by: Josh Hawkins

How should we measure the impact of online advertising? A good question considering the ubiquity of contextual and search advertising, increasingly rich display ads, and rapidly expanding use of video pre-rolls, which together will total close to $19.5 billion in 2007. According to Professor Chan Yun Yoo, at Kentucky's School of Journalism, it's really a question of memory, recall and brand building, as opposed to a simple "click-through" measurement exercise.

In a recent round of research, Yoo suggests that marketers using online advertising should look at two distinct memory models when answering this question: "explicit" and "implicit" memory. Explicit memory involves retention of specific information contained in advertising, such as product features, endorsements and tag lines. Implicit memory, on the other hand, is a summary impression that is only evoked during subsequent encounters with the brand. What he finds is that regardless of whether a consumer pays attention to an online ad or clicks on it, the ad has a significant impact of implicit memory and subsequent brand judgments. Therefore, impressions based on exposure to online ads can be just as important as click-through if marketers expand their ROI expectations to include brand building. 

The brand building impact of online advertising is given a lot of lip service, but there have been few attempts to develop a framework to analyze this dimension of interactive marketing. With his research, Yoo concludes that advertisers and marketers should use impression-based metrics in conjunction with performance indicators like click-throughs to determine the effectiveness of online advertising.

But this line of thinking begs several important questions for marketing professionals. For example, how should we go about measuring "brand building" that results from exposure to online advertising? Which form of recall based on these different memory constructs has the strongest impact on behavioral intentions, which is at the end of the day what we really care about. And, if one form of memory is a more powerful predictor of purchasing behavior over the other, can we manipulate it? 

On the first question, the first step should be to clarify what we're talking about when it comes to branding - a notoriously squishy concept. Most often when we talk about branding we really mean brand association, e.g., traits, attributes, features, solutions we associate with a particular brand. There are several measurement approaches available to get at the the brand building effects of exposure to online ads. Analytics packages and various ad tracking services (e.g., AdIndex) can be used to segment website traffic into different groups, one exposed to online advertising, another control group that isn't exposed. With this technique, a sample from each group is intercepted with a survey as visitors leave the site. The survey uses a battery of questions to measure brand association and simple means tests are applied to look for significant differences between the groups. Selection bias and other measurement problems aside, marketers can gain access to general information about brand impressions and add a new ROI metric for online advertising beyond the common ratio of impressions to click-through. 

The other questions are more complicated and harder to get at. It's unclear from Yoo's research which memory model is more important for determining behavior, or how the two interact to shape, for example, purchasing decisions. However, there is some research from related fields that suggests more focused attention and recall of specific brand information is more likely to result in behavior. And attention motivation is typically associated with personal self-interest or a trusted or socially important reference group. This obviously supports word-of-mouth marketing strategies. And by extension of Yoo's research, exposure to online advertising could condition subsequent brand exposure via recommendations and endorsements. The most powerful ROI for online advertising may actually be to prime the pump for word-of-mouth brand communications - a good question for future research by Yoo and his grad students 

Leave a Comment