Be Wary of Online Ad Campaign Metrics

The first client I ever worked with was president of a restaurant chain with locations around the world. He was famous in our firm for demanding advertising that “makes the cash register ring.” There’s nothing to argue with there and we were determined to do our best to give it to him.

What couldn’t escape my notice, however, were all the things that got between our good intentions and that lovely sound at the end of the transaction. Even as a marketing newbie, I remember thinking it was a little simplistic to expect advertising to not only initiate, but complete the sale. I knew that wait times, table cleanliness, product quality, and staff training—among many other things—contributed to (or detracted from) our success.

All other things being equal, it does make sense that two different ad campaigns would have two different effects on sales. In this respect, they would be measurable. But running two different campaigns is not always practical; it especially wasn’t in pre-Internet days. Nor are all other things frequently equal—one of the intractable nuisances that plagues any form of advertising research.

Now that it’s easier than ever to track online advertising campaigns (which surpass cable TV, broadcast TV, radio, and newspaper spending, according to research firm EMarketer), there are increasingly abundant ways to use metrics to improve our marketing. By putting the proper pixels on the proper pages of a website, for example, a company can see how many visitors clicked on a banner ad and ended up purchasing a product online or subscribing to a newsletter. What’s even better is that we can now know if someone who was served an ad—and perhaps never took the initiative to click on it—later made a purchase. That, in and of itself, is remarkable. What we can’t know, however, is how many messages, in what forms of media, that consumer was exposed to prior to making the purchase. That’s where the risk arises.

Online metrics are magnificent, but they’re only a small piece of the pie, and you need to think twice about giving sole credit to the ad that is identified with the sale. Think about your own consumer behavior: How often do you respond to the first ad you see for a purchase of any kind? Depending on the product category, it could be days, weeks, even years before a brand’s advertising convinces you to buy the product, and it will usually result from the combination of all the brand’s marketing messages to which you were exposed. Just because metrics are precise doesn’t mean they’re correct.

Tough Questions About Numbers

Consider retargeting, for example, which is the practice of serving ads to someone who stopped by your website after they’ve moved on to other online activities. Was the consumer that purchased after clicking on a retargeted banner actually converted by that banner, or was it just a convenient link? Might they have purchased with or without being re-served the ad? Questions like these can’t be easily answered and as a result, a given transaction’s “return” might be ascribed to the wrong (or partial) “investment,” resulting in a flawed return-on-investment analysis. Still, there’s something deceptively alluring about those online metrics. They look so firm, so final, so authoritative.

Be careful. Marketing consulting firm CoreBrand has conducted years of research about the long-term effects of marketing investments and has demonstrated a link between consistent spending and growth in the power of a brand. Those companies that maintain consistency in their marketing efforts over time gain efficiencies in their ability to influence customers—an effect that can neither be seen nor realized in the short term. (See “The Case for Brand Accretion.”) As the world of marketing metrics becomes increasingly oriented toward instant gratification, it will be easier to make a short-term decision that’s not in your long-term interests.

When it comes to tracking real ROI, by all means gather as much data as you can. Make sure you think globally; consider how other aspects of your comprehensive marketing program may be affecting the metrics you’re seeing. And be wary: Even the best online data can be incomplete or at times, incorrect. Companies that judge their efforts only by the immediate gratification of easily measured metrics may miss the real story.

In business, as in life, just because you can’t measure something doesn’t mean it’s not important. I can’t put a metric on how much I love my wife, but I know it’s a whole lot. What’s the ROI on vacuuming your store? Of re-striping your parking lot? Of being kind to the shipping guy? Of smiling at your employees? It all adds up. Measure ROI in as many ways as you can, in as much detail as you can, and as close to real-time as possible. Just don’t confuse metrics with success. They’re not always the same thing.