Thanks to digital marketing, we can hope to achieve the marketer’s dream…deliver the right message in the right way to the right person at the right time.
“Right message” is a creative question. “Right way” is media optimization. Marketers and their agencies focus on those. “Right person/right time” is an audience creation question and marketers and researchers aren’t good at that yet.
What is a high performing audience? Evidence from my recent projects suggests that if marketers can get the “right person/right time” part of the equation right, the returns would be enormous. Recently, I have been involved in four studies which all suggest that precision targeting of the right person at the right time can have a 10X effect on marketing ROI. Yes, TEN TIMES the marketing ROI! Even if we lower our sights to only 2-5X, it is hard to imagine that there is anything marketers could do that would have greater yield than getting “right person/right time” right.
But consumer segmentation is such a standard arrow in the quiver, how can it be that we aren’t paying enough attention to getting that right? Because we have blinders on from the way that we buy advertising via mass media and the way legacy segmentation works.
Marketers’ blinders. Mass media doesn’t think in people, it thinks in mass audiences and profile summaries. It thinks top down not bottom up…it doesn’t think about individual profiles, it thinks in aggregated profiles of those who watch a TV show. Mostly, it still thinks in (ugh) demographics rather than behaviors.
Researchers’ blinders. Mostly legacy segmentation studies are primarily attitudinal because they have a purpose of inspiring the marketing team, providing a north star for innovation and messaging. This might have strategic value but it is unlikely to drive up campaign ROI very much. They #fail to connect to digital profiles in your DMP. We #fail to judge segmentation usefulness by the only validity check that really exists…if I target a particular segment in a prescribed way with brand communications, does it drive up marketing ROI against my objectives? This just isn’t in the DNA yet. And finally, we fail to measure if audiences that we defined in much more sophisticated ways to be reached programmatically, were even the user types our campaign reached! We measures impression delivery, but to whom?
So let’s fix all of this by solving for segmentation and audience creation simultaneously.
How segments and audiences relate to one another. You need to drop the implicit assumption that segments and audiences are the same thing. Think of high performance audiences as the end goal and consider segments to be building blocks for audiences where there can be other building blocks as well, brought together in a Rubik’s Cube to solve the problem of maximizing lift in response to communications. For example, it is entirely conceivable that an audience (the subset of users in a DMP to be targeted with a particular communication in a particular way) is constructed from:
- Attitudinal consumer segments
- Shopper attitude segments
- Frequent shopper data profiles
- Contextual factors (e.g. tagged as an intender, or it is raining outside)
- Geo factors (near a certain store)
Two other big distinctions.
- People go in and out of an audience because they have a contextual and temporal aspect, what Google might call moments, while segments tend to be durable classifications a consumer permanently falls into.
- Audiences should be enumerated by the marketer not the media property in a digital age.
For example, those in the right attitudinal and behavioral segments should come into a Starbucks retail store audience WHEN they are physically traveling in business or shopping areas (not home), irrespective of what app or mWebsite they happen to be using, in order to maximize response to an ad intended to drive footfall.
Audience creation in this way might sound complicated, but it is right in the adtech wheelhouse to do this. However, because DMPs are dumb containers of data, they need to be properly directed to create these audiences. That is where you (research and marketing team) come in. Done right, segmentation simultaneously solves for inspiration and for anticipating what it will take to enable high performing audiences to be constructed.
Marketers are becoming very interested in research on research experiments that have practical value by driving marketing ROI to unheard of levels. For this reason, testing how to simultaneously solve for segmentation and the creation of high performing audiences is now gaining significant traction.