OTOinsights "Implications for User Engagement with Search Engine Result Pages"

ImplIcatIons of User engagement

wIth search engIne resUlt pages

JeremI Karnell • Dan BerlIn • greg slama

as consumer behavior continues to rapidly migrate to an on-demand model where

the customer controls the purchase cycle, search engine usage continues to rise as a

primary means of web navigation. with the massive increase in multimedia content

that is uploaded, stored, and consumed online, major search engines have continued

to evolve beyond simply indexing pages of text and are now providing results

based on a full array of video, audio, and image formats. the inclusion of these

digital assets is widely known as Universal search. as Universal search becomes

a more permanent fixture within search engine result pages (serps), users’

interactions with the new generation of results should be of particular interest to

companies who are focused on generating quality search engine traffic.

It is well established that rankings on a serp have much to do with whether a user is

likely to click on a given result and higher rankings can actually have a positive effect

on a brand’s perception and ‘likeability’ (marketingsherpa, 2009). with the onset of

Universal search results, this study sought to measure the user impact of this new

generation of search engine result pages. specifically, we wanted to understand if

there was a difference in emotional engagement with the results and if that would

impact click propensity in paid and natural search.

execUt Ive sUmmary : :

Copyright © May 25, 2009, One to One Interactive www.onetooneinteractive.com 1

oBJectIves for thIs stUDy InclUDe the followIng:

To compare user engagement when searchers are exposed to universal search

stimuli vs. more traditional text-only results and provide direction for future

media optimization:

• Identify the potential impact of search engine evolution on

• future search engine marketing (SEM) campaigns

• Determine changes in user behavior when presented with various

• digital assets consistent with current Universal search parameters

• Asses the viability of Universal search becoming a more

• prominent part of SERPs

• Use preliminary data to inform a formal study to glean more

• quantitative and actionable insights

QUalItatIve caveat

Although consistencies in the data lend confidence to the analysis, the results of

this research must be viewed as directional rather than conclusive. This study raises

more questions than answers, but provides the groundwork for future studies.

InsIghts

• SERPs which include Universal Search results reinforce eyeballs to

• stay focused on the first page’s top paid and natural search results

• The inclusion of Universal Search results increase emotional

• engagement during interactions with a SERP

stUDy oBJect Ive : :


PAID SEARCh

Bid

Quality Score

Click Through Rate (CTR)

NATURAl SEARChRelevance of content

Relevance of content

Keyword use it Title attributes

Site architecture

link popularity (anchor text & topical relevance)

Domain age

fIgUre 1: Search Ranking Factors

Currently, one of the most effective and efficient ways to search for digital content is

through the use of a search engine such as Google or Yahoo. In fact, a recent Forrester

report indicates that search engines are the primary method of finding Web sites: 72%

of adults said that they use search engines to find Web sites (li, 2008). These search

engines rank Web site pages and ads based on various criteria to determine relevance

to a particular search (Figure 1).

We are now moving into a new generation of search optimization where digital assets

beyond pure text come into play. Introduced by Google in May 2007, Universal Search

attempts to index, categorize, and display the most relevant information for a search

query regardless of format (audio, video, images, news, local results, and other rich

content). This search method is quickly replacing the traditional text only results and

is constantly adjusting to better conform to user behavior and intent. This study serves

to highlight the impact that Universal Search results have on consumer engagement

and click behavior. Our findings underscore the need for marketers to consider the

optimization of all digital assets to maintain a competitive advantage as we move to

the next generation of SEM.

IntroDUct Ion_serps & seo : :


Seventeen participants who were interested in changing their cable service were

recruited by a third-party recruiter. The participants were invited to OTOinsights’

Quantemo™ lab for a one-on-one session. They were first asked to choose five

search terms from a list of ten terms that may be used by individuals who are

looking to change their cable service (Figure 2).

For each of these search terms, four Google SERPs were created: one with only the

natural links (FIGURE 3: Natural Only), one with the natural links and paid links

(FIGURE 4: Natural w/ Paid), one with only the Universal links (FIGURE 5: Universal

w/o Paid), and one with the Universal and Paid links (FIGURE 6: Universal w/ Paid)

(see page 4 for FIGURES 3-6).

The SERP was presented for 1 minute while the participants’ physiological traces were

recorded and calculated using OTOinsights’ Quantemo™ Engagement Index (QEI).

Additionally, the participants’ eyes were tracked using a Tobii 1750 eye tracker. After

the minute, participants were asked which links they would click first, then second.

They were then instructed to fill out the surveys which comprise the emotional

component of the QEI.

1 PATENT PENDING

Cable Internet

Cable Television

Comcast

DirecTV

Comcast on Demand

FIGURE 2: Study Search Terms

Internet Service

Comcast DVR

hD Service

ComcasT Triple Play

Verizon FiOS


se arch engIne resUlts page (serps) : :

FIGURE 3: Natural Only

FIGURE 5: Universal w/o Paid

FIGURE 4: Natural w/ Paid

FIGURE 6: Universal w/ Paid


While viewing the Web pages, participants were connected to

OTOinsights’ Quantemo™ neuromarketing research system. Quantemo™

simultaneously records biophysical signals in addition to eye and click

tracking information.

