Marketing and Research Consulting for a Brave New World
Subscribe via RSS

Throughout my career…notably Unilever, The NPD Group, Vivaldi Group, Chief Research Officer at the ARF and now in my own consulting practice since 2010, my studies of existing brands find a consistent central idea.

Customer churn occurs primarily among those consumers whose attitudes are not consistent with their behavior.  It is why you lose some of your customers and why you gain from your competitors.

The first time I reported on this “brand favorability balance sheet” and showed it was predictive of brand trends was in 1996 in my often-cited paper in the Journal of Advertising Research. We analyzed 27 brands from 5 product categories among 4,071 survey respondents. We surveyed the same respondents twice…one year apart…about the same brands and categories. On average, we found that 10% of non-brand buyers for a given brand DID have favorable attitudes towards the brand (based on modeling attribute agreement). On the second wave, one year later, we found that favorable non-buyers were EIGHT TIMES MORE LIKELY TO CONVERT vs. non-favorables.  For this reason, we called them “Prime Prospects.” Conversely, we found that the behaviorally loyal buyers who defected were disproportionately those who did not have commensurately strong feelings towards that brand (for this reason we called them “Vulnerables”). So, brand attitudes were predictive of consumer behavior one year later.

In research I am involved with now via the MMA…26 years later…I am getting consistent results regarding the percent of non-customers who are favorable to the brand of interest. However, this time around we are learning more about ad responsiveness.  It turns out for the study that was just completed, the responsiveness to advertising came from Prime Prospects. However, the vast majority of consumers are non-customers who are not prime prospects.  This is a proof point for the contention in the white paper on Outcomes based marketing that broad reach strategies are a recipe for wasting a large portion of your marketing budget.

This study also provided strong confirmation of the Movable Middle theory.  For this study, Movable Middles (those with a 20-80% probability of choosing your brand) were expected to be FIVE TIMES more responsive to advertising VS. NON-Movable Middles and the results actually surpassed that (multiplier calculated as the absolute lift in conversions for Movable Middles divided by absolute lift among non-Movable Middles, expressed per capita).  

So how does a marketer act on this?

First, think of your consumer franchise as customers plus prime prospects and agree that you want to allocate advertising impressions disproportionately more to your consumer franchise.

Next, understand the difference between segments and audiences.  Segments define the type of consumer you are looking for while audiences are collections of consumers you can reach at scale with advertising. No audience is 100% the segment you are looking for but you can select audiences with high concentrations of that segment.  A concentration level that is 50% higher or more vs. the national average is an attractive audience. Marketers have a variety of ways to find these attractive audiences. Marketers can on-board a seed sample and build an audience at scale via lookalike modeling, they can find those linear TV shows or networks that deliver attractive audiences (e.g. identified by onboarding segments to smart TV providers), or they could use their own first party list of customers pushed for activation (they always have a high percent that are Movable Middles). For CPG, Movable Middle audiences can be identified at scale from frequent shopper data and pushed for activation. There is an activation strategy for virtually every marketing channel.

Let’s say you are solving for performance and want to target Movable Middles who are 15% of the consumer population nationally but that you can create (or find) audiences that deliver 25% Movable Middles.  It doesn’t sound like much but at the ROAS multipliers we are seeing, this alone could be expected to improve the ROI of your advertising by 50%. Across all brands studied so far, I have seen multipliers as low as 2 (still good) and as high as 50+. What is your brand’s multiplier?  How can you intelligently plan media if you don’t know this?

If you are solving for conquest, you can go through this same process looking for Prime Prospects, i.e. favorable non-customers. My instincts tell me you might want to differentiate your marketing and lean in more with TV and video advertising to this audience.

Some will push back and ask, “How do we build the brand and “fill the favorability bucket” if we don’t advertise to unfavorable non-buyers?  Don’t worry…by targeting audiences with high concentrations of responsive IDs, there are plenty of non-favorables who will also receive media too because no audience is perfect; in a sense, every audience has a significant rate of misclassification. You’re just shifting the odds in your favor.

If the marketing team leverages this in their media plans, this can lead the CFO to view paid media as a profit generator and not just a cost of doing business. And that means advertising might be the last thing to be cut out of the budget…not the first…in uncertain economic times.

Tags: , , , , ,


3 Responses to “Over 26 years, evidence for targeting your advertising impressions”

  1. As always – Great stuff here Joel!

  2. Michael Donahue

    It is critically important and definitely possible to identify a brand’s most profitable users among your Movable
    I know how to do that because I am now working with MASB to help marketers improve brand financial value.