We continue to believe that done well, market segmentation is not just a critical marketing tool, but can be the defining element of your overall strategy. In our previous blog post, we shared of the most common difficulties in getting to a workable needs-based segmentation.  So what can you do to avoid those mistakes and unleash the power of an effective segmentation?  There is no easy answer, but here a few things to keep in mind as you work as you work through this critical task:

• Segment at the right level – Insightful segmentations are generally at the lowest level of the value chain that you can make actionable.  If you are selling automotive parts to car manufacturers, for example, you might get some insight from segmenting car purchasers, but it would be difficult to act on this directly.  It is certainly more actionable to segment automotive OEMs, since they are your direct customers; but even that may not be sufficient – BMW likely has different needs for a diesel sedan built in Germany than for an electric vehicle built in the US – so application or platform might be the right way to segment.
• Segment with an eye towards your differentiation – remember, winning strategies are found in your sweet-spots, those intersections of under-met customer needs with things you do uniquely well.  So the right segmentation for you is not the right segmentation for your competitors.  Starting with an understanding of your differentiation and then exploring why some customers value it a lot and others very little is a great way to start…
• … Building on that, fewer needs are generally better – segmentation works best when you can zero in on the couple of customer needs that really determine whether or not you can win in that segment.  Starting with a long list of universal needs can add unnecessary noise to the process.  Worse yet, there is what we’ve come to call the 2^N problem.  Stated simply, if you have N different needs and a given customer can be high or low on that need, then you have 2^N potential segments.  No matter how theoretically pure this approach feels, if N is anything greater than three, you almost always end up with an unworkable number of segments.  The sooner you can agree on the two to three needs that really matter, the smoother the process will be.

• Use ‘why?’ questions to match customer characteristics to needs – Don’t confuse customer characteristics with needs.  Only a distinct set of needs defines a segment.  But segmentation works best when we can identify a link between characteristics and needs.  The key usually starts with ‘why?’ questions.  For example:

• Q: Why do some customers value technical support more than others?
• A: Because some customers have experienced internal engineering groups, and some don’t (reason behind the need)
• Q: Which customers do not have experienced internal engineering?
• A: Usually, those are start-ups with fewer than 200 employees (company characteristic)
• Q: Why don’t start-ups have internal engineering?
• A: Usually they have been focused on driving the revenue side of their business through design and have outsourced value engineering (the logic that links the need to the characteristic)

• Accept the imperfect – good segmentations do not need to be exact.  In fact, it is okay to have an ‘all other’ segment for that last 10 or 15 percent of the market that doesn’t fall neatly into a category.  The key to making segmentation actionable is basing it on information that you can know before you go on a sales call.  That is the only way to proactively tailor your communications, your offering and potentially your price.  You may guess wrong sometimes as to which segment a specific customer falls into, but as long as you collect the feedback you can adjust as you go.  A segmentation need not be perfect to be useful, as long as you recognize that segmentation is iterative, and building the feedback mechanisms to adapt over time.

• Focus on the implications – Sometimes when teams really struggle with how to segment, we flip the problem on its head and start with implications – what are you willing to do differently across segments? Change service levels, develop custom products, charge different prices?  If you can answer this question clearly, sometimes you can get to a segmentation indirectly.  For example, by asking why you would be willing to provide a higher service level for some customers than others.  This type of ‘reverse engineering’ can often shed light on the customer characteristics and needs that should be defining our segmentation approach.  If nothing else, it forces your team to realize that segmentation is not just a mental exercise for internal reporting.

All of these recommendations also highlight the importance of involving a cross-functional team – consistent with how we think about strategy, there is no monopoly on good ideas.  So involving more functions and perspectives in the segmentation process is almost always a good idea.  At a minimum, it will prevent groupthink, and help avoid creating a segmentation that looks good on paper but cannot be implemented. This tends to be true because involving people outside the marketing department will force the, sometimes awkward, discussion on implications – what will actually be different if we implement this segmentation and are we comfortable doing that?

As a last piece of advice, and perhaps the most important of all, don’t give up.  Segmentation is hard and takes concerted, cross-functional effort that does not come natural to many companies, especially those who have historically been driven by sales and/or technical excellence and may not understand or value strategic marketing.  But taking the time to do segmentation properly is almost always worth it.  For those who get it right it can be the centerpiece of their strategy and position them for ongoing success.