Transcript Slide 1

Online Advertising Recall:
What Works – Image or Information?
Eric Van Steenburg
University of North Texas
Introduction
• Online ad click-through rates (CTR) are falling (Drèze &
Hussherr 2003; Hollis 2005).
• Trust in the value of online advertising has nosedived,
leading to a decline in online advertising expenditures
(Business Week 2007).
• Who decided CTR is the be-all, end-all measurement for
the success of an online ad?
• Who decided that online ads were purely for sales lead
generation and could be measured like bingo cards,
coupons, and 800 numbers?
Background
Researchers divided into two camps:
1. Measure advertising effectiveness by CTR.
2. Measure effectiveness by awareness using recall
and recognition techniques.
The function of Internet advertising is to influence people in
the same way as advertising in traditional media – brand
enhancement (Drèze & Hussherr 2003).
If the banner ad is repeatedly noticed but not clicked on,
recognition and awareness of the ad and brand name is
generated (Briggs and Hollis 1997).
Background
Advertisers list brand awareness and recognition among
the top three objectives for online marketing activity (Hollis
2005). Yet they tend to follow the CTR method of analysis.
That leads to even more questions:
•
•
•
•
•
•
Should the ads be static or interactive?
How big should the ad be?
Should they be use color or not?
Where should they be located on the page?
Should they feature a promotion or not?
Should they even attempt to generate CTRs or try to reinforce a
brand?
Questions
• Can online ads support a brand?
• How well do consumers recall and recognize ads
designed to support a brand?
• How well do consumers recall and recognize ads
designed to support a product?
• Do consumers recall one type of online ad more
frequently than another?
Theoretical Framework
• Elaboration Likelihood Model (Petty et al. 1983) –
Consumers use either a central or peripheral route to
information processing. It can be a good determinant for
advertising effectiveness in a Web environment based on
consumer goal-directedness and involvement (Wang et al.
2009).
• Need for Cognition (Cohen et al. 1955) – Operating at the
elemental level, it measures the extent to which
consumers have an intrinsic motivation to engage in
problem solving activities. Research has shown the effect
of NFC on message elaboration, message recall,
motivation to process information, attitude change, and
message persuasiveness (Haugtvedt & Petty 1992).
Methodology
Drèze & Hussherr (2003) found that Internet users:
1. Avoid looking at ads while online
2. Perceive banner ads in their peripheral vision.
3. Affected most by the banner message rather than
how the message was conveyed.
Experimental design (2 x 2 x 3)
• Ad Type (Image vs. Information) x NFC (high vs. low)
x Recall (unaided vs. partial vs. full)
• Manipulated Ad Type
• Measured Need for Cognition
• Measured Recall
Methodology
Unaided Recall
•
•
While looking at the website pages, do you recall seeing one or more
banner ad? 1–Yes 2–No
If you answered “yes” to recalling seeing one or more banner ad,
please enter the brand(s) or company(s) advertised.
Partially Aided Recall
•
•
•
For the following brands or companies, please click all those that you
know or have heard of.
For the following brands or companies, please click all for which you
could identify the logo.
For the following brands or companies, please click all for which you
recall seeing a banner ad.
Fully Aided Recall
•
•
While looking at the website pages, do you recall seeing the following
banner ad? 1–Yes 2–No
If you answered “Yes” please enter the name of the brand advertised.
(Drèze & Hussherr 2003)
Methodology
This study is being done to examine Internet users reaction to news websites. Please
read the following information about a top story in the news before beginning.
WASHINGTON, D.C. – In the wake of a growing scandal and the
largest car recall in automotive history, Toyota announced today it is
considering firing its long-time CEO, Akiro Toyoda, the grandson of
the company founder. Should the world’s largest automaker fire
Toyoda, it would mark the first time in company history that a
member of the family was not at the helm.
You will now see a series of web pages where you can read the latest information on this story.
After viewing the web pages, you will be asked a few questions to check your reading comprehension.
Please click "next" to continue
Just Do It
Think Different
The Relentless Pursuit of Perfection
Nike’s new Glider.
Because not every race starts at 7 a.m. on a Saturday
Apple Macbook
Lighter than air
The new hybrid Lexus
Perfect for today’s Economic Climate
Methodology
Filter Questions
• Reading Comprehension
• Favorite news sites
• Attitude toward Internet news sites
Measured
•
•
•
•
Internet usage
Need for Cognition
Recall
Attitude Toward Brand
Demographics
Results
General Linear Model (Petty et al. 1983; Lee & Thorson 2009)
Within Subjects
• Recall significant (p-value < .000)
• Interaction between Recall, Ad Type, NFC significant (p < .05)
• Interaction between Recall, Ad Type not significant
• Interaction between Recall, NFC not significant
Between Subjects
• No significance (all ps > .1)
Covariate (ATB) not significant
Results
Recall – Unaided
• NFC has negative correlation to message complexity.
High
Low
Results
Recall – Partially Aided
• High NFC consumers had much greater recall of complex ads
High
Low
Results
Recall – Fully Aided
• Low NFC consumers have much greater recall of Image ads
High
Low
Conclusions
• Image-based online advertisements are more readily
recalled by low NFC consumers after provided cues to
their message.
• High NFC consumers recalled complex messages more
easily after only a few cues.
• Level of NFC in consumers appears to go against theory
when it comes to complexity and recall of online
advertisements.
• For marketing managers, message association may be
more important than message content ~ have an online
message strategy to augment mix.
Future Research
Do incentives in ads affect recall?
• Banners that mention neither price nor promotional offers have
the most effect on click-through and post-impression rates (Rettie
et al. 2004).
How does repetition of ads affect recall?
• There are relatively strong diminishing returns to early repeated
exposures that taper off as exposures continue. The rationale is
that the first exposure provides sufficient opportunity to elicit a
response, similar to print advertising (Calder & Sternthal 1980).
Thank you