The recorded biophysical measures are combined into a single representative

measure of physiological engagement: the Quantemo™ Physiological Index, or

QPI. The QPI serves as a single point of reference for the overall level of physical

engagement (or disengagement) exhibited by a research participant. higher

QPI scores represent stronger physiological engagement, while negative QPI

scores represent weaker physiological engagement.

After viewing a page, the participant fills out two surveys: the Geneva Emotion

Wheel (GEW) and an Aesthetic likert (1-5) scale. Developed by the researchers at

the Swiss National Research Center in Affective Sciences, the GEW is designed to

obtain self-report information on a wide range of felt emotions elicited by a

particular event (in the case of this study, observing a Web page) (Scherer, 2005).

The QPI, aesthetic score, and emotional descriptor scores are combined to form

the Quantemo™ Engagement Index, or QEI. Calculating the QEI produces a single,

representative and holistic measure of user engagement that allows researchers to

correlate the objective physiological data of the QPI with the subjective, self-report

data of the ratings and emotion scores.

QUantemo™ engagement InDex : :


Ins Ights : :

SERPs which include video and image results reinforce eyeballs to stay focused

on the first page top paid and natural search results.

The “Universal w/ Paid” SERP is what visitors will encounter when utilizing search

engines. As such, we examined the eye-tracking and click results from this SERP

type to determine where each user first looked and clicked. It is no surprise that

the top paid links received a large number of first fixations and clicks

(see Figures 7 and 8, respectively).

however, notice the location of the video results in both graphs. The video was

typically fixated upon before both the first natural and the right paid links. The video

link also received more clicks than the multilinks and the second natural link. These

findings have great implications for those companies who are already optimizing their

Web site: it may not be enough to optimize just the text. For example, consider the

Google multilinks which are an indented list of site topics determned by the search

engine when a site is ranked in the first natural position. Figure 8 shows that the video

results are likely to receive more clicks than the multilinks.

UnIversal w/ paID search clIcKs

FIGURE 8: Universal with Paid

12

10

8

6

4

2

01st

lInK

% o

f p

ar

tIc

Ipa

nt

s

top paID

other lInK

vIDeo rIght paID

mUltI-lInKs

2nDlInKs

news

7

3

6

3

1

4

2

3

1

3

1

2

1

1

1

Image

1

2

1st click

2nd click

UnIversal w/ paID search fIrst fIxatIon

fIgUre 7: Universal with Paid First Fixations

45%

40%

35%

30%

25%

20%

15%

10%

5%

0%

% o

f p

ar

tIc

Ipa

nt

s

39%

22%

17%

11%

6% 6%

top paID

page top

vIDeo 1st

lInKrIghtpaID

1st

pIctUre


Figures 9 and 10 show the first fixation and clicks for the “Universal w/o Paid” SERP,

respectively. The importance of optimizing media is also highlighted from these results.

Notice that the image links received more first fixations and clicks than the multilinks

and the second natural link. The advent of Universal search demands that companies

optimize not only their site text, but also their media to ensure that their images, videos,

and other content are properly indexed and show up in Universal search results.

There were also interesting differences in the participants’ gaze patterns on the

different SERPs. Figures 11 and 12 show 30 seconds of eye-tracking for a “Natural w/

Paid” and a “Universal w/ Paid” SERP, respectively.

On some of the pages without Universal results (those without images or videos), the

gazes seemed to follow Nielsen’s F-pattern (Figure 11) (Nielsen, 2006). however, on the

pages with Universal results, the gaze pattern seemed to follow the ‘golden triangle’

identified by Enquiro (Figure 12) (MarketingSherpa, 2009). Additionally, notice the

drop-off in views below the images in Figure 12. The images seem to keep the users’

eyes near the top of the page. These findings align with the second insight of the study:

Universal results were more engaging to the participants. Since the images and videos

caused engagement, the participants were less likely to explore the rest of the SERP.

fIgUre 10: Universal Clicks

UnIversal

fIgUre 9: Universal First Fixations

35%

30%

25%

20%

15%

10%

5%

0%% o

f p

ar

tIc

Ipa

nt

s

29%

25% 25%

13%

8%

1st

lInKpage top

Image mUltI-lInKs

vIDeo

% o

f p

ar

tIc

Ipa

nt

s

UnIversal clIcKs

20

18

16

14

12

10

8

6

4

2

0other lInK

1st

lInKImage mUltI-

lInKs2nD

lInK

8

11

9

26

2

2

21

vIDeo zIp coDe

232

1st click

2nd click


fIgUre 11:Natural w/ Paid Eye Tracking

fIgUre 12:Universal w/ Paid Eye Tracking


Videos and image results increase engagement on a SERP

Utilizing the QEI as an overall measure of engagement,

we found that the SERPs with images and videos were

more engaging to the participants. Figure 13 shows the

results of the physiological and emotional engagement

data produced by the QEI. The results in the middle

show the QEI for the Universal results, which had a

higher engagement level than those without Universal

results. This is not surprising given that images and

videos will give the user ‘instant gratification’ with a

single gaze. That is, instead of having to read text, the

user will quickly process the images and videos, increasing

engagement. These findings underscore the impact that

brand marketing has on search marketing strategies.

