A Build – Measure – Learn cycle is applied widely in circles where agile, iterative and incremental approaches are used. It is central to Lean Start-Up and also when prototyping around a business model.

However, to start with the word ‘build’ risks addressing the wrong problem. Before we build we should decide what we want to ‘learn’, and even before this we must take account of our context and measure the contours and constraints of our environment. In almost all circumstances, I would propose a Measure – Learn – Build cycle.

When installing a new product, preparing for the provision of a service, initiating a transition or transformation, or any other kind of project, we need to know where we are; the as-is situation, the to-be situation, and then we can imagine the paths which would lead from start to finish. This is all about the starting situation.

Next step, before selecting which path to move along is for us to learn just a little bit about each path, and only as much as we need. To make too many steps would be costly and time-consuming, before we know the right way to go; and therefore we can progress only as far as we need orientation to make a decision. It really is about minimum effort for maximum feedback (mini effort, max feedback).

It’s like scouting; make sure you can find the way back before making a definitive decision, don’t give the game away too early and don’t get lost in the detail. Ensure also, that you are making plentiful use of user perspectives. One person’s promising path is another person’s penury or purgatory. Then when you are sure of your way, explore a small part of the road. Be prepared to build each brick, bridge, bump and boundary in order to oust obstacles and bolster the benefits.

I believe that this cycle of measure – learn – build cycle is missing an important step. A more complete cycle would be measure – learn – plan – build. Now, I can hear replies of “Not more steps”. “We already have enough steps” “And now there are three steps before we build”. Later on, I can explain more about the number of steps, but first I want to justify the ‘plan’ step.

This ‘plan’ step is more than it seems. Learn how to plan in your project. It would seem as if we are going to plan the actions that will be executed during ‘build’, which is entirely true; and so what? In fact, the real importance of the ‘plan’ step and the reason to keep it in the cycle is its vital role in the dynamics of teamwork. If you don’t plan together, you won’t build much together.

To prepare a plan is to develop and to agree upon a work breakdown structure, to define the responsibility matrix, to work as a team to elaborate a critical path network, master plan with milestones, or to participate in the planning game.

The plan includes taking account of the threats and opportunities that may occur during build so that the team is all the more prepared to make the most of them, to persist, to pivot, or to prevent the iteration from going any further.

We all understand that these cycles are iterations built incrementally or to obtain the maximum possible feedback to enable decision making. We could also say that because it’s a cycle, it doesn’t matter where we start; measure – learn – plan – build, or learn – plan – build – measure. But, plan – build – measure – learn does matter, because it reminds us of the quality wheel from Deming and Shewhart; plan – do – check – act.

Wait a moment, plan – do – check – act is a production cycle; first plan the work, work the plan, check at the end, learn how to do better next time. A problem solving cycle is more like check – think – plan – do; and so is a project cycle. Check going in and check going out in fact. Avoid the habit of doing the wrong things in the first iterations; because there will be too many of those false first steps. Make sure that you do measured learning. Maybe you can get a 20% or a 30% improvement. “If everything could be done twice everything could be done better.”

There are compelling arguments for starting with the end in mind; a mindset of first ‘check’ or ‘measure’, second think and learn, third plan or explore third and fourth build or do. It’s a project cycle, it corresponds to learning and change, and also it matches the team growth cycle: form – storm – norm – perform.

Let’s see how that works in Reinventing the Wheel, Cont.