• by May 20, 2005

In the last week alone, I have had three clients separately ask the exact same question: "How much is my online advertising and paid search campaigns contributing to my retail store sales?" This usually would not raise an eyebrow, except that last week two other clients asked me: "How much is my offline media influencing my online sales?" Agencies like the one I work for receive these cross-media questions all the time, but as of late I have noticed they are being posed with dramatic frequency. To me it is another sign that we have reached an industry tipping point. For both big and small advertisers alike, the online medium is shifting from isolated little brother status to "Oh my God we better figure this thing out cause we're about to spend a boat load of cash with Google!"

To answer the questions posed by my clients is not that easy. With the proliferation of different media choices and the advertising that accompanies them, one cannot simply conclude that a jump in retail sales was due to a spike in banner advertising at about the same time. Consumer behavior and their purchase pathways are far too complex to draw simple correlations, even though your gut might be telling you differently. It's possible that an Internet user sees an online ad and then goes directly to a retail store and buys the product. But life is usually not so straight-forward! More likely, the consumer sees a company's online ad, two days later sees its TV spot, later that night does a search on Google for its product, visits the company's Web site, goes to sleep, again sees the TV spot during the morning news, and decides to visit the store a week later.

So when clients ask how much their online presence contributes to retail, assigning a value in the context of complex consumer behavior becomes tricky business. The value of online cannot be considered in isolation but rather as a contributing factor in a chain of influences. A multitude of variables beyond advertising have to be considered - seasonality, promotions, competitive activity, and partnerships, to name a few.

If you've ever been in a meeting where the words "Econometric Modeling" were mentioned, chances are the questions in this article seem very familiar. Econometric modeling is the statistical practice of determining the contributing factors to a desired outcome and assigning a value to each one of them. When the volume of a client's sales move up and down, does that variance have more to do with online media activity, TV advertising, the interaction between the two, or something else, perhaps? Econometric modeling attempts to parse out those influences to answer what the advertiser really wants to know: "What am I getting for my channel investments?" "And, what should I optimally spend with each?"

Just as consumer behavior runs across multiple channels, determining return on investment and optimal spend levels has to be considered holistically, in the context of the entire media mix. Online advertising budgets over the past year have risen to become anywhere from 3 percent to 20 percent of a Fortune 500 company's media spend, depending on the business category. This has given rise to a lot of nervous clients who want to know exactly what they are getting in return. The consequence has been an ever-growing minion in the online agency world, lovingly called "The Geek Squad" (to which I proudly belong). And No, I am not talking about those Best Buy flunkies. We are the group of Ph.Ds and MBAs whose job it is to assign an equation to the purchase path. We are the ones leading the charge up the crumbling wall, peering over, and yelling to the retail clerk, "Hey, 6.7 percent of that sale belongs to us!"