fIgUre 14: 28 Second Gaze Plot

QUantemo™ engagement InDex

fIgUre 13: Engagement levels for 4 SERP Types

0.70

0.60

0.50

0.40

0.30

0.20

0.10

0.00

lo

we

r

en

ga

ge

me

nt

0.07

0.37

0.08

natUral only

UnIversalw/ paID

natUral w/ paID

0.44

UnIversalonly

hIg

he

re

ng

ag

em

en

t

Results had images and videos

Copyright © May 25, 2009, One to One Interactive www.onetooneinteractive.com

A prime example of the salience of images, particularly brand logos, is evident from the gaze plot in Figure 14. This

figure shows 28 seconds of eye tracking for a particular participant. Notice the amount of time that the participant

spends on the Verizon and Comcast logos in the image area. It took the participant 50 fixations to get to the images,

but then spent 38 fixations on the images themselves. There is no doubt that these logos were interesting to the

participant. This has the implication that further research is warranted to determine if SERPs with brand images

result in a higher propensity to make a purchase.

11

ImplIcatIons for marKeters

• Creating a holistic strategy which considers all search interactions

• will provide a competitive advantage and drive more engaged visitors

• through your SEM activities

• Begin to think about current digital assets and how they can be

• leveraged to gain additional reach with the search engines

• Top Paid results are the highest focal point within typical search

• results and provide a great branding opportunity

• It will continue to be important for brands to maintain high positions

• within Paid and Natural results to generate click volume and increase

• brand perceptions

• Natural Search top results will continue to be the primary driver

• of search engine traffic and should be a major focus on optimization

• initiatives

ImplIcatIons for fUtUre stUDIes

• Does media in searches change users’ propensity to purchase?

• Does brand exposure on a SERP change users’ engagement and

• click behavior?

• Do positional changes in non-text results impact users’ click

• propensity and engagement?

conclUsIons

The findings from this study highlight the importance of optimizing media (video &

images) for companies who want to drive traffic to their Web site from search engines.

Traditional SEO efforts may help get links in the top few positions of the natural results.

But, the Universal search paradigm demands that companies pay closer attention to

their optimization of digital assets to increase the likelihood of inclusion in search

results. This study provides evidence that videos and images are not only changing

a user’s first fixation, but that these media formats positively increase the user’s

emotional engagement during the time they were reviewing the Universal SERP. This

underscores the importance of branding and the strategic role it should play in search

engine marketing efforts.

Copyright © May 25, 2009, One to One Interactive www.onetooneinteractive.com

•

12

li, C. (2008). how Consumers Find Web Sites. Cambridge, MA:

Forrester Research, Inc.

MarketingSherpa. (2009). 2009 Search Marketing Benchmark Guide.

Warren, RI: MarketingSherpa llC.

Nielsen, J. (2006, April 17). F-Shaped Pattern for Reading Web Content.

Retrieved March 19, 2009, from Jakob Nielsen’s Alertbox:

http://www.useit.com/alertbox/reading_pattern.html

Scherer, K. (2005). What Are Emotions? And how Can They Be Measured?

Social Science Information , 44 (4), 695-729.

worK s c IteD : :


amplIfyIng engagement

New knowledge about human behavior brought to light by social and neuroscience

has fundamentally called into question the old mental models of how advertising

and marketing work. Gone is the notion that consumers make decisions in a

linear think-feel-do way and behavior is guided by rational-only principles. Instead,

memories, emotions, associations, and thoughts play a primary role in how

individuals relate and ultimately engage with brands.

OTOinsights is a primary research offering that is breaking new ground in neuro-

marketing to offer clients advanced and scientific levels of insights into how their

consumers engage with them across the landscape of new media channels.

To learn more about OTOinsights, visit www.otoinsights.com

complete one-to-one solUtIons for BranDs, agencIes, anD pUBlIshers

OTOinsights is a One to One Interactive company. Established in 1997, One to One

Interactive is the first enterprise to assemble a complete solution for brands, agencies,

and publishers executing one-to-one marketing strategies. By bringing together one

of the nation’s leading digital marketing agencies, the world’s most comprehensive

portfolio of permission marketing platforms, unique performance-based social media

networks, and cutting-edge neuromarketing research techniques, the companies

of One to One Interactive build informed and creative customer/constituent strategies

on the belief that digital media’s ability to enable engaging one-to-one dialogues is

the future of marketing.

To learn more about One to One Interactive, visit www.onetooneinteractive.com

OTOinsights

529 Main Street, Charlestown, MA 02129

617.425.7300

www.otoinsights.com

[email protected]


OTOinsights "Implications for User Engagement with Search Engine Result Pages"

Technology

search method

search query

viability of universal

universal search stimuli

universal search attempts

search engine usage

particular search figure

study search termsfor