Analysis of a Study on Online Advertising and Click Through Rates

The researchers investigated the influence product involvement has on people's responses to banner advertisements. Other studies indicate that consumers with low involvement are less likely to want more information on the subject of the ad. The researchers cite other studies that indicate people with low involvement focus on peripheral cues such as pictures, endorsers, appetizing food, and expert sources. They also cite studies that show the size of the stimulus has the most effect on consumer's attention.

PURPOSE/MAIN IDEA:

Building on the ideas that consumers with low product involvement are less likely to want more information, and focus on peripheral stimuli, researchers set up their study. The study involved three hypotheses. The first hypothesis (H1) stated that people with high product involvement "are more likely to click a banner ad than those with low product involvement" (p. 624). The researchers' second hypothesis was divided into two parts (H2a and H2b): people with low product involvement will be more likely to click on an ad if the ad is large or animated.

SAMPLING:

Researchers used a nonprobability sample, recruiting participants through 165 separate LISTSERV lists. A total of 751 participants were randomly assigned to one of two groups.

PROCEDURE:

Each group was exposed to three banner ads - including American Express, Kodak film, and American Airlines. Each ad was placed at the top of a website containing a movie review. The first group was shown an American Express ad, a small American airlines ad, and a non-animated Kodak ad. The second group was shown the same American Express ad as the first group, as well as a large American Airlines ad and an animated Kodak ad. The American Airlines ads were identical except for size. The Kodak ads had similar variations.

Each participant's product involvement level was rated before exposure to the three advertisements. Subjects were then shown the first ad (American Express). If they did not have previous exposure to the ad, they had the option of clicking on the ad or clicking 'next' to move to the next page. If the banner ad was clicked, the participant was directed to a linked target ad. The same procedure was followed for the remaining two ads. Subjects were then given a questionnaire to determine demographic information.

RESULTS AND CONCLUSION:

Researchers found a positive association between level of involvement and clicking selection, supporting their first hypothesis. Click-through rates for the large American Airlines ad were higher than that of the smaller ad among the subjects with low product involvement. For those with high product involvement, the difference in click-through rates was not significant. The animated Kodak ad yielded higher click-through rates than that of the static ad among participants with low product involvement. There was not a statistically significant difference among those with high product involvement. These findings supported both halves of the second hypothesis.

CRITIQUE:

There were several limitations to the study, pointed out by the author, that affect the validity of the test results. Since the samples were gathered through a nonprobability method, the samples don't represent general internet users. The subjects independently decided to participate. The web surfing environment was very artificial. Participants could only click on the ad or on the 'next' link. This could have reduced the chances of people interacting with the sites in their normal manner. Questioning participants about whether they had previously seen the first ad may have primed them. Participants' attention may have been focused towards the remaining banner ads more than normal.

If random sampling had been used, the sample would have better represented the general population's behaviors. Utilizing more realistic web pages would have increased the validity of test results. Web users typically have several options on any given web page, while test subjects had only two. As the author suggests, asking participants about prior exposure to the first ad would have been more effective at the end of the study. Then researchers could be sure they hadn't inadvertently diverted added attention to the ads.

However, the apparent link between low product involvement and high click-through rates could be very useful in advertising. If ad designers are targeting an audience with low product involvement, using peripherals in their ad may increase their response rate. Conversely, if advertisers target those with high product involvement, their time may be better spent developing content in their ad, rather than peripherals.

Source:
Cho, C. (2003). The effectiveness of banner advertisements: Involvement and click-through. Journalism and Mass Communication Quarterly, 80(3), 623-645.