January 25, 2023
Commentary
Data-Infused TV Buying Is A Tricky Business
- by Eric Fischer , Op-Ed Contributor, December 14, 2017
In grade school, we learned fractions have three parts: numerator, denominator and quotient. And the denominator can never be “0.”
Today, we seem to be inundates with case studies of how data-infused TV buying drives efficiency vs. historical TV buying. Translating that statement into a fraction, the ROI for the data-infused approach’s (numerator) would be greater than the traditional method (denominator), with the quotient being greater than “1.”
Seems logical, but as someone who’s been in the middle of this my entire career, something struck me that didn’t seem right.
Given the decades-old hand wringing about not having good data to tie results back to TV spending, how can this fraction even exist in the first place? In other words, if we couldn’t prove how TV spending performed before (the denominator is “0”), how can we prove this way is better? Better than what?
advertisement
advertisement
Today, as marketers try to develop hard ROIs against their TV spend, it’s only logical to use data to better target audiences and eliminate waste. And while the ability to attribute success against TV spend is a much-needed first step for many marketers, the macro question remains: Are we confusing the ability to establish a data-infused performance baseline with success vs. the traditional age\sex demographic approach?
Developing a baseline provides for an initial starting point on a graph. Subsequent campaigns provide additional data and dimension, so trends can be developed. What may not be possible is comparing a baseline against what was done before and claiming victory. It may well be the case, but it’s a fallacy to think we have the data to prove it.
Claims of increased performance from a data-infused TV buying approach vs. historical buying, should be taken with a grain of salt. We’re essentially assigning an outcome to an equation where our denominator is “0.” Confidently comparing the old way vs. the new way may also seem a bit duplicitous — it carries the assumption both marketers and their agencies had high degrees of certainty regarding the ability to measure historical TV buying performance.
Astute marketers and agencies are measuring results of data-infused TV buys across multiple campaigns. They are comparing and optimizing accordingly, rather than falling prey to the faulty logic of comparing these campaigns against the traditional TV buys where no definitive baseline was established.
The contrarian may see the inability to forge conclusive comparisons against the traditional TV buying approach as the opportunity to raise the question of whether the results associated with data-infused buying warrant the premium placed against that inventory.
The “sell” side will continue to profess the benefits of this data-infused approach. They need to find rationales for raising the unit pricing in the face of shrinking GRP supplies, as well as finding new ways to sell distressed inventory.
For the “buy” side, the essential question remains: How can the benefits of a data-infused approach be proven in the absence of knowing the ROI of the previous approach?
The answer may be the realization that these two approaches are apples-and-oranges. The real solution may lie in a deeper dive as to what the benefits of audience segmentation and addressability could mean for the brand. How we choose to proceed is the real question.
Agree? Disagree? Let me know what you think.
You raise some interesting questions, Eric. Contrary to what some think TV ad campaigns are invariably evaluated based on results---be they ad awareness, sales, bolstering a marketer's image, impressing "the trade", etc. or combinations of these and other factors. What is being proposed, when it is boiled down to fundamentals, is that each TV commercial exposure in each program "vehicle", must be evaluated separately as to its specific contribution to ROI. Hence the constant, DR -driven chant about "attribution". Here is where we go astray as it is virtually impossible to isolate the impact of any individual commercial exposure from the rest of the ad campaign, as well as the effects of other variables such as word-of-mouth endorsements, how well the product is distributed, the impact of competitive promotions by rival products, etc. Even if branding advertisers abandoned their retail partners entirely and switched to a 100% direct response model, it would be difficult, at best, to determine what the exclusive result of any single ad exposure was---unless the ad was shown only once to each consumer---an unlikely event.
My point is that it's is fine to target your ad exposure tonnage as best as you can, but that's all you are doing---dealing with tonnage----which means that the audience of each commercial is duplicated by exposure to the identical sales message an hour ago or yesterday or last week, while some of the other variables noted above also play a role.. Consequently, you are better off monitoring the cummulative effects of your campaign, as it builds reach and frequency, rather than trying to judge the contribution of every component of media audience as if it played out in isolation. It doesn't.