ArticlePDF Available

Abstract and Figures

Attaching an ad banner on a clip in a video-sharing website such as YouTube has become common although eye-tracking studies have concluded that this fails to secure visitors' attention. To date, there have been no studies verifying whether ad banners on a video clip can ensure eye fixation from viewers. Through eye-tracking, this study investigates whether YouTube visitors fixate on ad banners, what the correlations between fixation duration on banners and overall fixation counts are, and the extent to which site visitors are able to recall details of ad banners and of the clip viewed. Using a Miramatrix eye-tracker to record YouTube viewers' eye movements, this study showed that nearly all fixated at least once on an ad banner in a clip. However, less than 10% were able to correctly recall the ad content viewed. Nevertheless, about half of viewers were able to correctly recall clip details. Fixation duration on the banner and fixation counts on the clip are negatively correlated, but the relationship between fixation duration and counts on the banner was insignificant. This study sheds new light on YouTube advertising through the use of eye-tracking and advises advertisers to be attentive in selecting clips on which ad banners will appear.
Content may be subject to copyright.
International Journal of Electronic Commerce Studies
Vol.7, No.1, pp. 49-76, 2016
doi: 10.7903/ijecs.1404
Chatpong Tangmanee
Chulalongkorn University
Chulalongkorn Business School, Chulalongkorn University, Bangkok,
Thailand, 10330
Attaching an ad banner on a clip in a video-sharing website such as
YouTube has become common although eye-tracking studies have
concluded that this fails to secure visitors’ attention. To date, there have
been no studies verifying whether ad banners on a video clip can ensure eye
fixation from viewers. Through eye-tracking, this study investigates whether
YouTube visitors fixate on ad banners, what the correlations between
fixation duration on banners and overall fixation counts are, and the extent
to which site visitors are able to recall details of ad banners and of the clip
viewed. Using a Miramatrix eye-tracker to record YouTube viewers’ eye
movements, this study showed that nearly all fixated at least once on an ad
banner in a clip. However, less than 10% were able to correctly recall the ad
content viewed. Nevertheless, about half of viewers were able to correctly
recall clip details. Fixation duration on the banner and fixation counts on the
clip are negatively correlated, but the relationship between fixation duration
and counts on the banner was insignificant. This study sheds new light on
YouTube advertising through the use of eye-tracking and advises advertisers
to be attentive in selecting clips on which ad banners will appear.
Keywords: Fixation, Recall, Eye-tracking, Ad Banner, YouTube
According to a fair amount of empirical evidence, Internet banners may
not be so useful for marketing promotion as was thought decades ago1, 2.
Hervet et al.3 confirm that Internet users avoid looking at banners. This
effect was not incidental, but was due to viewers’ direct intent to refrain
from looking at banners. Moreover, viewers showed poor recall and
International Journal of Electronic Commerce Studies
recognition of content that appeared on banners; however, these studies
were mostly in print, and not online contexts2, 3.
In an attempt to prove apparent banner blindness, Dreze and Hussherr1
claimed that online viewers may be aware of banners on a webpage since
they shared “bandwidth with other elements of the page in which they are
being displayed” (p. 10). As such, viewers may not look at the banners. Also,
the viewers may have expected certain locations (e.g., horizontally across
the top of the screen) to contain a banner. Their focus is therefore moved to
content in other locations. Through eye-tracking, Lee and Ahn4 confirmed
that banners with animation attracted less attention than static designs.
However, the animation still unconsciously affects visitors’ attitudes.
Empirical research has confirmed the banner blindness phenomena and
indicated that online visitors do not pay equal attention to all the content on
a screen5. In other words, viewers allocate their attention to online content in
one location of the display more than in other locations. As a result, to draw
the attention of the Y generation (age 18-31) viewers, website advertisers
must include a sufficiently large image, especially of celebrities, and a
search feature, while minimizing the amount of text on the screen5. Using an
eye-tracking approach, viewers tended to scan a webpage in a top-down
manner to locate cues for the brand they were looking for6. Similar patterns
of webpage viewing were also found by Resnick and Albert7.
Despite the effects of banner blindness and the varying degree of
attention online visitors pay to website content, there has been a recent
practice by which ad banners appear at the bottom of a clip in a
video-sharing website. For example, in Figure 1, the clip, as framed by a
dotted line in Area B, is being watched while a banner, as framed by a grey
line in Area C, appears at the bottom of the clip frame. Area A as framed by
a thick dark line displays the other content, besides that in areas B and C.
The clip in Figure 1 is a movie trailer for the film Die Hard 5 and the banner
advertises a real-estate project in Bangkok, Thailand. Specifically, the
banner is asking the viewer to take a certain action as indicated in the
“register now” statement. The blurred region was intended to give contrast
to Area 1 from the other areas and was only utilized for clarity.
The reasoning behind this placement strategy was that when the
website visitors start watching the clip, they may fixate on the ad banner and
eventually click on it to learn more. As a result, we considered it interesting
to examine the correlations between fixation duration and fixation counts on
the banner, or alternatively, fixation counts on the clip. To the author’s
knowledge, such empirical research has yet to be done.
Chatpong Tangmanee
Figure 1. YouTube display of a trailer of the movie Die Hard 5 (Area B)
featuring a banner (Area C) advertising a real estate project in Bangkok,
Among video-sharing websites, YouTube is listed in many sources as
one of the most popular8, 9, 10. This remarkably high level of acceptance can
be attributed to three components11. First, YouTube allows a display of
videos with quality comparable to well-established video play technologies
including Windows Media Player, RealPlayer or QuickTime. Users can
upload videos of different formats which YouTube will subsequently
convert to a uniform easily-playable format. Second, the actionable features
of YouTube allow a straightforward sharing process. Furthermore, YouTube
clips’ HTML markup tags help users to easily embed them in other websites.
Finally, YouTube assigns to a video clip a series of meta-data. The data
includes: (1) typical identifiers such as a distinct 11-digit identification
number, the date when the clip was added, or the uploader’s identification;
or (2) a list of related videos. Unlike videos distributed by peer-to-peer
services, YouTube videos contain “relationship” attributes that provide a
link to other videos with similar descriptions chosen by uploaders.
According to Cheng et al.11, 12, these three features account for the
significant level of acceptance YouTube has earned. For these reasons,
YouTube has been chosen as the context of the current study.
International Journal of Electronic Commerce Studies
Among indicators of advertising effectiveness, recall is of principal
interest in this study. Typically, online advertisers expect that viewers
should be able to recall what they have seen on a computer screen in an
online campaign. The greater the recall, the more effective the campaign is
expected to be. A few empirical studies have confirmed that there is a small
amount of recall of banner promotions amongst viewers3, 4, 13. However,
virtually none of these studies has examined recall in the context of video
clips placed on YouTube or in conjunction with viewers’ fixation behaviors.
As a result, the four objectives of the present study are: (1) to examine
whether viewers of YouTube clips have fixated on ad banners appearing at
the bottom of clips, (2) to determine what correlation exists between
fixation duration on banners and fixation counts, in addition to (3) the
correlation between fixation duration on banners and fixation counts on
clips, and finally, (4) the extent to which visitors recalled content in the
banners, and details of the clips.
The remaining sections are outlined as follows: the literature review
addresses findings of previous research on eye-tracking and electronic
commerce and will include a discussion of the rationale behind the current
study; the third section explains the research methodology followed by a
report of the results; the final section provides conclusions, including the
contribution of the findings to the field and study limitations.
2.1 Eye-Tracking Method
Tracking what a person visually fixates on is challenging. Subjects’
responses to a questionnaire about content viewed points to issues of
validity and reliability of eye-tracking in research. These issues include:
inaccuracy in recall of what subjects claim to see and secondly, lack of
recognition of what was seen. However, advances in eye-tracking
technology are able to provide some valid insights. The moment at which
people see something may coincide with a cognitive gap for which they
require further information to cope. Basically, an eye-tracking device
records a subject’s eye movement as being exposed to visual stimuli and
aggregates the data in both qualitative (e.g., footage of what is looked at)
and quantitative ways (e.g., coordinates of areas of fixation). The process of
tracking how a person watches items on a computer screen involves two
issues: saccades and fixation. The saccade is the simultaneous movement of
the eyes in the same direction as it jumps from one fixation point to another.
Although simultaneous, the eyes must pause on a specific area of the screen
and this is known as the fixation. In order to piece together a complete
Chatpong Tangmanee
image at which one is looking, an individual will fixate on one spot and then
move the gaze. The image of the fixation takes place in one’s fovea. It
provides a highly detailed, clear image of what one is focusing on. Given
the tiny area of fixation, one must shift one’s gaze in saccades around the
area of interest in order to fully form an image.
