Commentary
Better Will Always Outweigh Perfect In Audience Targeting
- by John Piccone , February 20, 2015
Of course, there were leaps made in measuring the relationship of TV advertising to sales, but perfect did not seem to get in the way of better. The media marketplace understood that over time, things would naturally progress -- if the drum beat of “what if?” and satisfaction with “better than, though not perfect” set the tone for decisions.
Now it seems that drum beat is getting louder as we enter a new dawn of television advertising, a dawn where the influx of better data has expedited the interpretation of what matters and what doesn’t. And although the mission to understand the relationship between TV advertising and business outcomes has not changed, the tactics that were once thought tried-and-true are now up for reinterpretation.
advertisement
advertisement
Today media outputs such as GRPs, CPM and Day Part Mix rule the roost for both pre- and post- campaign pricing and value metrics. And ratings have for years been the principal metric for audience attention through general personal attributes such as sex, age and gender.
Should we expect these fundamental elements to go by the way side because better data is on the table? Of course not. Metrics such as CPM and GRP are stable means through which to transact. However, to categorize audiences by only day-part viewing or age and gender is not good enough. And this is where tactics will change quickly.
Thanks to first- and third-party anonymous purchasing data, media outputs are being tied to business outcomes such as transactions, store visits and basket size volume. As a result, marketers can see how X number of GRPs at Y CPM contributed to Z dollars in actual sales -- specifically measured against a host of new audience attributes such as “Spent on average $250/month at Macy’s” or “Goes to Pizza Hut 3 times per week.”
But it isn’t enough to understand that those people were reached and took an action at these respective merchants. What also matters is understanding the customers who were not exposed to paid television advertising yet also spent time and money there. Thanks to second-by-second commercial viewing data, we can now measure “the lift” of exposed versus unexposed audiences on sales activity. And when marketers can prove their TV advertising had a positive return versus unexposed audiences, TV advertising moves from a cost center to a profit center, just as search engine marketing has done.
Last but not least, by measuring sales against different audience types, marketers will get a clearer picture of who is and who isn’t their current customer. This will not only allow them to steal share from the competition but will also allow them to reinterpret previous television advertising activity to determine the right media outputs to drive the best business outcomes.
In today’s hectic marketplace, the spoils will go to those who don’t let perfect get in the way of better.
Very well said John. And it's not just with TV Ads. The same problems (or I should say "challenges") exist in Radio Ads and Print Media. Whenever you try to work out a campaign that deals with disparate technologies (such as you have mentioned - TV Ads and website hits, etc.) you are going to make compromises and doing it "just right" or "just better" will always be more fruitful than finding "the best" way...
John, with all due respect, in times past, even if the advertiser didn't know exactly who was exposed to his TV ads---only that they "reached" about 70% of the target group per month, an average of 4 times------ the results, if any, were reflected by sales response----or, if ad awareness was the goal, by awareness studies. So, what's so different now with the "better data" you are talking about? You mention second by second commercial "viewing"data and the ability to track the "lift" in sales that results from commercial exposure that access to such information gives the advertiser. That's great, but what "viewing" data are you referring to? Surely not the set usage ratings used by various "big data" researchers? The scenario, you paint, reads like all of this is happening right now, yet in reality, it's mostly theoretical and is not being practiced to any meaningful extent by anybody. Some of it may come to pass----no one, least of all myself----is against progress, when it is real progress and is based on operationally valid research----- but it's not accurate to imply that until now, advertisers were flying blind, using only outmoded yardsticks like GRPs to gauge the impact of their TV ad campaigns. Those with any brains at all routinely pre-test their campaigns,they track both sales and awareness relative to audience delivery, they test alternative media mixes, weights and scheduling concepts via single source scanner panel facilities, they conduct ongoing commercial recall and attitude-change studies, etc. etc. And they are well aware of whats' working and not working. If we are talking about better ways to buy media, especially in the targeting area, that's another matter and here some improvements are needed. But these aren't coming until we find a way to measure commercial "viewing" ,not just set usage, and relate this to product usage, as well as more sophisticated indicators like consumer mindsets, ad receptivity, etc.
Ed - you are right to point out that there have been excellent proxies in place to measure the impact of TV. In addition to these tactics the arrival of first and third party direct matching data sets, anonymized through third party matching companies, is indeed happening today. And although not perfect, it will at the very least inform the accuracy of historical models and separate the wheat form the chaff moving forward.
Bravo John. While perfection is always the ultimate (even if unattainable) goal, the only way to bring things forward is: Progress now, perfection later