Banner Ad Location Effectiveness Study: Theories | WebReference

Banner Ad Location Effectiveness Study: Theories

Theories and Methodology

Banner Ad Placement Study

Hypotheses

We tested three theories for the impact of location on click-through:

  1. Ads next to the scroll bar will generate a higher click-through rate. Users are more likely to click next to the scroll bar because their pointer is already on the scroll bar, therefore it is convenient to click on the ad. (Link to study)
  2. Click-through will be higher if the ad is placed approximately 1/3 of the way down the page as opposed to at the top of the page. It is more likely that the user's eye will first focus 1/3 down the page. (Link to study)
  3. Two ads on a page (same product) will generate a higher click-through rate. It will be convenient for a user to click on the second ad at the bottom of the page, after reading the content of the page. (Link to study)

All data was collected on Webreference.com between 3/25/97 and 4/15/97.

Variables

Dependent Variable:
Click-through rate = # clicks on the banner ad/number of page impressions

Independent Variable:
Location of ad on the web page (3 different locations were tested).

External Variables:
These variables were identified as important to the determination of click-through, but were external to our study. Therefore, after each variable is a description of how we attempted to isolate the effects of the variable from the study.

1. Banner ad content
The same ad was used in both locations for each of the three studies. Then, the relative impact at each location was measured by comparing the differences between the click-throughs. Two different companys' ads were also tested for each study.

2. Matching of the ad content to the web page
The ads were run on the same page at both locations for each study. Then, the relative impact at each location was measured by comparing the differences between the click-through rates. To the extent possible, advertising was matched to the content of the page. Due to limited time available to run the studies, placing the ads on high traffic pages, linked to the front page of the site, was a high priority.

3. Banner ad burnout
This is the phenomenon reported by Double Click where the number of click-throughs for a given ad declines quickly with each view of the ad. Our goal was to control this variable by using all new advertising for this study. In the actual studies, however, the goal was not always obtainable. The following ads had new artwork shown for the first time on Webreference.com during this study: Star Wars/Hasbro, ShopSite, and Digital Frontiers.

4. Ad size
Only two ad sizes were used for this study. A 468 x 60 pixel ad was used as the full-size ad running across the page. The 125 x 125 pixel ad was used in the first study to measure the top of page vs. side of page location. The 468 x 60 pixel ad could not be run at the side location, which was the rationale for including the smaller size ad.

Note on click-through:
Click-through percentages should be evaluated by their differences within each study. The absolute percentages are not comparable accross studies due to the external variables.
Back to Home

Comments are welcome

 


Created: Apr. 21, 1997
Revised: Apr. 27, 1997

URL: http://webreference.com/dev/banners/theories.html