As a result, an eye-tracker will record the location on which a person
fixates and track the movement of his or her eyes. Figure 2 shows the
outcome from an eye-tracker of a subject’s eye movements during a visit to
a Wikipedia webpage. This viewer’s first fixation was on the navigation
section labeled Number 1. The next fixation point, together with the
saccadic path, were also recorded. Researchers often pay attention to the
cluster of recorded fixations in a particular region since they provide valid
evidence of a viewer’s visual attention14.
While Figure 2 provides a qualitative summary of a visual
representation of how an eye-tracker records fixating behavior, it is not the
exact data actually recorded primarily by an eye-tracker. In fact, the tracker
adopts a screen mapping coordination approach, that is, what a person
fixates on a screen is observed and recorded as an (x, y) coordinate. The
Miramatrix eye-tracker, for example, considers a computer screen as a map
of (x, y) coordinates where the screen’s upper left corner is deemed the (0, 0)
coordinate, and the X axis represents the screen’s vertical edge while the Y
axis represents its horizontal edge. This is a unique definition that users of
the Miramatrix eye-tracker must be aware of since it is in the opposite of the
typical definition in which X and Y denote vertical and horizontal axes. A
visual representation of this mapping is given in Figure 3.
With the coordinate mapping technique, researchers are allowed to
define an area of interest (AOI). The AOI is a certain area on a computer
screen which researchers define by shape in an attempt to explore whether a
fixation occurs in the area14. To declare the AOI, researchers adopt the
coordinate format. For example, one could define an AOI in Figure 4
surrounded by the four coordinates as (140, 870), (1000, 870), (1000, 990)
and (140, 990). In a usability assessment, it is important to detect if viewers
have fixated on a given AOI. Olmsted-Hawala & Bergstrom15 were able to
adjust the display of fields on an online form based on the samples’
fixations on particular fields that were defined as AOIs in their study.
In the present study, we identified two AOIs on which to analyze
fixations. The first AOI is the banner that appears at the bottom of a video
clip and the second is the exclusive display of the clip. Note that the former
embeds in the latter. When we refer to the clip, it thus means the display
area of the clip with the banner removed. Referring to Figure 1, our first
International Journal of Electronic Commerce Studies
AOI is Area C and the second is Area B with Area C removed. Irwin16
recommended two variables associated with an AOI: fixation duration and
fixation counts. The two constructs are valid indices of cognitive processing
16, 28.
Figure 2. An example of fixation points on Wikipedia
(, accessed on
September 24, 2014)
Figure 3. Mapping coordinates on a screen display
Chatpong Tangmanee
One fixation refers to one pause recorded by an eye-tracker to confirm
that a viewer has fixated once on an area of interest (AOI). Given a
coordinate of a fixation, researchers are able to determine whether the
fixation is in the AOI. For example, if the eye-tracker records the two
coordinates as a subject is visiting the webpage in Figure 4 as (132, 750)
and (215, 930), we could have a program script to detect that the former is
not in the specified AOI, but the latter is. Bergstorm and Schall14 contend
that a fixation is the eyes’ brief rest on a certain area during saccadic
movement. It is possible that in a video viewing session, one may fixate on
and off the video clip. Consequently, fixation counts refer to the total
number of recorded fixations on a given AOI in one session and the fixation
duration refers to the total amount of time (in milliseconds) during which an
eye-tracker records the viewer’s fixation on a specific AOI.
In addition to studies of advertising on a computer screen, eye-tracking
can be applied in many contexts including landscape architecture and
printing technology. Prior to the data collection section, a researcher
typically needs to calibrate the device, after which the actual tracking
process can begin. The process is non-intrusive and can be applied to any
person including those wearing glasses or contact lenses. However, on
occasion, the subject may wear rather thick lenses or may have
oddly-shaped eye balls making calibration of the tracking device impossible.
Such subjects were therefore excluded from the current study.
Figure 4. Mapping coordinates on a screen display with one AOI identified
2.2 Banner Blindness
A large volume of findings in previous research has verified the
phenomenon of banner blindness1, 18, 19, 20. Banner blindness is commonly
International Journal of Electronic Commerce Studies
defined as the absence of fixation on banners3. Yet, Hervet et al.3 used an
eye-tracking device to prove that 82% of subjects fixated on banners at least
once. They claim that viewers were not blind to banners, but instead they
sometimes deliberately avoided fixating on them. Resnick and Albert7
confirmed based on their eye-tracking experiment that viewers who had
specific goals when visiting websites were less likely to fixate on ad banners
than those who browsed web pages for non-specific purposes. In other
words, banner blindness is less evident in a browsing visit than with a
webpage viewer pursuing a specific task.
Ad banners often compete for viewers’ attention with other content on
the same webpage. A number of studies have therefore examined banner
design features that could lead to increased attention or secure viewers’
fixation2, 21. Lee22 contended that website visitors became blind to banners
due to information overload. According to this idea, in response to
continuous information flow, online viewers simply ignore much of the
information they encounter. They pay attention only to what they consider
“highly relevant” to themselves. As such, what is obvious to some may be
invisible to others. In addition, banners located at the top of a computer
screen were relatively more noticeable than those at the right of the screen5.
Combining the habituation-tedium theory and ergonomics domains, Portnoy
and Marchionini23 claim that banner blindness is not solely the result of
viewers’ intention to ignore the banner, but also depends on the frequency
with which viewers have previously encountered banners. During search
sessions, viewers are not equally aware of their online surroundings.
Consequently, they could exhibit inattention to banners because (1) the
stimuli did not draw their attention or (2) they intended to ignore the
stimuli23, 24. Cho and Cheon19 examined why people avoided advertising on
the Internet and found that perceived goal impediment is a significant
antecedent of advertisement avoidance.
2.3 Examination of Fixation and Recall
Online advertising campaigns often rely on visual stimuli.
Consequently, advertisement designers assume that stimuli will be able to
initiate viewers’ fixation, and retain their attention. In a print ad medium,
researchers25, 26 used eye-tracking to confirm that a greater fixation rate of
magazine readers correlated with better brand recall. Seo et al.27 noted the
significant effects of brand image on visual attention in a Korean electronic
commerce website.
Use of an eye-tracking device allows researchers to accurately
calculate the number of fixations in an AOI. A single count is determined
when a visitor starts gazing on a banner (Area C in Figure 1) until their
Chatpong Tangmanee
fixation moves out of the banner area. As such, if a viewer fixates on and off
the banner on a screen, this could increase the number of fixation counts in
one session. Researchers may count fixations on an ad banner to quantify
the extent to which a web viewer has seen it. Through eye-tracking, Hervet
et al.3 verified that about 9 in 10 subjects had fixated on an ad banner at
least once. Based on these studies, it appears the actual occurrence of banner
blindness is low. The location in which a banner appears on a screen and the
degree of animation in the banner content has an interactive effect on the
fixation count on banners28.
Previous research has attempted to examine factors that contribute to
the amount of attention spent in fixation and high fixation counts. Through
eye-tracking, it has been shown that facial images attract higher fixation
counts than corporate logos in a financial service website29. Seo et al.27
discovered that famous personalities (what they call “human brand image”)
in an endorsement on an electronic commerce website were able to attract
longer durations of fixation than typical presenters. An eye-tracking analysis
of a recruitment website revealed that Internet-based job seekers had a
higher number of fixations on text than on graphic images30. These findings
imply that online viewers fixate on particular digital details more than others.
Hamborg et al.13 confirm that animated banners were able to trigger viewers’
fixation better, thus leading to better recall of advertised products and their
details than static ones. Fifty-five percent of viewers fixated at least once on
static banners, while 94% fixated on animated banners. However, 62% of
viewers who responded to survey questionnaires falsely responded that they
had not noticed any banner on the experimental website. Using an
eye-tracking technique, Simola et al.31 verified that animation of ad content
and abruptly appearing ads could draw viewers’ attention. This distraction
was evident in both browsing and reading tasks. According to the review by
Higgins et al.32, the distraction may consequently draw viewers’ attention,
yet no empirical evidence confirms this statement.
In the context of search engines, the location for display of content is
critical (i.e., whether it is on the top, right or at the bottom of the screen),
given the large number of stakeholders on a webpage displaying search
results. The fixation counts on content at the bottom of a page should be
fewer than those at the top of the page. This speculation was empirically
confirmed by Guan and Cutrell33. In fact, Goldberg et al.34 verified through
eye-tracking that users of search engines process information on the results
page from left to right in the same column. Guan and Cutrell33 further
remark that the results of search engine optimization and those of sponsored
items may receive different fixation counts. The conclusion to draw from
these studies is that a search engine web designer must be very careful in
International Journal of Electronic Commerce Studies
displaying various types of content on a search result page since viewers
fixate only on some items and may miss others.
