The compounding of good

There is a particular kind of work that disappears the moment it succeeds. You build the thing, it starts running, and within a month nobody remembers it was ever built — least of all the people it now quietly serves. This is usually treated as a failure of recognition. I have come to think of it as the highest form of success a system can reach.

The social sector runs on heroics. A coordinator stays late to fix the rota by hand. A caseworker memorises the workaround for the form that never quite fits. None of it scales, all of it depends on the particular person doing it, and every one of those people is exhausted. Heroics are what you get when the system underneath is missing, and a heroic culture is just a slow way of burning through the most committed people you have.

Subtraction comes first

The instinct, when you arrive, is to add. A new tool, a new dashboard, a new process with your name on it. The discipline is to do the opposite — to walk the work as it actually happens and ask, of each step, whether it needs to exist at all. Most improvement is removal. The duplicate form, the report nobody reads, the approval that approves nothing. Take those away and the work gets lighter before you have built a single new thing.

A copy repeats the wording. A system returns with shape.

What is left after subtraction is small enough to understand, and small enough to hand over. That last part matters more than anything that came before it. A system you cannot give away is not a system; it is a dependency wearing a system's clothes. The test is simple and slightly humbling: if you vanished tomorrow, would the thing keep running? If the answer is no, you have not finished.

The compounding part

Here is the quiet mathematics of it. A good system, properly handed over, does not pay out once. It pays out every week, to people who never met you, for as long as it runs. The coordinator who inherits the rota inherits the thinking baked into it. The next caseworker starts from the cleaned-up process rather than the mess. Each of them adds a little, and passes a slightly better version on again.

This is why we give so much of it away. Not out of charity — out of arithmetic. A tool released into the open is taught by everyone who touches it and returned to everyone who needs it. The good compounds precisely because no one owns the compounding. You give up the credit, and in exchange the work outlives you.

None of this is romantic. It is mostly spreadsheets and patient documentation and resisting the urge to be needed. But it adds up to something I have not found a better word for than good — the kind that keeps arriving long after the person who set it in motion has forgotten they did.

So we measure ourselves by an odd metric: how quickly a client stops needing us. The faster that happens, the better the work was. It is a strange way to run a practice, and the only one I have found that I can defend on a bad day.