I have had the honor of being the thought-leader and chief geek in residence of two ground-breaking white paper studies that prove how targeting done right in a digital age is superior as a guiding principle for media planning, as opposed to the idea of maximizing reach.
In fact, these two studies actually tie together quite nicely. The most recent study found that consumers in a brand’s Movable Middle, that is, the 15% or so of category buyers who have a 20-80% probability of buying your brand exhibit FIVE TIMES the ROAS (return on ad spending) versus non-buyers of your brand. The study from 2017 (The Persuadables, linked below) found that heavy and moderate brand buyers are much more responsive to advertising for your brand than non-buyers. The tie-in? Heavy brand buyers are many times more likely to be in your Movable Middle! (That’s right – your heavy buyers are not as loyal as you think and probably buy other brands more than they buy you!)
The white papers for these studies can be found at the following links (ungated, free download):
Why are targeting and reach-based media strategies for your advertising incompatible? Reach-based approaches assume that the “next ad impression” is best deployed against the person who has not yet seen the ad. Targeting says “bollocks” – yes, this Brooklyn boy has been watching a lot of Britbox and Ted Lasso – and serve the “next ad” to someone for whom we have a prior expectation of being more responsive.
And who is that responsive consumer? They are in the Movable Middle. We proved this via simulation but also through math. If you consider a logit function – which best describes the effect of factors that move a consumer between a 0% and 100% probability of choosing your brand – it turns out that the first derivative is maximized at a baseline probability of purchase of 50%. In other words, it is NOT the non-buyers who are most likely to respond to your advertising and add incremental sales, it is those who choose you 50% of the time! So, reach-based strategies offer a desperate attempt to get more unresponsive eyeballs to see your ad. Not good.
But what about bringing non-buyers into the brand? It turns out that targeting the Movable Middle does better on that metric as well. The reason is that you do not target a segment, you target an audience that has a high concentration of the segment you are looking for but still has misclassification. Those misclassified by your Lookalike Modeling are non-buyers who share defining characteristics and they will, in fact, respond more than the general pool of non-buyers.
I know this seems heretical to many of you, so I encourage you to prove it for yourselves. Design research to test this hypothesis. It could be a test of ad exposure post-analyzed with frequent shopper data. It could be based on your first-party data where you analyze lift within segments defined by heavy, moderate, light, and lapsed buyers. Then bridge over to the Movable Middle via surveys to demonstrate what I found – that Movable Middles tend to reside within your heavy and moderate buyers. As a survey protocol, I suggest you use a constant sum question to get at a respondent’s preferences for one brand versus another, and to tag someone as being in the Movable Middle.
Besides, heavy and moderate brand buyers, where else might we find a high concentration of Movable Middles? Lots of audiences! Social media fans and followers like a brand in real life before they like it on social media! Those who shop at a particular retailer. Those in higher brand development markets. Those who watch a certain TV show or cable channel. Really, there is a Movable Middle strategy that is possible for every marketing/media channel.
I know we all yearn for a simpler time, when planning for the most reach among women 18-49 was the guiding principle and TV ratings and MRI data were about all you needed, but as the former Chairman of GroupM said,”Marketing has to look a lot more like digital; it can’t go the other way!”
Credits: The Movable Middle research was done via my relationship with the MMA trade association, in partnership with Neustar, and supported by Numerator graciously donating their receipt scanning data. The Persuadables research was supported by Viant and NCS.
Header Image: Melanie Deziel, Unsplash