 28
th
International Confer-
ence on Research and Development in Information Retrieval, Salvador, Brazil, 15-19 Au-
gust, SIGIR ’05, ACM, New York, NY, 154-161.
Jones, R. (2008). SEO Site Structure 101. [Internet] Accessed: 22
nd
December 2008 at
http://searchenginewatch.com/3632183
Keane, M. T., M. O’Brien & B. Smyth. (2008) Are People Biased in Their Use of Search
Engines? Communications of the ACM, 51, 49-52.
Klöckner, K., N. Wirschum & A. Jameson. (2004). Depth and Breadth-first Processing of
Search Results List. In CHI ‘04 Extended Abstracts on Human Factors in Computing Sys-
tems, Vienna, Austria, April 24-29, CHI ‘04. ACM, NY, 1539-1539.
Laffey, D. (2007) Paid Search: The Innovation That Changed The Web. Business Horizons,
50, 211-218.
Lewandowski, D. (2008). The Retrieval Effectiveness of Web Search Engines. Considering
Results Descriptions. Journal of Documentation,
64(6), 915-937.
Marable, L. (2003). False Oracles: Consumer Reaction to Learning the Truth About How
Search Engines Work, Results of an Ethnographic Study, Consumer WebWatch, Research
Report.
O’Brien, M., M. T. Keane & B. Smyth. (2006). Predictive Modelling of First-Click Behav-
iour in Web-Search. In Proceedings of the 15
th
international Conference on World Wide
Web, Edinburgh, Scotland, May 23-26, ACM, NY, 1031-1032.
Richardson, M., E. Dominowska, R. Ragno. (2007). Predicting Clicks: Estimating the
Click-Through Rate for New Ads. In Proceedings of the 16
th
Intl Conference on World
Wide Web, Banff, Alberta, Canada, May 08-12, ACM, NY, 521-530.
Szetela, D. (2008). PPC Landing Pages: PPC Visitors have ADD Accessed: 19
th
December
2008 at http://searchenginewatch.com/3632150
Todd, M. (2006). Getting High Traffic from Search Engines is Wasted on Poor Sites. New
Media Age, June 29, 15-15.
Van Waes, L. (1998). Evaluating On-line and Off-line Searching Behavior Using Thinking-
Aloud Protocols to Detect Navigation Barriers. In Proceedings of the 16
th
Annual interna-
tional Conference on Computer Documentation, Quebec, September 1998, 180-183.
Zhang, Y., B. J. Jansen & A. Spink. (2009). Time Series Analysis of a Web Search Engine
Transaction Log. Information Processing and Management, 45(2), 230-245.
