18 Apr 2019 It’s a Fine Line Between Simplifying and Stereotyping
Segmentation is a powerful weapon in the marketer’s arsenal but a poor one can be catastrophic. Avoiding oversimplification is crucial.
Simplification is at the core of any segmentation exercise because people, purchase occasions, consumption occasions, businesses, employees (or whatever it might be) aren’t all the same. They all respond differently to elements of the marketing mix. Treating people as if they were the same makes no commercial sense, and treating everyone differently is inefficient, if not impossible.
Where segmentations can often go wrong
Sometimes segmentations fail and there are many reasons why. Some are too all-encompassing. Others clash with senior stakeholders’ deeply held beliefs. Some are statistically fantastic but pull the market apart in all the wrong places.
Others don’t make the grade because research consultancies, client-side researchers and end users are used to seeing the market divided up into a small number of segments that cannot hope to offer any real clarity, targeting potential or sizing accuracy.
It’s almost trivial to say that the fewer segments you create, the muddier they are. But real problems kick in when time and money are spent on conveying the essence of a segment that only represents a stereotype. It’s clear this has happened when, with closer examination, most members of the segment don’t look how you might expect – a segment of ‘Innovation Seekers’, for example, where a third claims to be resistant to change.
Sometimes this happens because you’re trying to squeeze the market into five or six segments to avoid overwhelming the people who need to use them. This is an understandable strategy, but the consequence is that many individuals in any given segment don’t really reflect what it’s supposed to be about. This might not always matter if we’re just trying to paint a broad descriptive picture. But often that’s not what we’re doing.
How can we tread the line between simplifying and stereotyping?
Segmentation is powerful for identifying and sizing market opportunities, quantifying barriers to product usage, or measuring spend and loyalty.
All of these require quantitative data, and here’s the issue: if your brand share of Innovation Seekers is 30%, it’s easy to believe you’re doing a great job meeting the evolving needs of this group. But what if this is the 30% that shouldn’t be in that segment?
It’s critical to start with far more segments than you would ever circulate broadly in an organisation, let alone develop a strategy against. By creating 11 or 12 segments you can choose to entirely disregard a majority of them without stripping away huge swathes of the market. This means you can simplify effectively: focus on five or six segments and create collateral and strategies around these.
Then you can feel confident, knowing that these segments are clear. Everyone in a segment is actually similar and reflective of the description. And they’ll be more likely to respond as a unit to whatever activities you put in place.
We need a different way of rolling out segments
A careful roll-out plan is required. The right people need to be involved throughout to ensure the paring down process is timely and features the right segments. The model of consultancies signing off a segmentation with the insight department and then rolling it out into activation stage just doesn’t work here. The right people have to consider the activation potential together as we’re evaluating which segments to take forward.
These principles need a shift for everyone and at all phases of a segmentation project. Researchers need to find more granularity in the right places to support a successful segmentation. And clients need to be open to a different segmentation research journey to maximise impact and accuracy of decision-making.
At the end of the day, if you build a segmentation with only five or six segments, you may well get a useful, broad understanding of your market. But the amount of noise you’ll introduce to each segment means that you’re confusing segments for stereotypes.