Segmentation is an ART, with a little SCIENCE mixed in. It is rarely the "answer", but when used appropriately, it can change the way you market and manage your business. If you look at it as a journey rather than a project, you are much more likely to be pleased with the outcome.
I recently took this segmentation journey alongside one of my key B2B clients. It was a three year adventure with a few peaks and valleys, but through the process we learned a lot about the business and ways to effectively move it forward.
Some of the lessons we learned along the way:
- No detours. We established a roadmap early in the process; with buy-in from all stakeholders, including senior management. Our roadmap looked something like this…
It was easy to communicate, follow and execute. We had bumps along the way, but were always able to stay on course because we had prepared and sold the roadmap early in the process.
- Understand what you know – and what you don’t know. You cannot underestimate the value of understanding business history when starting the segmentation journey. In this case, the business had at least five flavors of segmentation that had been applied over the last ten years. None of these flavors had stuck. Why had they failed? Was it an institutional problem or a methodology problem? We needed to understand the reasons "why" before we started the process; otherwise, we would be doomed to the same outcome.
- Take the time to define your hypotheses. I know, I know, I know – who wants to sit around brainstorming hypotheses? For segmentation, it is critical. If you do not have a sense of what the segments "should" be – you are not ready for a segmentation journey. Time spent upfront to brainstorm and define the expected outcomes will ensure your success at the end.
- Allow yourself to be surprised. There are many different scientific approaches to segmenting. Try them all. Be open to the possibility that your hypotheses could be flawed. Ours were. We learned through the process that one of our foundational assumptions was flawed; and in fact, an alternative set of variables, which we had paid much less attention to, was actually more effective at identifying and defining the segments.
- Do not settle. Call it gut or intuition (or magic) – but you will know the "right" segmentation when it emerges from the data. We did. After weeks of running and re-running the variables, when the final framework surfaced from the data, our whole team knew that was it. How? After having immersed ourselves in the business and the data for months, we knew that the essence of what we were segmenting had been captured in the solution. We knew it was implement-able. We also knew that it was re-defining and potentially game-changing.
- Do not let implementation issues destroy it. Implementation of segmentation is hard work. You must be practical. Understand the persona of the segments that have emerged as well as what distinguishes them. Be practical with how you score your customer database – it will not be perfect. For example, one of our defining variables was based on the nature of processes customers followed. There was not an exact metric in the customer database that could measure that specific variable. However, there were a number of variables in the database that were close, and in combination, could be used as a surrogate. We used the surrogate – and you know what, it got us close enough. The pursuit of perfection with segmentation is a pathway straight to failure. Compromise when needed; find a way to make it work.
Segmentation is oftentimes more ART than SCIENCE. The point is… Does it work? Does it move your business forward? Is it implementable? Consider KS&R the next time you decide to take this journey; you may be surprised what you discover along the way.