In addition to fixation counts, fixation duration is another interesting
research construct. It is defined as the amount of time a viewer spends
fixating on a banner. Few studies have addressed fixation on ad banners18, 22,
27. Among the few who have is Lapa21 who found that fixation duration
decreased when the format of successive web pages was the same. Given
identical layout, Internet users may learn quickly to refrain from looking at
the same screen area as they are aware of what to expect from previously
viewed screens.
As seen in Figure 1, content on the screen can be classified into three
types based on the three major locations of display: the ad banner (Area C),
the video clip (Area B excluding the ad banner) and the rest of the screen
(Area A). Presenting an ad banner on a clip is a common service on a
video-sharing website. While viewers enjoy the clip, they are assumed to
have seen the banner. We believe that the fixation duration on the banner
should be long enough to attain high fixation counts if an ad banner is to be
effective in retaining the attention of viewers. Consequently, Hypothesis one
is that the relationship between fixation duration and the fixation count is
statistically significant. Although such a relationship may be anticipated,
there is, as yet, no empirical evidence to validate it.
In general, web-based advertising and ad banners compete with other
content on the same page for viewers’ attention. On video-sharing websites,
the ad banner is perceived as complementing and not competing for viewers’
attention with the clip since the placement of the former is embedded in the
latter. Hypothesis two is therefore the correlation between fixation duration
on the ad banner and the fixation counts on the YouTube clip itself is
statistically significant. Besides the clip and the ad banner, other content on
a YouTube screen may also play a role in online marketing strategy.
Intuitively, it is understood that other content should be plain so as not to
draw viewers’ attention which should focus mainly on the ad banner or the
The rate of fixation on a YouTube ad banner is not the only indicator
of online advertising success. Online marketing practitioners are often
interested in examining if visitors are able to recall the content of banners.
As such, for purposes of this study, ad content recall is considered to be the
extent to which a visitor can remember the details of a banner and recall
them without aids or prompts. Given that the banner is supposed to
compliment the clip, we wanted to see to what extent viewers can also recall
details associated with the clip. Examining whether viewers can recall clip
Chatpong Tangmanee
and banner content is consistent with Till and Baack’s method35 used to
investigate the effects of creative television commercials. According to
Hamborg et al.13, about 11% of site visitors were able to correctly recall
products on banners but none were able to recall written details. Similarly,
only 8% of viewers correctly identified whether an ad banner had been
present30. Moreover, Bayles and Chaparro7 discovered that less than 50% of
viewers could recall the presence of ad banners. Based on an eye-tracking
examination, Kuisma et al.28 confirmed the effect of interaction between
animation of ad content and ad format on viewers’ attention measured by
fixation counts. Animated ads displayed vertically garnered the most
attention. Furthermore, animation was found to significantly improve recall
of content. Kuisma et al.28 discovered one unexpected finding. Their
subjects were able to recognize ad content without actually having to look at
the ad (i.e., there was no fixation on the ad). They further discuss ways in
which online consumers are able to allocate their focal attention to irrelevant
ads when they are involved in multiple tasks on a screen.
Recall can be measured in two ways: unaided and aided. While the
former is to test if a person can properly describe details of a stimulus, the
latter is to assess if the person can confirm the stimulus once it is presented.
Aided recall is thus similar to recognition36, 37. Consequently, the present
study focuses on unaided recall. Moreover, an effective advertising project
should attract viewers’ attention and reinforce the ability to recall details of
3.1 Population and Participants
The target population of this study were people who have watched a
clip on YouTube at least once. Since Chulalongkorn Business School in
Chulalongkorn University allowed the authors to use an eye-tracking device
at a university lab, the subjects chosen for the research were undergraduates
from the business school. Thus, no claim is made that the college students
chosen as subjects in the study represent the target population. However, a
large proportion of those who are active users of video-sharing websites are
in the same age range of the chosen participants in the current study.
According to a 2010 survey by comScore, people between 15-24 years old
were the largest group of viewers (as compared to other age groups) and
spent an average of 42.1 hours per month watching online videos38. Two
years later, the comScore survey showed that 90% of Internet users were
young people in this age group each of whom reported to watch an average
of 202.5 videos on the Internet39, 40. General Internet user age profiles
International Journal of Electronic Commerce Studies
provide support that users in the 15-24 year age range represent a significant
portion of current Internet users. For example, the largest group (20%) of
Internet users in Thailand are 15-24 years of age41. A 2013 survey by Pew
Research shows that the largest (30%) group of Internet users worldwide is
between 13-25 years of age42. We therefore believe that those selected to
participate in the current study represent the largest group of Internet users
including those who enjoy watching videos online.
We announced a call for research participants in two undergraduate and
two graduate classes at Chulalongkorn University. A total of 107 students
participated voluntarily in the study. Some of the subjects were rewarded
extra points for class, but they learned of this only after completing the
participation. As such, their participation was voluntary. Table 1 presents a
profile of the participants. Just under 30% of subjects were men, 42% were
18-19 years old while 68% were undergraduates. Only 37% of the
participants had normal vision.
Table 1. Participant profiles (n = 103)
N (%)
Subject vision
3.2 Variable Operationalization
The six major variables in this study were: (1) whether YouTube
viewers fixate on an ad banner placed at the bottom of a YouTube clip, (2)
the duration of fixation on the ad banner, (3) fixation counts on the ad
banner, (4) fixation counts on the YouTube clip and (5) recall of what was
advertised on the banner (banner recall) and (6) recall of content featured on
the clip (clip recall).
A Miramatrix eye-tracking device was used to verify whether visitors
gazed on an ad banner and measured the duration (in milliseconds) that the
viewer fixated on the banner. In addition, eye-tracking was used to count the
actual number of viewer fixations on a banner and clip. The Miramatrix s2
system was connected to a desktop computer with a resolution of 1024 x
768 pixels. The eye-tracker’s resolution and sampling rate were at most
Chatpong Tangmanee
and 60 Hz, respectively. Eye movements were captured by a camera placed
beneath a 17" computer screen located about 50 cm from the participant.
The use of the eye-tracker helps us to reliably and validly detect a
viewer’s fixation on a screen. Had the viewer been asked using a
questionnaire, their response may have been subjective. It is uncommon that
a person who has enjoyed a YouTube clip would be able to reliably recall
whether they had looked at a specific banner14.
Given that our participants were college students, we strove to select
movie trailers that would be of interest to that segment of viewers. The
visual stimuli selected were thus three clips on YouTube viz. movie trailers
of Die Hard 5, The Smurfs 2 and one supplied from the channel Cieon, a
YouTube clip supplier. Die Hard 5 and The Smurfs 2 were selected to
represent the two genres of action and animation movies, respectively. The
third clip was selected from the Cieon collection. We selected the clip from
Cieon to ensure a variety of movies for the rest of the participants. There are
three reasons why we selected these clips. First, we want our findings to be
generalizable to the population of users in this age group. This is why using
only one clip was deemed inadequate. Second, we chose to use clips of
comparable length on which one ad banner would appear. Each of the three
clips lasted between 1.5 - 2 minutes, a sufficient period for participants to
enjoy the content of the clip and to track their gaze. In other words, this
length of time was considered sufficient to observe whether subjects fixated
on the clip or on the ad banner, but was not too long for research purposes.
Third, all clips selected had banners of comparable size located in the same
location. Figure 5 shows two sets of four coordinates that define two AOIs
in the current study. The first set (cxi, cyi) where i = 1, 2, 3, 4 is for the
location of the clip and the second (bxi, byi) where i = 1, 2, 3, 4 is for the ad
banner. For instance, the Die Hard 5 clip fits within coordinates (225, 136),
(225, 560), (866, 136), and (866, 560). Although the other clips chosen may
not have exactly the same coordinates as those in Die Hard 5, they were of
comparable size and location.
Though all clips were in English, we do not believe this posed a
problem for the subjects (Thais) to follow them since they were brief.
During the three-day observation before the actual data collection, all three
selected clips met our requirement criteria; all three were therefore kept for
use in the current study.
The decisions regarding the movie selections were mainly to enhance
the finding’s reliability. For instance, the selected movie clips must be of
varying interest to the sample. Consequently, we opted for clips available
through the Cieon collection, in addition to the action (i.e., Die Hard 5) and
International Journal of Electronic Commerce Studies
animation (The Smurfs 2) movie clips. Also, our selection of clips of
comparable length (i.e., 1.5 - 2.0 minutes) had implications on the study
reliability. Had we selected clips of far different lengths, the audience may
not have recalled viewing the banners and the movie clips.
Banner and clip recall measurements were unaided. After viewing the
YouTube clip, subjects were asked if they remembered which product was
in the banner. In fact, they were asked to write down the advertised product
name. Unaided recall was used because we wanted to test viewers’ memory
of banner content which was measured as banner recall. This was the first
step before testing for other aspects of memory35.
