The Movable Middles of your business are the most important group of category buyers you may not be aware of. According to arithmetic and in-market evidence, they create FIVE TIMES the ROAS (return on ad investment) of all other category customers.
If a consumer’s chance of choosing your brand is between 20% and 80%, they are in the Movable Middle. According to research conducted by the MMA and Neustar, allocating money to targetable audiences with a high proportion of Movable Middles increases ROAS by 50% and improves non-buyer conversion by 13%.
People that don’t regularly respond to your adverts are known as low unresponsive. However, since that is where the great majority of your brand’s non-buyers’ dwell, a conquesting campaign with the right content may be effective.
As illustrated in Figure 1, the number of these groups is expected to fluctuate over time, as does their relative susceptibility to advertising. The size of the ball represents the proportion of category consumers in that brand preference group. The “inclination of the hill” indicates how simple or difficult it is to convince people to respond to your campaign.
This provides a solid and simple foundation for the marketer’s conundrum: how can I successfully execute and mix performance marketing with brand marketing? Consider this: performance marketing is all about reminding clients of what they already know about the firm by directing ad impressions to the Movable Middle for the maximum sales bump. The purpose of brand marketing is to move low unresponsive to movable middles. People must think or feel differently about your brand as a consequence of your advertising in order to complete this challenge.
What do you want “converts” to understand or experience? That is the subject of this paper. What distinct ideas are needed to turn non-buyers into “Movable Middle” customers?
So, what do we know about the Movable Middle right now? What about the low amount of people that don’t respond? What about your most loyal clients? How can we develop predictive consumer categorization models? This may be helped by discriminant analysis from brand monitoring or brand equity research.
This technique is particularly beneficial for mapping out a path for customers to take from Low Unresponsives to Movable Middle and, ideally, High Loyals.
The methodology is known as Comparative Discriminant Analysis, and this article will lead you through it using the fictional Coca-Cola brand iced drink Bok Tea as a case study. Although the iced tea business is large and lucrative, the brand is suffering. The brand team should want to move clients from Low Loyals to Movable Middles in order to achieve long-term growth (and ultimately High Loyals).
Basics of Discriminant Analysis
Discriminant analysis is a kind of statistical method in which the groups or dependent variables are determined ahead of time (what data scientists refer to as “supervised learning”). We prefer to employ constant sum questions across a consideration set, despite the fact that other survey approaches may work. The goal is to provide a chance of purchase estimate for your brand for each survey respondent who is offered a category buy. As a result, the categories are the first need. To begin, you must first define the groupings. Multiple discriminant analysis is then used to find the discriminating characteristics, from which discriminant maps are constructed.
The fundamental properties of the grouping variable are as follows. They are all-inclusive, mutually exclusive, and distinct from one another. We’re looking for membership groups within each level of the loyal pyramid in the case of Bok Tea.
Coca-ability Cola’s move to the Movable Middle of Bok Tea will provide a two-fold benefit: higher market share and advertising responsiveness/profitability. When advertising becomes practically self-sustaining, more money is spent on it, which leads to even more growth… a vicious cycle of brand expansion!
Construction of the Target Variables
The dependent variables, or goal variables, are constructed in this manner. For the first discriminant level, we start with our sample. To do a Movable Middle analysis, all respondents in the survey must buy the category. To conduct the study, we enlisted the help of 10,000 category buyers. As we go up the pyramid, the study’s sample size shrinks as we reach customers at the next level—or higher.
As we can see, as we go further up the pyramid, the base sizes become smaller. On the first level, those who are Aware but Bok Tea Low Unresponsive get a ‘0’. Those who are Bok Tea Movable Middle or High Locals get a ‘1’.
Creation of the Dependent Variables—The Road Map
Comparative discriminant analysis’ strength comes from its ability to use numerous parts of the survey to identify what pushes purchasers to go up the brand pyramid. And at each stage, we’re essentially doing new research. Not one, but four sets of instructions are shown on this map (in our example).
It’s unusual to include six or seven different sections of the survey in a roadmap analysis. However, for the purpose of example, we’ll limit our Bok Tea research to four different brand combinations.
We’ll look at the following survey elements to see how to go up the brand pyramid for Bok Tea.
Primary reasons for purchasing iced tea (brand non-specific)
Favorite Bok Tea flavors
Rating the Bok Tea brand
Negative statements about Bok Tea—brand barriers
On their quarterly tracker, Bok Tea asks respondents 21 primary reasons why they purchase any brand of iced tea. Bok Tea is trying out 30 distinct flavors. Following that, respondents are asked to rate Bok Tea on 15 different allegations about the tea. Finally, respondents are asked whether they expect any of 14 negative Bok Tea encounters.
The graph below shows how for each item, performance rating, and barrier, we’re creating a new variable. The coding for these variables is 1/0. The item variable gets a 1 if someone ordered anything, but a 0 if they didn’t. If a respondent agrees with a statement about Bok Tea, the variable earns a 1, otherwise a 0. If there is a barrier, give it a 0; otherwise, give it a 1. Below is a list of the variables that were produced.
In the figure above, there are four different kinds. As a consequence, we undertake four discriminant analyses for each level of the brand pyramid. That is, a total of 12 analyses are being performed.
Interpreting the Analysis
This method is known as comparative discriminant analysis since it uses the same analysis for all levels of the brand pyramid. The only variable that fluctuates is the dependent variable. In this way, we may identify which items, performances, and barriers drive the reaction from one level to the next. In other words, we’re sketching up a plan to extend the Bok Tea brand.
The findings for each step of the study would take up much too much room to display. Below is an example of the statistics output from discriminant analysis.
The highlights are the Bok Tea Attributes that excel at discriminating between the Low Unresponsives and the Movable Middle.
According to a simple interpretation of these graphs, the top four features distinguish the most between customers who are Movable Middle of Bok Tea and those who are Low Unresponsive.
Summarizing the Brand Path
The study’s last step is to compile a list of the discriminators used in the four survey areas’ comparative discriminant analyses. Below are the differentiators between Movable Middle and Low Unresponsives.
In order to guide clients to the next level of the Bok Tea brand pyramid, we may identify iced tea purchase reasons, favorite flavors, Bok Tea attributes, and possible hurdles from this summary graphic. Remember that this study comes with four summary tables, one for each path in the brand pyramid.
Comparative discriminant analysis is one of several effective, adaptive methods that have existed since the beginning of mathematics and are still important in today’s branding strategy methodology. When applied to the Movable Middle, you now have a solid research that can guide advertising communication and innovation initiatives that can deliver short-term performance and long-term brand growth.