- Posted by Eric Paquette
- On July 8, 2020
In more typical times, a market segmentation can provide the foundational marketing understanding to guide a brand’s strategy for several years. Done well, a market segmentation can inform targeting, positioning, messaging, experience, product, pricing, communications, loyalty and other key elements of brand strategy. Clients often ask how long a segmentation will last and I have always advised that in the absence of some sort of seismic disruption to the category or brand, they can be leveraged the next five years – give or take a couple of years depending on how dynamic the brand or the category is. Having developed hundreds of segmentations and looked at many more, that rule of thumb has held true (unless it was developed using foolish data constructs).
Unfortunately, I suspect that a global pandemic and months of stay at home orders just may qualify as a “seismic disruption” for many brands. In a number of categories, there have been significant shifts in the underlying attitudes, needs, emotions and behaviors that would be used to define a market segmentation. For many brands, that may mean it is worth revisiting, or at least testing the market segmentation to make sure it is still sound and provides an actionable framework for brand strategy.
But how do you know if the old segmentation structure still works or not? Fortunately, there are relatively easy ways to objectively pressure-test whether your old segmentation still works well.
If you have included your segmentation in your brand tracking, it should be possible to extract most of the necessary data from a recent wave. If not, or if you don’t have a recent wave of brand tracking data available, a relatively straightforward custom survey can be used to gather the needed information.
You should answer the following four questions about your segmentation to determine if it is still sound.
1) Have the segment sizes changed dramatically?
Segment sizes typically do not change very quickly. It is unusual to see movements of more than a couple of percentage points year over year, and in many instances, shifts are not even that large. If you find that there have been large shifts in segment sizes, that is one clue that the segmentation may need to be revisited.
It is possible that the segment sizes can change and the segmentation is still solid. If the key underlying attitudes, needs and behaviors that define your segmentation are unchanged, and simply different numbers of people hold those attitudes, needs and behaviors, the segmentation itself may be fine. You would also want to ensure that your target segments haven’t shrunk to the point that they are no longer sizable enough to support the brand.
But more often than not, when you see big movements in segment size, it is an indication that the underlying foundation of the segmentation has shifted.
2) Are the segments still different in terms of their value to your brand?
A good segmentation will yield segments that are different in terms of what they are worth to your brand. For a typical consumer brand, a primary target segment ought to be at least a 150-200 index in terms of their current or potential value to the brand. Another indication that a segmentation is out of date is erosion in the value metrics for your target segments or increasing similarity in value of each segment. When this happens, it is an indication that the underlying attitudes, needs and behaviors for your brand and category have changed and a new segmentation that reflects those changes is needed.
3) Are the members of each segment still highly similar?
Most segmentations are developed using some measures of similarity. When developing a segmentation, the idea is to develop segments in such a way that all the members of a particular segment are as similar to one another as possible, while different segments themselves are as different as possible from one another.
When monitoring segment membership over time, most will use some sort of classification routine like a discriminant function analysis or choice model to predict segment membership based on a subset of the original segment questions. Most of those classification routines assign a new respondent to the segment that represents the best fit, but they do not typically take into account whether the “best “fit is actually a good fit.
The plot below gives an illustration of what I mean. On the left, in the original segmentation, you see three tight, homogenous clusters on key segmentation attributes. On the right, you see predicted segments when the underlying market dynamics have shifted making the segments more dissimilar and blurry.
Monitoring Segment Homogeneity
It is possible to calculate measures of similarity for the segments to see whether they have become noticeably weaker. If so, it is another strong indication that a segmentation has outlived its usefulness.
4) Are new attitudes, needs or behaviors now related to measures of value for your brand?
Finally, we can look to see whether there are new attitudes, need, psychographics, behaviors, etc. that have taken on elevated importance and / or are highly related to value for your brand that now should be included in the definition of your segments. This involves adding a new dimension or set of dimensions to the segmentation analysis that you are using to actually develop the segmentation. Depending on the number and the nature of the added dimensions, this could cause minor or more significant shifts in the segment definitions.
Think about today’s COVID-19 environment. It would not be at all surprising that a segment of consumers is paying more attention to how companies are treating their employees than they were before given both the health risks faced by some employees and the economic environment that many face. If these measures were not included in defining the segmentation originally, it may be necessary to develop a new segmentation that incorporates them, assuming they new needs and attitudes are not believed to be temporary.
This last dynamic is actually one of the trickier ones to interpret and navigate in an environment like today. You would not want to develop a new segmentation based on a set of attitudes and needs that are going to be highly temporary. While some of the changes that we’ve all faced over the last several months have caused some temporary changes, there have almost certainly been others that will be more permanent, at least for a portion of the market.
About Brookmark Research:
Brookmark is a research operations company focused on managing the critical components of the research process including project management, survey design, programming, hosting, sample design & management, data management and reporting solutions. Brookmark provides a full outsource solution for all your research needs.