Figure 5. Coordinate references on the current study’s clip (cxi, cyi), where
i = 1, 2, 3, 4 and those of the ad banner (bxi, byi), where i = 1, 2, 3, 4
Once subjects responded to the banner recall question, they were asked
to answer two additional questions that measured recall of the clip. In the
two questions, we asked the subjects to write (1) the title of the movie clip
and (2) the day the movie was to open in theatres. The subjects earned 0
points if both answers were incorrect, 1 point if partially correct and 2
points if both were correct. The two questions were intended to measure
accuracy of recall of the clips.
Note that the data in the current study were collected in a captive
environment. We required that a subject watch the YouTube clip while
eye-tracked, after which the subject responded to a few questions. A typical
research design with a pretest or control group was not employed. However,
researchers adopting an eye-tracking approach often refer to such data
collection as an experiment14.
Chatpong Tangmanee
3.3 Procedures and Analysis
Once participants arrived at the lab, they received a short note
explaining the purpose of the research and importance of their participation.
The participants were asked to take a look at a movie trailer on YouTube,
after which they were asked to respond to a brief questionnaire. We did not
mention about the ad banner. Although they were not directly asked for a
consent, the participants learned from the note that they were free at any
time to leave the lab, if uncomfortable with the circumstances.
When they were ready, the eye-tracking device was calibrated to each
participant, which took about 1 - 1.5 minutes after which the participant
started viewing the YouTube clip. The aim behind the calibration for each
participant was to ensure instrument reliability. Participants were randomly
selected to watch one of the three clips. While viewing, the subject’s eye
movements were recorded for further analyses. After the clip ended, they
were asked to answer the recall questions, and then filled in their
demographic information and received a ballpoint pen for their contribution
to the study. Of 107 students that signed up to participate in the study, four
did not show up, leaving 103 usable records. We then processed the data in
order to derive the values of the six major variables. A research assistant
examined the data recorded in the eye-tracker in order to (1) check whether
the participant fixated on a banner and (2) to count the number and measure
the duration of fixations on the banner and count the number of fixations on
the clip. The assistant subsequently combined the fixation data and the
recall data to form the data set for further analysis and hypothesis testing.
Descriptive statistics of all variables were recorded. The analyses of
fixation and recall were mainly done through percentages and cross
tabulation. Moreover, the Pearson’s r correlation was used to test the two
hypotheses only if the fixation duration and fixation counts were normally
distributed. If not, non-parametric correlation replaced the Pearson’s r.
As seen in Table 2, 97% of subjects fixated at least once on the
YouTube ad banner. The analysis on the minors who had not fixated on the
ad banner indicated that their focus was neither on the ad banner nor the clip
but on the content in other areas. This further implies that in this study, there
was no banner blindness3. The subsequent analyses are therefore based on
the 100 samples who we found through eye-tracking to have gazed on the ad
banner at least once. As seen in Table 3, among those who had fixated on
the ad banner, only 8% were able to correctly recall what was advertised on
International Journal of Electronic Commerce Studies
the banner. However, when asked about the clip, 6 out of 10 could give the
correct movie title and 5 in 10 were able to recall the correct air time. In
short, about 50-60% of the subjects were able to correctly recall the clip
details while 8% could remember the clip titles properly.
Table 2. Fixations on YouTube Ad Banner (n = 103)
N (%)
No Fixation
At least one fixation
As shown in Table 4, the participants fixated on the ad banner for an
average of 573.01 milliseconds. The average number of fixations on the ad
banner and on the YouTube clip were 17.88, and 73.96, respectively. Based
on skewness and kurtosis, we conclude that the variables were not normally
distributed since their absolute values exceeded one43. As a result, the
correlation analysis had to be non-parametric. We thus substituted Pearson’s
r with Spearman’s rank correlation.
Table 3. Content recall of clips and banners viewed (n = 100)
N (%)
Banner Ad
Correct recall
Incorrect recall
Movie clip titles
Correct recall
Partially correct recall
Incorrect recall
Movie clip details
Correct recall
Partially correct recall
Incorrect recall
In the first hypothesis we attempted to test if the correlation between
duration of fixation on an ad banner and fixation counts on the banner was
significant. Yet, the correlation found (-.100, p < .324), shown in Table 5,
was insignificant. For the second hypothesis, the correlation between
duration of fixation on the ad banner and fixation counts on the clip was
negatively significant (-.486, p < .000). The following section provides
further discussion of the findings.
Chatpong Tangmanee
Table 4. Descriptive statistics for variables (n = 100)
Standard deviation
Duration of fixation on ad banner
Fixations on ad banner (times)
Fixations on YouTube clip (times)
Table 5. Correlation matrix using Spearman’s rank correlation analysis
Duration of fixation on ad banner
Fixations on the ad banner
Fixations on the YouTube clip
Note: *in parentheses are p-value above which are correlation coefficients.
This research used an eye-tracking approach to explore YouTube
viewers’ fixation on and recall of an ad banner placed on YouTube movie
clips. A total of 103 students from the Chulalongkorn Business School were
selected randomly to view one of three movie clips. While viewing, their
eye movements were tracked, recorded and converted into a count of
fixations on ad banners and video clips. Despite the relatively small size of
the sample, as college students are in the age range of a large portion of both
Internet users and YouTube viewers32, 42, findings based on this group of
subjects are deemed to have validity.
Through eye-tracking, it was shown that 97% of the subjects fixated at
least once on the ad banner placed at the bottom of YouTube clips. This is
confirmation of the scant chance of banner blindness if that term means
complete lack of fixation on a banner. Compared to the number of fixations
on ad banners reported in previous research13, the rate in the present study
(97%) was found to be similar to the rate of fixations on animated banners
(94%), but considerably higher than that on static banners (55%) reported in
the same experiment. Such a high figure may suggest that placing an ad
banner on a YouTube clip could be a better strategy than presenting it via
other channels. In addition, a sizable number of studies have reported
banner blindness under different conditions which included undertaking
various tasks during a visit to a website, or alternatively, a number of
banners appearing in different locations on the webpage in question35, 44.
International Journal of Electronic Commerce Studies
Our attempt may be the first to examine banner placement on YouTube
clips and we were able to show that nearly all viewers fixated at least once
on the banner in this location. It is thus reasonable to claim that embedding
a banner on a YouTube clip is a relatively feasible strategy for some types
of advertising. The no-blindness incidence may be accounted for by the type
of task undertaken during a visit to YouTube. Previous work has suggested
that viewers whose purpose of a visit is to browse items online are more
likely to notice and fixate on a banner in the website than those with another,
more specific goal7, 28. People presumably visit YouTube to relax and enjoy
watching video or movie clips. YouTube viewers are therefore likely to
notice a banner embedded in such a clip.
We report that our research subjects on average spent .573 seconds
gazing on an ad banner on a YouTube clip. In other words, the viewer
allocated about half a second to fixate on the banner. In terms of the number
of fixations on the banner and on the YouTube clip, it was found that
subjects fixated on average 17.88 and 73.96 times, respectively. Again,
subjects in this study, at least, were not blind to the ad banners. Compared to
the findings in previous projects22, 45, the number of fixations and the
duration of fixation on banners in Koster et al.45 were 4.02 times and 0.851
seconds while those in Lapa21 were 3.01 times and 2.69 seconds. Our
finding concerning the average fixation count (i.e., 17.88 times) was
considerably higher, but that of the average fixation duration (i.e., 0.573
seconds) was considerably lower.
In Koster et al.45, personalized banners obtained more fixation counts
than non-personalized ones. Similarly, in Lapa21, banners with content
related and relevant to that on the webpage on which the banners were
displayed received more fixation counts than those with unrelated irrelevant
content. The more frequent fixations and the shorter length of each fixation
in our study may indicate a poor match in content between the banner and
the movie clip. As such, the viewers might not have wanted to spend time
looking at them. Also, the number of fixations on the YouTube clips was
roughly four times higher than on the ad banner. It is reasonable to assume
in the current study that the clips were therefore considered more engaging
to viewers than the ad banner. Had the banner been personalized or had
content been relevant to the webpage matter21, 45, fixation on the banner
would have been more frequent or of a longer duration. The significant
negative correlation between the number of fixations on the clip and the
fixation duration on the ad banner also supports the assertion that the clips
were of greater interest to subjects than the ad banner. Nonetheless, as we
are aware of no previous empirical work investigating issues about fixation
Chatpong Tangmanee
on YouTube clips, and as such, there is no benchmark against which to
compare these figures. This suggests an area for future research.
Among those who fixated on the ad banner in this study, only 8% were
able to recall what was presented on the banner. Although 8% is a relatively
low percentage, it is consistent with results in previous studies13, 46 in which
about 8-11% of viewers were able to correctly recall the presence of banners
or their content. According to Koster’s work45, subjects were able to recall
an average of 21.6% of all ad content. Such a high level of recall could have
been due to their use of personalized banners. Once personalized, subjects
could have recalled the ad better than in our study (8%). In addition to
banner recall, the subjects in the present study were able to recall details of
the movie clip (e.g., title and air time) more accurately than banner content
(see Table 3).
