Ad measurement—3 ways to effectively evaluate marketing performance

The advertising industrial complex reports in averages. Average CPM, average CPC, average TACoS (total advertising cost of sales), average ROAS (return on ad spend)—the list goes on. Our industry literally turns on these numbers. This hurts us all—especially brands—in profound ways. It’s a primary, yet silent, obstacle to advertising regaining credibility in the modern business world. Media’s largest platforms—Google, Meta, Tik Tok, Amazon—all have incredible data and reporting capabilities. Want to know how often moms with dogs are checking out or rebuying your product in Bel Air vs. Hollywood during the noon hour? Want to compare ROAS on Tuesday mornings in Atlanta and Savannah? They can tell you—and much more. That detail makes it possible to put marketing on a foundation of unprecedented precision. A-List & Creativity Awards Enter for a chance to be recognized as the best in advertising Submit your entry here Yet we default to averages to make work easier and more profitable. We use averages to keep commissions rolling on media that might be underperforming. And we use averages to keep from doing the in-depth labor that precise data would beckon; it would make many (if not most) businesses unprofitable on both the buy and sell sides of media. In the process, we lose a key to real efficiency: the ability to see exactly which things are working and not working and understand what is incremental. On an average basis, it’s easier to look good. Let's pretend Adidas spends $10 million on search and every dollar is laid out in an illustrative Excel sheet. Row 1, the first $1, has a ROAS of 50:1 or more because it’s a brand term (the first thing bought). The 5-millionth dollar, likely for a specific shoe model, is still producing a 6:1 return. But by the time we get to $8 million, the dollar is fishing for revenue in category search “running shoe” or competitive conquesting “Nike Running Shoe” to earn 25 cents on the dollar. The last $2 million produces negative return. That’s good and acceptable if it’s a strategic hunt for new customers, but bad if it’s simply hiding within the average. Similarly, in paid social, prospecting and remarketing campaigns too often get lumped together in a blended ROAS. The blended performance looks good, but we can’t be sure what’s working. We’ve lost the vision necessary for improvement. The same problem happens with time. Over a month or quarter, averages mask extremes and distort time. They invite impulsive moves where some media is diagnosed as bad too fast. Some products take more time and more touches to produce returns; those particularities get revealed with more tailored and micro measurement. Why does this happen? The answer is labor economics. Understaffed agencies and clients alike lack the resources to tackle nuance. And commission structures incent underpaid agencies to play the averages that support higher spending. We can overcome the tyranny of averages by taking collective action. We need to be willing to do some things differently and ask different, harder questions. NextGen Marketing Summit Join us for a crash course on Gen Z, DTC marketing and the creator economy Register today Look line by line For starters, we need to challenge ourselves to look at every line. In the search example, it’s identifying the pieces and time slots that are most productive for the given strategy. It's asking for brand performance to be separated from the overall campaign and 10 other questions such as this one. That way we can evaluate media investments for what they are actually doing. Test for incrementality Do this ruthlessly. We’re going to hear the term iROAS more, meaning “incremental ROAS”—return on ad spend that’s beyond what would have happened anyway. To test and prove it, we need to lean on market-mate tests, media mix modeling and old-fashioned structured incrementality tests (as in, ad-exposed and non-exposed comparisons). That way, we can answer what the last dollar does. Sharpen goals Especially when budgets tighten, we need to pursue and measure what’s most important to the strategy. It’s not always ROAS. For example, it could be increasing long-term customer value or lowering the cost to reactivate lapsed customers. It could be ROAS for new customers, but lifetime value (LTV) for existing customers. There’s nothing average about the demands of market competition and marketing. Boards, CEOs and investors are asking much harder questions to get a granular picture of what’s really propelling their brands. We’ll find the answers by trading what’s expedient for what’s truly effective. Check out the latest Ad Age events and award programs

Ad measurement—3 ways to effectively evaluate marketing performance

Featured Stories