Over the past few weeks I have had occasion to work with a few established organizations and some startups. I know by now it should not surprise me, but the engagements almost always have to start with a discussion about customers and segmenting. It is usually something with which the clients are struggling.

There are two behaviours that I believe cause people a lot of difficulty in thinking about customers. There is a tendency to look at customers by thinking of anyone with whom you are, or you want to have interaction; even if that relationship is not a customer one but rather a partner. The second common approach is creating segments based on characteristics (age, technology expertise, location etc.).

I subscribe to the job-to-be-done approach to thinking about customers. This is the Harvard Business School approach championed by Clayton Christensen and Mark Johnson, and used widely by Alexander Osterwalder and members of this community. It is built on the concept that the job your customer has to get done, not the customer, is fundamental unit of analysis. This concept can be extended to include the problem-to-be-solved. I won’t go into lengthy, detail explaining the approach here, my objective is to share how I implement that point of view. It is worth doing more research if you are not familiar with "job-to-be-done".

When we start the discussion I ask the people to create a three-column table. In the first column I have them brainstorm anyone who might be interested or be helped by the vision, product or service the business owners have in mind. This first exercise should be as creative as possible, don’t let them self-edit. It will work out as we go along.

After the first column is populated, I ask the participants to then, in the second column, identify the job, or the problem, that each of the potential customers in column one have, with which the participants could help them. This last point “with which the participants could help” is very important. It is not just any problem or job.

Once the second column is populated I ask the participants to then list, in column three, what the customer needs from you in order to get that job done, or solve that problem.

We then look down each of columns two and three, and try to find commonalities that allow us to group these ideas together. For column two, the groupings of common jobs-to-be-done, or problems-to-solve, become the “customer segments” of the business model. The groupings of column three, commonalities about needs, become the “value propositions” of the business model.

On the first issue (customer versus partner), I have found a very simple question helps clarify thinking in this area – who is helping whom here. If you are helping someone else to get a job done, or solve a problem, then they are customers. If the other party is helping you to do something, they are partners not customers. That is not to say the two are mutually exclusive. It is quite conceivable the party may be both customer and partner. In that case you may be involved with a multi-sided platform.

It is a simple exercise that doesn’t take overly long to do. It requires no special tools or knowledge. It can be done on flip charts, or whiteboards, with markers or with pens and sticky notes. Then the findings are transferred to the VP and Customer components of the BM canvas.