Given the study’s exploratory nature, two possible explanations are
posited for this difference in recall. First, most viewers may have a tendency
to pay more attention to content in a relatively larger area of the screen than
in a small area. Since the banner at the bottom of a clip is often much
smaller than the clip, viewers may see and perceive the banner as an
annoying peripheral object and this may hinder recall. A second explanation
pertains to the matter of competing content, i.e., the ad message was recalled
with less accuracy than the clip, perhaps because the clip content was much
more engaging to watch. The second explanation is supported by the fact
that the number of fixations on the clip was four times higher than on the ad
banner and the significant negative correlation between fixation duration on
an ad banner and number of fixations on the YouTube clip discovered in the
present study. The longer that YouTube viewers fixated on the banner, the
fewer fixations on the clip. If the clip content draws much of the viewers’
fixation, they will necessarily spend less time focusing on the banner. In
other words, had a less interesting clip been selected, it is possible that
fixations on the ad banner would have been greater or could have led to
greater recall of banner content. If the clip content is too uninteresting,
however, no one would want to watch it, leaving the embedded ad banner
unviewed. Consequently, the selection of an appropriate video clip is a
challenge for online advertising designers. As these explanations are
speculative, they require further research to validate.
From an online advertising perspective, the trivial correlation between
fixation duration and fixations on ad banners is a disappointing finding in
this study. Designers of online advertising might have expected a more
significant positive correlation, i.e. in a successful YouTube advertising
campaign, the higher the number of fixations on an ad banner, the longer the
duration of each fixation. This could be assumed to be an indication of
International Journal of Electronic Commerce Studies
viewers’ greater interest in the message on the ad banner. Although
insignificant, the slim negative correlation coefficient in the current study
might be a result of the message on the ad banner competing with the clip
content for viewers’ attention. While viewers were watching the clip, they
could encounter some less engaging scenes during which their peripheral
vision would pick up the ad banner embedded at the bottom of the clip. This
should then lead viewers to fixate on the banner. Once a more engaging
scene is encountered, interest on the banner might be interrupted. This is
thus a matter of competing content between the YouTube clip and the
embedded ad banner. The slightly negative relationship in the current study
could indicate that the clips chosen were much more interesting than the ad
banner. Viewers may have had a relatively high number of fixations (17.88)
on the banner, but did not fixate on it long. In the review of the literature
only the work of Koster et al.45 examined similar issues. The trivial
correlation between the number of fixations on ad banners and the fixation
duration found in the present study is consistent with what was found in
Koster et al.45 However, as only two empirical research investigations came
to this conclusion suggests further research is warranted.
The findings of this study have both theoretical and practical
implications. Theoretically, the results have shed new light on the use of
eye-tracking techniques for analysis of the effects on viewers of online
advertising in the context of YouTube. Two major points of conceptual
import stem from our findings. First, YouTube viewers are not blind to ad
banners placed at the bottom of a movie clip, which is contrary to previous
reports. However, they may not correctly recall banner content. Viewers
nonetheless retain significant recall of movie content on YouTube clips. The
second theoretical contribution is the finding of a significantly negative
correlation between fixation duration on the ad banner and fixation counts
on the YouTube clip. This confirms the competition for viewers’ attention
between the YouTube clip and the ad banner. The idea that YouTube
viewers may enjoy a clip, notice the ad banner embedded at the bottom of
the clip and ultimately click on the banner to further learn about the ad is not
supported in the context of the present study. Such finding should prompt
attention from other researchers and spur further studies to shed light on the
These findings offer a practical contribution to two groups of
stakeholders: online advertisement designers and clip-sharing website
managers. As the people in charge of placing ad banners on a video clip,
online ad designers have to conceive of a design that can secure the clip
viewers’ attention as well as accurate recall of the ad message. The present
study’s findings confirm the low probability of banner blindness if the ad
Chatpong Tangmanee
banner is embedded at the bottom of a YouTube clip. Nearly all viewers in
the study noticed the ad banner as demonstrated by having one or more
fixations on it. Noticing the banner may have been a result of what was
referred to by Hsieh et al.47 as “dishabituation.” Typically, a web site visitor
may assume the location of a banner and successfully avoid looking at it.
However, if the presence of the ad banner does not conform to the viewer’s
expectation, they will be more likely to notice the banner.47 In other words,
the dishabituation of a banner location could enhance a viewer’s attention to
the ad banner. According to Hsieh et al.47, an image or animation may better
trigger the dishabituation effect and consequently yield more attention from
the viewer as compared to plain text content. An alternative way to
dishabituate website visitors expectations is to delay the display of an ad
banner31. According to Resnick and Albert7 and Simola31, viewers expect
that the display of a webpage’s top section will be faster than that of the
bottom. While waiting for a webpage to download to the screen for example,
if all sections excluding the banner appear on the screen, a viewer is likely
to notice the location where the banner is being downloaded and
subsequently wait to see the content of the banner.
In addition to viewers’ awareness of advertisement message, designers
of online advertising will desire successful recall of the ad content. The
present study verified that YouTube visitors could not correctly recall what
was on the banner, but their recall of details from the video clip was fairly
accurate. Thus, designers of online advertising must adjust their campaign
strategy to enhance recall of ad content. Findings in previous studies offer
suggestions in this regard. First, animation in a banner enhanced viewers’
recall of an online advertising campaign13, 28, 31. It was found that completely
animated banners drew more of viewers’ attention and resulted in greater
recall than static banners.13 Moreover, animated ad content received a high
number of fixations and relatively long fixation duration31. Interestingly, the
number of clicks on animated banners did not differ from that on banners
with a static design. As a result, designers of online advertising would be
well advised to adopt animation in designs of ad banners since it could
trigger recall, although it may not lead to a significant number of clicks. The
subsequent challenge is to tailor online advertising by using the right
amount of animation so as to achieve expected outcomes.
The other implication from the study is that ad personalization may
trigger recall. According to Koster et al.45, personalized banners were able to
trigger more recall of banner content than non-personalized ones. The
banners in their study were embedded in a news website and personalized
according to viewers’ demographic characteristics including age, gender,
interests, profession, and location45. Based on findings in Koster et al.45,
International Journal of Electronic Commerce Studies
online advertisement design could benefit by the inclusion of personalized
ad banners making use of YouTube viewers’ personal backgrounds prior to
embedding in a video clip. Given competition for viewers’ attention from
content between the ad banner and the clip found in the current study,
personalization of banner content may improve viewer attention on the
banner over that reported in Koster et, al.45 As a result, designers of online
advertising will have to balance personalization of banner design with the
selected clip on which the banner will appear.
Website managers of clip-sharing websites such as YouTube are the
second group of stakeholders who could benefit from the findings in the
present study48. The aspect of key importance for managers is the selection
of a clip that can successfully balance viewers’ attention between the ad
banner and clip content. There are two possible practical suggestions to
achieve this. First, managers should be very well aware of the presentation
of clip content to be able to pinpoint precisely the moment in a clip during
which an ad banner should appear in order to balance competing content
between ad and clip. For example, during a typical movie clip, there may be
relatively less engaging moments during which the ad message could pop up
to draw viewers’ attention and retained until the scene draws the viewer’s
interest again. For example, in the 1:36-minute clip Frozen on YouTube
(accessed on November 26, 2014) the segment from 1:20 to 1:33 may be an
opportune place for an ad banner since it is right after the climax of the clip,
but still at a point where viewers are still watching to see the end of the clip.
Second, clip-sharing managers may work beneficially with online
advertising planners on clip selection. Our findings confirm that a movie
clip may engage viewers to the extent that they were unable to recall the
content of the ad banner embedded on the clip, but still were able to recall
clip details correctly despite most viewers fixating on the banner. According
to findings from previous work, the point when an online viewer is most
likely to notice an ad banner is while still browsing, and not while engaged
in a particular site7, 28. Should the chosen clip be a movie, it should have a
moment where viewers could allocate attention to absorb an ad banner or
the clip should be something other than a movie that will permit the viewer
to notice the ad. For instance, the clip could be a documentary or routine
presentation featuring a community leader attending a local event. Viewers
of such routine types of content may focus on the presentation only
moderately and still be able to allocate attention to absorb content from an
ad banner embedded at the bottom of the display. Nonetheless, this is just a
suggestion for which more empirical research is needed to confirm
Chatpong Tangmanee
The study had three limitations. First, the context of Internet use is
constantly changing. What can be found on YouTube today may not be
valid on a future day. This imposes a limitation on the generalizability of the
findings. More frequent research to keep up with changes is one remedy.
