When Building Is Cheap, Saying “Not Yet” Gets Expensive

There is a particular discomfort I have had to get used to this year: telling a fast, capable, eager team to wait, when everything in the tooling is begging them to start. Building is close to free now. A feature that used to be a week of commitment is an afternoon of curiosity. And the cheaper starting got, the more expensive it became to start the wrong thing.

For most of my career, the cost of building was its own safety mechanism. You did not chase every idea because you could not afford to. The sheer effort of implementation forced a rough triage before anyone wrote a line: is this worth a week? That friction was annoying, and it was also doing quiet work, filtering out the half-formed ideas before they consumed anything real.

Remove the friction and you remove the filter with it. When the effort no longer says no on your behalf, someone has to, and that someone is usually the person holding the whole thing together.

What I noticed first was how quickly the backlog stopped meaning anything. When starting is cheap, everything looks worth starting. Someone has an idea in a call and it is prototyped by the afternoon, and now there is a working thing that wants to be finished, staffed, maintained, and defended, and none of that was ever decided on purpose.

The build was free. Everything downstream of the build was not. A prototype is cheap to make and expensive to own, and the gap between those two facts is where a lot of quiet waste now lives. You did not spend a week building the wrong thing anymore. You spent an afternoon, and then three weeks living with it.

So the leverage moved. It used to sit in the middle of delivery, in execution, in keeping the build on track. Now the build mostly takes care of itself, and the leverage moved to the very front of the pipeline, to the decision about what should be allowed to begin at all. That is the least glamorous part of the job, and it turned out to be the part that matters most now.

Running several client engagements at once, the temptation compounds. Each one has an idea that could be tried today, and let us just prototype it is now the easiest sentence in the room. On one of them I let three things start in a single fast week, because there was clearly capacity and no obvious reason to say wait.

All three worked. Two of them we then spent a fortnight quietly unwinding, not because they were badly built but because they should never have been built, and a working feature is much harder to kill than a proposal. A proposal you can decline in a sentence. A working thing has momentum, a small constituency, and a sunk cost that everyone can see.

What made it expensive was not the building, which was cheap, but everything that attached itself to the two features once they existed: the client who had seen them and now expected them, the small integrations that had grown around them, the quiet assumption on the team that they were part of the plan. Killing them meant unpicking all of that, and none of it had been decided on purpose.

Cheap to make had turned into costly to remove, and I had been the one holding the gate open. The speed I had let run free at the start came back as a bill at the end, and it was a bigger bill than the speed had ever been worth.

So now I treat not yet as real work, not as a bottleneck. Being deliberate about what gets started, when starting the wrong thing is exactly as fast and cheap as starting the right one, is where a lot of my attention goes. It feels wrong, because saying no to something ready and easy runs against every instinct a delivery person has.

But a fast team aimed at a vague idea does not save you time. It just arrives at the mess faster, with more already built, and hands you a bigger thing to undo. Guarding what gets started earned its keep more than anything I did once the work was already moving.

It also changes what good prioritisation looks like. The old skill was choosing the highest-value item to build next, from a list that was expensive to add to. The new skill is deciding what does not get built at all, from a list anyone can add to in an afternoon. Choosing what not to build turned out to be harder than choosing what to build next, and it leaves nothing on the board to point at.

The hard part is that this looks like caution in a room that is excited to move, and caution is an unpopular thing to be when the tooling makes everything feel possible. It is worth being unpopular for, because the cost of the alternative arrives later and lands on the same person.

So I will ask the thing I have had to learn: now that building is cheap on your team, who is guarding what gets started, and is that person treated as doing real work, or as getting in the way?