ROI is the new rallying cry for marketers beset by budget squeezes, clutter, fragmentation, a do-it-today mentality and a tougher competitive landscape than any in the past half-century. It impacts compensation negotiations, agency reviews and the endless search for competitive advantage. In response, media agencies have rushed to provide "econometrics," a scientific process that directly links advertising to sales results.
Every large manufacturer of consumer goods in the U.S. and Europe has or is developing an econometrics modeling capability, either internally or through agency partners. Indeed, isolating that tricky "50 percent that works" has become the real cornerstone of much of the new thinking in media planning. But the imperative is honored mostly in the breach: Promises to increase advertising return on investment are too often couched in vague terminology and left open as a catch-all to excite advertisers without being specific enough to mean a quantifiable shift in measurable delivery.
Making the quest even more difficult is that, like beauty, the "return" in ROI is in the eye of the beholder. Take, for example, a manufacturer of laundry detergent:
To the brand manager, advertising is an investment in maintaining the brand's medium- and long-term image and loyal users. The brand manager knows that advertising does not typically work by making people leap off their sofas and rush down to Wal-Mart because they've just seen a 30-second spot.
To the financial controller, ROI is the incremental profit derived from an immediate uptick in sales from advertising compared with media costs. Advertising is simply a cost of doing business.
To media buyers, ROI means maximizing delivery of target-audience impressions in the planned programs and dayparts for the client's budget. It's less about consumer behavior and more about realizing reach and frequency targets irrespective of sales.
Is econometrics another "mystery black box" solution? Will it convince the CFO, the brand manager and the media buyer to love and respect each other? Well, no and probably not, respectively. Sales-modeling techniques that have been developed during the last five to 10 years try to answer the ROI questions for all parties at the same time. These techniques, grouped together under the term econometrics, involve building statistical-regression models of brand sales, breaking down all of the constituent elements that influence a consumer's brand-purchasing decisions.
We know that every brand sale is based on a variety of influences: primarily absolute and relative price (the total cost of the item and the brand's price compared with that of its direct competitors) and brand distribution. Competitive advertising pressure, category innovations (and other nonadvertising influencers) and advertising itself all have varying degrees of effect as well. Using econometric modeling (and detailed client data), it is possible to strip out sales prompted by all the nonadvertising factors. What is left is a measure of the short- and long-term effects of advertising on brand sales. But it's a "look-back" technique that can measure only the factors we know to look for and have data to build models with. Larger questions concerning evolving markets, new-product launches and consumer sentiment must still be addressed.
The deeper we look into the effectiveness of advertising, the more we learn about the importance of building contextual relevance among the brand, the message and the media environment. We have to marry this contextual research to our historical modeling and business-planning skills to build a better picture of what the future might hold.
So, can econometrics tell us which 50 percent of the ad budget to eliminate? Not quite. We are creating models based on historical behavior. In the economist's perfect world, everything other than adspend in a model remains the same, and consumers make the same rational choices again and again. But as we all know, life—and the brand buyer —is not quite like that. Still, econometric modeling can give marketers a much better understanding of what has worked in the past and what hasn't—a great first step in figuring out what will work in the future.