Second, the data was collected in a captive environment. The subjects
selected were those with access to a lab, and in the present case, college
students. It is suggested that the use of eye-tracking methods in further
advertising research extend to cover other types of online visitors or other
websites. Given the exploratory nature of the current study, we excluded
certain variables that may have confounded the recall of a banner on a
video-sharing website. This could be another limitation of the current study
findings. As such, future research could examine possible confounding
effects such as a subject’s disposition to a movie clip. The outcome of such
research would further build on knowledge about online advertising on a
We are thankful for partial financial support from the Chulalongkorn
Academic Advancement into its Second Century project. We also want to
express appreciation for constructive comments from IJECS reviewers.
[1] X. Dreze, and F. Hussherr, Internet advertising: Is anybody watching?
Journal of Interactive Marketing, 17(4), p8-23, 2003.
[2] R.S. Moore, A. Stammerjohan, and R.A. Coulter, Banner
advertiser-web site context congruity and color effects on attention and
attitudes. Journal of Advertising, 34(2), p71-84, 2005.
[3] G. Hervet, K. Guerard, S. Tremblay, and S. Chatourou, Is banner
blindness genuine? Eye-tracking internet text advertising. Applied
Cognitive Psychology, 25(5), p708-716, 2011.
[4] J. Lee, and J.-H. Ahn, Attention to banner ads and their effectiveness:
An eye-tracking approach. International Journal of Electronic
Commerce, 17(1), p119-137, 2012.
[5] S. Djamasbi, M. Siegel, and T. Tullis, Generation Y, web design, and
eye-tracking. International Journal of Human-Computer Studies,
68(5), p307-323. 2010.
[6] E. Tzanidou, S. Minocha, and M. Petre, Applying eye tracking for
International Journal of Electronic Commerce Studies
usability evaluations of e-commerce sites. Paper presented at the
workshop on 'Commercial Uses of Eye tracking' held at the 19th
British HCI Group Annual Conference, Edingburgh, UK, September
5-9, 2005.
[7] M. Resnick, and W. Albert, The impact of advertising location and user
task on the emergence of banner ad blindness: An eye-tracking study.
International Journal of Human-Computer Interaction, 30(3),
p206-218, 2014.
[8], YouTube Stats (Subscriber, Ranking, & Video Statistics).
Retrieved on October 25, 2014, from
[9], Top sites. Retrieved on October 25, 2014, from
[10] C. Lorenzo-Romero, M. Alarcon-del-Amo, and E. Constantinides,
Determinants of use of social media tools in retailing sector. Journal of
Theoretical and Applied Electronic Commerce Research, 9(1), p44-55,
[11] X. Cheng, C. Dale, and J. Liu, Statistics and social network of
YouTube videos. Paper presented at the 16th International Workshop
on Quality of Service, Enschede, June 2-4, 2008.
[12] X. Cheng, C. Dale, and J. Liu, Characteristics and potentials of
YouTube: A measurement study, in peer-to-peer video. In E.M. Noam
and L.M. Pupillo (Eds.), The Economics, policy, and culture of today’s
new mass medium (p205-227). New York: Springer Science, 2008.
[13] K.-C. Hamborg, M. Bruns, F. Ollermann, and K. Kasper, The effect of
banner animation on fixation behavior and recall performance in search
tasks. Computers in Human Behavior, 28(2), p576-582, 2011.
[14] J.R. Bergstrom, and A.J. Schall, Eye tracking in user experience
design. Boston: Elsevier, 2014.
[15] E. Olmsted-Hawala, and J.R. Bergstrom, Think-aloud protocols: Does
age make a difference? Paper presented at the SIGCHI Conference on
Human Factors in Computing Systems, Atlanta, GA, USA, April 10-15,
[16] D. Irwin, Fixation location and fixation duration as indices of cognitive
processing. In J.M Henderson and John M. (Eds.), The interface of
language, vision, and action: Eye movement and visual world
(p105-133). New York: Psychology press, 2004.
[17] M. Bayles, and B. Chaparro, Recall and recognition of statistic vs.
animated banner advertisements. Paper presented at the Human
Factors and Ergonomics Society Annual Meeting, Minneapolis,
Chatpong Tangmanee
Minnesota, USA, October 8-12, 2001.
[18] M. Burke, A. Hornof, E. Nilsen, and N. Gorman, High-cost banner
blindness: Ads increase perceived workload, hinder visual search and
are forgotten. ACM Transactions on Computer-Human Interaction,
12(4), p423-445. 2005.
[19] C.-H. Cho, and H.J. Cheon, Why do people avoid advertising on the
Internet? Journal of Advertising, 33(4), p89-97, 2004.
[20] D.R. Fortin, and R.R. Dholakia, Interactivity and vividness effects on
social presence and involvement with a web-based advertisement.
Journal of Business Research, 58(3), p387-396, 2005.
[21] C. Lapa, Using eye-tracking to understand banner blindness and
improve website design. Unpublished master thesis in information
technology. New York: Rochester Institute of Technology, 2007.
[22] C. Lee, Banner blindness: An effect of information overload on the
World Wide Web. Retrieved on October 14, 2015 from
[23] F. Portnoy, and G. Marchionini, Modeling the effect of habituation on
banner blindness as a function of repetition and search type: Gap
analysis for future work. Paper presented at the CHI'10 Extended
Abstracts on Human Factors in Computing Systems, Atlanta, GA,
USA, April 10-15.
[24] C.Y. Yoo, K. Kim, and P.A. Stout, Assessing the effects of animation in
online banner advertising: hierarchy of effects model. Journal of
Interactive Marketing, 4(2), p49-60, 2004.
[25] W. Wedel, and R. Pieters, Eye fixation on advertisements and memory
for brands: A model and findings. Marketing Science, 19(4), p297-312,
[26] R. Pieters, A review of eye-tracking research in marketing. In N.K.
Malhotra (Ed.), Review of Marketing Research, Volume 4 (p123-147).
United Kingdom: Emerald group publishing Limited, 2008.
[27] Y. W. Seo, S.W. Chae, and K.C. Lee, The impact of human brand
image appeal on visual attention and purchase intentions at an
e-commerce website. Intelligent Information and Database Systems,
p1-9., 2012.
[28] J. Kuisma, J. Simola, L. Uusitalo, and A. Oorni, The effects of
animation and format on the perception and memory of online
advertising. Journal of Interactive Marketing, 24(4), p269-282, 2010.
International Journal of Electronic Commerce Studies
[29] S. Djamasbi, M. Siegel, T. Tullis, and R. Dai, Efficiency, trust, and
visual appeal: Usability testing through eye-tracking. Paper presented
at the 43rd Hawaii International Conference on System Science
Proceedings, Honolulu, HI, January 5-8, 2010.
[30] D. Allen, J.E. Biggane, M. Pitts, R. Otondo, and J.V. Scotter, Reactions
to recruitment web sites: Visual and verbal attention, attraction, and
intentions to pursue employment. Journal of Business Psychology,
28(3), p263-285, 2013.
[31] J. Simola, J. Kuisma, A. Oorni, L. Uusitalo, and J. Hyona, The impact
of salient advertisements on reading and attention. Journal of
Experimental Psychology: Applied, 17(2), p174-190, 2011.
[32] E. Higgins, M. Leinenger, and K. Rayner, Eye movements when
viewing advertisements. Frontier in Psychology, 5, p1-15, 2010.
[33] Z. Guan, and E. Cutrell, An eye-tracking study of the effect of target
rank on web search. Paper presented at the SIGCHI Conference on
Human Factors in Computing Systems, San Jose, CA, USA, April 30-
May 3, 2007.
[34] J.H. Goldberg, M. Stimson, M. Kewenstein, N. Scott, and A.M.
Wichansky, Eye-tracking in web search tasks: Design implications.
Paper presented at the 2002 Symposium on Eye Tracking Research &
Applications, New Orleans, LA, USA, March 25-27, 2002.
[35] B.D. Till, and D.W. Baack, Recall and persuasion: Does creative
advertising matter? Journal of Advertising, 34(3), p47-57, 2005.
[36] R.P. Bagozzi, and A.J. Silk, Recall, Recognition, and the measurements
of memory for print advertisements. Marketing Science, 2(2), p95-134,
[37] S.N. Singh, M.L. Rothschild, and G.A. Churchill, Recognition versus
recall as measures of television commercial forgetting. Journal of
Marketing Research, 25(1), p72-80, 1988.
[38], 19.2 million Spanish internet users watched 2.9 billion
online videos in October 2010. Retrieved on December 11, 2010, from
[39], UK engagement with online video descends with age.
Retrieved on June 8, 2012, from
Chatpong Tangmanee
[40], comScore releases August 2014 U.S. online video
rankings. Retrieved on September 17, 2014, from
[41] NECTEC (National Electronics and Computer Technology Center).
