Evidence that reach-based media planning is like a bad slot machine

“For decades, marketers have been operating under the assumption they need two audience strategies: One to fulfill short-term goals (hyper targeted) and another to cultivate long-term performance (broad reach). But new research from TransUnion and MMA Global has proven this to be a myth.” Emiko Seale, Sr. Director, The Knowledge Lab, Marketing Solutions, TransUnion

I was delighted to contribute to this analysis, as part of the MMA team.

(I urge you to download the whitepaper: https://gag.gl/XQTw4m to learn more about building an audience strategy with Movable Middles at the foundation.)

First, the findings from Ally study. TU continued to analyze the Ally “brand as performance” data with a new question in mind. Which collection of their 16 TruAudience rollups would be the top priority if following a Movable Middle based audience selection strategy? (This was based on on-boarded Dynata survey data where we flagged survey participants as Movable Middles.) With separate continual feeds of ad serving and conversions mapped back at an ID level, TU determined if those Movable Middle-rich audiences actually outperformed other audience collections that would be chosen by following other strategies.

Results proved that Movable Middle audiences delivered more short term incrementality AND brought more new customers to the bank over 18 months after the campaign ended. The table below shows the degree to which Movable Middle audience selection outperformed other strategies (for full list of alternative strategies tested, see white paper).

Why targeting Movable Middle audiences leads to more new customers. There are two explanations as to why this result actually should not be a surprise. Here I will share the intuitive one but for those interested in the math, please see my companion blog here.

The intuitive reason…

First understand that you direct ad impressions to audiences, not segments. The Movable Middle is a segment that exists at different concentration levels within audiences you can choose from. So, let’s say that, on average, 20% of consumers are Movable Middles for the brand of interest. Now, let’s say that 40% are Movable Middles within a particular audience (e.g. a list that was pushed for activation.) While this would be a great audience to direct disproportionately more media towards, it also has 60% of IDs who are NOT Movable Middles…in effect, it is as if they are misclassified. Now, misclassification occurs because they share important profiling characteristics of Movable Middles…that makes them lookalikes!

Hence, when you target audiences rich in Movable Middles you also reach non-customers who are lookalikes for Movable Middles (mostly customers). The ad campaign directed at such audiences is the nudge that some of those non-customers need to become customers.

Summary. TU has written a white paper based on an Ally bank case study where audiences rich in Movable Middles delivered more short-term sales PLUS more new customer acquisition over time. I have given the explanation as to why this result was actually to be expected and why Movable Middle audience selection would lead to growth, repeatably.

On the other hand, staying with the myth that it is best to maximize reach with your media plan for brand campaigns is like playing the slot machine that never pays off bad in the short-term, worse in the long-term. Heresy? With the possibility of doubling your advertising return, isn’t it worth an experiment to test this on your brand? Or, as Greg Stuart, the CEO of the MMA asked at the Possible conference in Miami in front of thousands, “Are you too busy to do better?”