An objective set where there is no constraint doesn't move the company, however well it is delivered

The theory of constraints isn't something you implement or skip. An organisation always has constraints, and the only choice is whether you manage them or they manage you. The simplest picture in this conversation is a hairdresser's. At first the constraint is one pair of hands, so we hire another person. Then the constraint is the premises, because no more chairs will fit in. If there aren't enough customers, the market becomes the constraint and the work moves to marketing. In the end what is left is opening a second salon. The constraint moves around, so the question of where it sits right now has to be asked afresh every so often.

And this is where it meets objectives. The theory of constraints tells you where to apply leverage, and it stops there. It has no operational layer: it doesn't tell you how to work day to day or how to cascade an objective down, and that is exactly what OKRs bring. The dependency is even sharper the other way. An objective placed where there is no constraint improves an area that already works well, so the system as a whole, meaning the rate at which the company makes money, gains little from it. That objective may be very well written, have measurable key results and be delivered on time. It is local optimisation carried out to perfection.

Hence the second axis of the episode, a plea for the right order: first define the problem, then look for a solution. The reflex goes the opposite way, because we are practically bombarded with ready-made frameworks and promises before anyone has named what is actually broken. The tool for diagnosis is the current reality tree: you write down the undesirable effects, connect them logically and work down to the root causes. The difference from the usual five whys lies in the rigour. You are not allowed to write it seems to me; every link between cause and effect has to be proven with an interview or a metric. A tree like that isn't drawn in an hour and isn't drawn with the board alone, because the people closest to the work usually know perfectly well where the problem is, it is just that rarely does anyone ask them. The same tools can also be used for something that doesn't always flatter us: the questions how will you know that objectives are helping your organisation, and how will you know they are not, are ready material for a current and a future reality tree.