Profiles of Internet users in Thailand. Bangkok: NECTEC press, 2012.
[42] PewResearch, Teens fact sheet. Retrieved on September 17, 2012, from
[43] S. Muylle, R. Moenaert, and M. Despontin, The conceptualization and
empirical validation of web site user satisfaction. Information &
Management, 41(5), p543-560, 2004.
[44] M. Pagendarm, and H. Schaumberg, Why are users banner-blind? The
impact of navigational style on the perception of web banners. Journal
of Digital Information, 2(1), p1-6, 2001.
[45] M. Koster, M. Ruth, K.-C. Hamborg, and K. Kaspar, Effects of
personalized banner ads on visual attention and recognition memory.
Applied Cognitive Psychology, 29(2), 181-192, 2014.
[46] N. Razzouk, and V.A. Seitz, Banner advertising and consumer recall.
Journal of Promotion Management, 9(1-2), p71-80, 2003.
[47] Y.-C. Hsieh, K.-H. Chen, and M.-Y. Ma, Retain viewer’s attention on
banner ad by manipulating information type of the content. Computers
in Human Behavior, 28(5), p1692-1699, 2012.
[48] C.E. Tucker. The reach and persuasiveness of viral video ads.
Marketing Science, 34(2), p281-296, 2015.
International Journal of Electronic Commerce Studies
... Second, visual attention is an important issue during overlay ads. Although almost everyone visually notices the ad, only a small amount are able to remember its content (Tangmanee, 2016). We present in Table 4 a summary of the content present in those two studies. ...
... advertising in OTV, in-stream ads, overlay ads, YouTube ad formats and metrics, computational systems development, and others). In addition, we provide a summary of those future research directions in Table 7. Furthermore, we encourage future studies to be supported by neuromarketing techniques and real data since important studies only happened due to this (please see Tangmanee, 2016;Bellman et al., 2012;Bellman et al., 2018;Belanche et al., 2017a;andVermeulen et al., 2019, for neuromarketing supported studies, andPashkevich et al., 2012;Campbell et al., 2017;and Krishnan & Sitaraman, 2013, for real data supported studies). In addition, since older people and non-student samples are rarely used, we suggest their use in future studies as there may be differences (Belanche et al., 2017b;Rubenking, 2019). ...
... mandatory vs skippable in-stream ads). Experiments tracking eye-movement and gazes during advertising exposure facilitated by eye-tracking devices, as used by Tangmanee (2016), can provide powerful insights since they enable a higher level of confidence independent of self-reported measures. ...
Streaming video (SV), such as YouTube, is a new media widely used nowadays. Nevertheless, there is a lack of knowledge about advertising in SV. In view of this, through a search in the rich depository of the Scopus database, this article presents the first integrative literature review about advertising in SV. Searching every article and conference paper related to the topic published until May 04, 2020, 59 studies were found and classified into two main topics: marketing studies (35), mostly focused on evaluating or exploring advertising in SV, and computational studies (24), focusing on the development of systems for the insertion of ads into SV. All knowledge present in these studies was summarized so that readers (both scholars and practitioners) could have easy access to the main contributions and information present in each study. Moreover, future research directions in six main themes are presented through a research agenda.
... Task completion time was recorded from the time that the experiment material was presented to the time that participants pressed the space key, which is a measure of the total time required to find a target icon. Eye fixation is defined as the eyes pause on a specific area (Tangmanee, 2016). The fixation count shows the total number of fixations on an experimental material. ...
... According to feature integration theory (Treisman and Gelade, 1980), similarity theory (Duncan and Humphreys, 1989), and guided search theory (Wolfe, 1994), the presence of different colors on app icons on the interface of smartphones helps the Moreover, our results indicated that the varied color significantly reduced the fixation count and fixation duration when finding a target app icon on the interface of smartphones. Eye fixation is defined as the eyes pause on a specific area (Tangmanee, 2016). Previous studies have shown an inconsistent relationship between eye movement and task completion time in visual search tasks (Kim and Nembhard, 2019). ...
Icon color and icon border shape are two key factors that affect search efficiency and user experience but have previously been studied separately. This study aimed to ascertain their separate and combined effects on smartphone interfaces. We conducted an experiment using eye tracking in addition to performance and experience measures to understand the effects of app icon color and border shape on visual efficiency and user experience. The results identified both features as essential attributes with interactive effects in the process of searching app icons on a smartphone interface. The study confirmed that varied colors across icons and a rounded square border shape helped to improve search efficiency, decrease cognitive effort, and lead to a more positive user experience. Relevance to industry Users of smartphones are often confronted with the problem of selecting a single app from a great number of apps. Visual design of app icons plays a key role in influencing visual search efficiency and user experience. The results of this study have implications for designing app icons on the interface of smartphones to improve search efficiency and elicit positive user experience.
... Apesar de apenas três estudos terem utilizado técnicas de neuromarketing (Bellman et al., 2012;Bellman et al., 2018;Tangmanee, 2016), é provável que os próximos estudos as utilizem com maior frequência, já que elas trazem maior confiança aos resultados. Ainda, o uso de eye tracking pode ajudar a superar um dos maiores problemas dos anúncios on-line: a falta de atenção visual (Liu-Thompkins, 2019). ...
... In the context of video interaction, eye gaze attention can provide implicit understanding of user interest while users are watching the videos. The implicit feedback has been investigated in different analytical scenarios such as to judge video quality [Sawahata et al. 2008], to summarize and compress videos [Gitman et al. 2014;Katti et al. 2011], characterize users [Tanisaro et al. 2015], or to assess relevance of advertisement banners [Tangmanee 2016]. Therefore to support such analysis, in GIUPlayer we record and visualize the gaze data while video is being played. ...
... This study used an eye tracking device to record and analyze participants' eye movement data. Many scholars have already used this device to examine Internet scanning behavior (Harrar, Le Trung, Malienko, & Khan, 2018;Lee & Ahn, 2012;Tangmanee, 2016), evidence that it is an effective method. All participants joined the experiment and signed a form indicating that they permitted the eye movement data collected. ...
This study used spillover effects and reversal theory to examine the influence of banner advertising on the attitude of the host website. A total of 146 volunteers were recruited for an experiment. A computer program and an eye-tracking device were used to record and analyze participants’ eye movements, and a psychological scale was used to measure the participants’ attitudes. The results showed that the negative attitudes generated by banner advertising do spill over to the host websites, which in turn negatively affects the website. When viewers have negative attitudes toward the host websites, their fixation time and revisit intention of the websites will be significantly reduced. These findings encourage advertisers to plan advertising strategies and provide valuable implications for banner advertising and websites.
... Unsurprisingly, it has also grown as a sortafter medium for advertisers looking for eyeballs. However, digital media, due to its inherent nature poses exciting challenges to both practitioners and researchers (Tangmanee, 2016). Such challenges have led to the discovery of new knowledge around advertising metrics and their effectiveness on the internet. ...
Full-text available
V ideo advertising on the internet is gaining popularity amongst new as well as established brands. Independently, these ads can either have high or low ad appeal. This research studies the interactive effect of product involvement with brand familiarity and ad appeals on outcome parameters like brand engagement, ad stickiness, and intention to purchase. Firstly, a pretest is conducted on 30 ads with 54 participants to segment the ads aired on YouTube into high/low brand familiarity and ad appeal. Four ads are then selected that are either high/low on brand familiarity and high/low on ad appeal. Further, an eye-tracking study using inputs from 103 subjects captures quantitative elements like ad completion rate (ACR), ad abandonment rate (AAR) and click through rate (CTR). It also captures behavioral aspects of eye gaze and fixations. This is then followed by a survey to corroborate the findings from the eye tracking data. Findings reveal that brands that are yet unfamiliar to a market can create engagement by focusing on ad elements. However, brands that are familiar to a market cannot automatically assume engagement. The strategy and treatment for each of the segments identified in the study have to necessarily be different to reach their desired outcome, which is explained by the interactive effect of product involvement with brand familiarity and ad appeal.
... In summary, attention influences brand recall (i.e. Breuer and Rumpf, 2012;Tangmanee, 2016;Yang et al., 2015). However, just because attention has been paid to an area does not mean that it is remembered (Drèze and Hussherr, 2003), although there is a high correlation between the amount of visual attention (time of permanence) and the retrieval of memory (Wedel and Pieters, 2000). ...
