Notes

When I look at a process I write down the first five steps. Not all of them, just five. Anything past five is a guess, and a written guess starts to look like a plan.

The habit has one purpose: to make the blockage discussable. If three people describe step three differently, that is where the problem is — not in the tooling.

What usually surfaces is this: nobody sees the work end to end. Everyone knows their own part well, and nobody knows the gaps between them.

A report's first job is not to measure, it is to tell. If a number does not become a sentence in someone's head, it is not working yet.

So before building a table I ask which sentence I am trying to write. When the sentence is clear the table gets shorter. When it is vague the table grows and nobody opens it.

In production this is sharper: how many days a piece takes, how often it comes back, and where it waits are three sentences. The rest is detail.

Fixing a mould and fixing a screen ask the same question: is this step actually necessary, or is it just habit?

At the bench an unnecessary operation is felt in the hand. In software it is not felt; it accumulates as one more field, one more approval, one more tab, and nobody notices.

Jewelry taught me this: a wasted step shows up as fatigue before it shows up as cost. If you can see the fatigue, you can fix the process.

Installing a new tool is easy; changing how work flows is hard. Swap the tool and the same work restarts somewhere new, with the same mess.

So when I start something I pick one step. Most efforts that try to change everything at once get quietly abandoned, because they made nobody's day easier.

If one step works, nobody has to be pushed into the second. If they have to be pushed, the first step did not work.