Purpose The aim of this manuscript is to examine the influence of congruence (perceived and effective) and the level of visual attention towards sponsors on recall and purchase intention in sports sponsorship by applying neurophysiological measures. Design/methodology/approach This study is part of neuromarketing research applied to sports. The experiment entails eye tracking with 111 men and 129 women (N = 24) with 24 sports posters of three different disciplines (sailing, tennis and F1), varying the congruence, the clutter of sponsors and the position (2x2x2). The data are analysed via analysis of co-variance and regression using ordinary least squares. Findings Brand recall is influenced by the clutter of sponsors present on the poster and by the time of fixation. Effective and perceived congruence influence the purchase intention, but the full time of fixation on the sponsor does not. The latter only influences purchase intention indirectly. Practical implications The results enable managers to implement better poster designs and sponsors to have objective measures of sponsorship. Originality/value There are few studies that analyse print media in sponsorship using neurophysiological techniques. This research is a pioneer in considering attention to sports posters to examine recall and purchase intention.
Dramatic advances in the design and deployment of sensors, edge communication and computing in recent years have enabled a significant shift and evolution in education. Pervasive learning in various learning environments, including face-to-face classrooms, online learning, virtual learning, and hybrid learning, is becoming our dominant learning paradigm for students who expect highly flexible and efficient options for studying at their own time, place, and pace. Especially, in our post-pandemic present, online learning will continue to gain in importance as a significant available and required learning component. This article provides a comprehensive review of the state-of-the-art systems and studies for assisting student learning and attempting to capture student attention and engagement in many different ways. It focuses on sensors and hardware, ranging from commercial multi-sensor eye-tracking devices to open-source, low-cost systems. We also explore and present system infrastructure, data features, data processing techniques, tools and software, and key technologies that are useful for enhancing many student learning environments. Additionally, the advantages, use cases, features, and limitations of different systems and techniques are explored and contrasted, where the results of our comparative analysis are summarized in tables for readers. This review article could assist both teachers and researchers to have a better understanding of current sensors, multi-sensor systems, and related technologies for capturing student attention and understanding their performance. It can also be applied to providing practical information for novel system design, promoting ongoing research studies, and fostering sensing and technology innovations for smart education.
Full-text available
The present study attempts to examine the effects of animated banner ads, as well as the moderating effects of involvement, on each stage of the hierarchy of effects model, and to explore the applicability of the hierarchy of effects model to the banner advertising environment through an online experiment. The results provide support for the notion that animated banner ads prompt better advertising effects than do static ads. Animated banner advertising has better attention-grabbing capabilities, and generates higher recall, more favorable Aad, and higher click-through intention than static ads. Furthermore, an individual’s product involvement moderates the effects of animated banner advertising on recall, Aad, and click-through intention. However, the study does not provide solid evidence of the feasibility of the traditional hierarchical model (Cognition -> Affect -> Behavior) in the online banner advertising environment. Several implications and limitations of these results are discussed, and future research is suggested.
Full-text available
The authors report the second in a series of experiments on recognition as a dependent variable in the study of learning and forgetting of television commercials. They investigate the impact of time since exposure, commercial length, and commercial repetition on recognition and unaided recall scores. The results indicate that recognition scores are not indiscriminately high, as commonly is argued, and that they do decline with time, contrary to what often is assumed. The data, in fact, show that recognition scores are more sensitive and more discriminating than, and covary with, unaided recall scores. The evidence indicates they warrant more consideration by advertisers.
Full-text available
As with all forms of advertising, exposure is a necessary prerequisite for Internet banner ad effectiveness. However, exposure does not guarantee a user's attention, an issue especially relevant to the Internet, where ad avoidance occurs most frequently. And if an ad is noticed, the message may or may not remain in the consumer's memory after cognitive processing. However, even if the advertising message is not consciously remembered, the exposure can be unconsciously processed and subsequently change the user's affective state. To investigate how attention levels influence users, this study uses eye tracking to measure the level of attention that results from an advertisement exposure and explores how different levels of attention influence users in conscious and unconscious ways. Also, we examine the effect of animation-one of the most popular attention-grabbing tools-on attention and how it moderates cognitive processing. By measuring and analyzing users' actual eye-movement data, we found that animation in banner ads not only attracts less attention than static ads but also reduces the positive effect of attention on memory. In addition, although more than half of the participants could not recognize the advertised brand, the animated banner ad was unconsciously processed and did influence attitudes toward the brand. The results suggest that animation in banner ads does not necessarily increase user attention, but that even if a user does not consciously notice a banner ad, the user's attitude toward the brand is influenced.
Eye Tracking for User Experience Design explores the many applications of eye tracking to better understand how users view and interact with technology. Ten leading experts in eye tracking discuss how they have taken advantage of this new technology to understand, design, and evaluate user experience. Real-world stories are included from these experts who have used eye tracking during the design and development of products ranging from information websites to immersive games. They also explore recent advances in the technology which tracks how users interact with mobile devices, large-screen displays and video game consoles. Methods for combining eye tracking with other research techniques for a more holistic understanding of the user experience are discussed. This is an invaluable resource to those who want to learn how eye tracking can be used to better understand and design for their users.
Creativity is an important component of advertising. This research examines the potential effectiveness of creative advertising in enhancing recall, brand attitude, and purchase intent. Our basic methodology compares a set of randomly selected award-winning commercials (Communication Arts) with a random sample of control commercials. The commercials were embedded in television programs and subjects for a naturalistic viewing experience. Studies 1 and 2 had aided and unaided brand and execution recall as dependent variables. For Study 3, brand attitude and purchase intent were the dependent variables of interest. Results indicated that creative commercials facilitate unaided recall, but that creativity did not enhance aided recall, purchase intent, or brand and advertisement attitude. The basic advantage of creative advertising in enhancing unaided recall was found to persist over a one-week delay.
Many video ads are designed to go viral so that the total number of views they receive depends on customers sharing the ads with their friends. This paper explores the relationship between the number of views and how persuasive the ad is at convincing consumers to purchase or to adopt a favorable attitude towards the product. The analysis combines data on the total views of 400 video ads, and crowd-sourced measurement of advertising persuasiveness among 24,000 survey responses. Persuasiveness is measured by randomly exposing half of these consumers to a video ad and half to a similar placebo video ad, and then surveying their attitudes towards the focal product. Relative ad persuasiveness is on average 10% lower for every one million views that the video ad achieves. The exceptions to this pattern were ads that generated views and large numbers of comments, and video ads that attracted comments that mentioned the product by name. Evidence suggests that such ads remained effective because they attracted views due to humor rather than because they were outrageous.
The most common medium for advertising on the World Wide Web (WWW) is through the use of banners. This study investigated recall and recognition of animated and static online banner advertisements. It was found that regardless of a banner's animation state, fewer than half the participants were able to recall the presence of an ad. Overall recall was lower than recognition, however, participants unable to successfully recall the ads were still able to effectively recognize them. Results also suggest that the use of animation may enhance user memory of banner advertisements.
Internet companies collect a vast amount of data about their users in order to personalize banner ads. However, very little is known about the effects of personalized banners on attention and memory. In the present study, 48 subjects performed search tasks on web pages containing personalized or nonpersonalized banners. Overt attention was measured by an eye-tracker, and recognition of banner and task-relevant information was subsequently examined. The entropy of fixations served as a measure for the overall exploration of web pages. Results confirm the hypotheses that personalization enhances recognition for the content of banners while the effect on attention was weaker and partially nonsignificant. In contrast, overall exploration of web pages and recognition of task-relevant information was not influenced. The temporal course of fixations revealed that visual exploration of banners typically proceeds from the picture to the logo and finally to the slogan. We discuss theoretical and practical implications. Copyright © 2014 John Wiley & Sons, Ltd.
The purpose of this study is to explore the emergence of ad banner blindness in the viewing of e-commerce home pages. Building on the literature on inattention blindness and banner blindness, this article assessed the gaze path of users in goal-directed and free-viewing tasks when viewing pages with advertising banners on the right side of the page and on the top of the page above the main navigation menu. The hypotheses are tested using an analysis of variance. Using an eye-tracking methodology, the results identify significant differences in visual attention for banner ad location and for task type. Banner blindness is strongest for advertising banners on the right side of the page and for goal-directed tasks. Neither participants’ ratings of page visual appeal or of page familiarity could explain the findings. The study contributes to the existing literature by resolving some of the cognitive factors that lead to banner blindness and supplementing previous research that focused on relevant perceptual factors.
Conference Paper
The purpose of this study was to examine how human brand image appeal affects visual attention using eye-tracker, a visual attention measuring apparatus, at an e-commerce website. Additionally, we examined the effect of human brand image appeal on purchase intention using a questionnaire method. We conducted an eye-tracker experiment, collected survey data, and conducted interviews with each participant using laptop and perfume products. After collecting 108 valid data, the human brand images were divided into three groups: high human brand image appeal group, low human brand image appeal group, and no human brand group. We applied MANOVA and t-tests to analyze the data. The results showed that the level of human brand image appeal has a significant influence on a consumer's visual attention and purchase intention towards a product. Both visual attention (human brand and product AOI) and purchase intention are highest for the high image appeal group.