Software is only useful in use

In every project, a moment comes when the roadmap starts to feel more important than delivery. A clean schedule, a wide feature set, a launch in six weeks. It feels comfortable.

It’s the opposite of comfortable. Software that isn’t in someone’s hands within weeks often never ends up in anyone’s hands at all. The world has moved, the budget is gone, the enthusiasm is spent, or the user who was meant to use it has long since moved on to something else.

Two weeks, not two months

I work in iterations of weeks, not months. Not out of principle-stubbornness, but because end-users only say something useful once they have something in front of them. As long as it’s on paper, you get politeness. The moment it works, you get the truth.

• The first delivery is usable, not complete.
• It gets refined based on real feedback.
• What no one uses gets cut.

Ten years of building has taught me: a feature no one has used has no claim to more hours either. Better to discover in two weeks that something doesn’t work than to find out after three months.

The feedback loop is the work

The moment something is in someone’s hands, the conversation changes. “That button doesn’t do what I thought” is worth more than a hundred workshop notes. The first real bug says more than two weeks of planning, because it forces you to look at what is actually in the way.

That’s why I’m willing to ship early, even when the piece is still raw. What’s live gets finished in steps. What lingers between design and handover, dies between design and handover.

Roadmaps are internal

Roadmaps are a handy instrument for me as the builder — not a promise to the end user. A roadmap the client broadcasts becomes a debt. A roadmap that sits quietly in my head stays a tool. Which of those two delivers you working software, you get to guess.

No milestones left on the shelf.

That is what being used means.

What that means for what clients get from me

No demo-only delivery. No screen that only behaves at handover. No feature that’s “live” but can be found nowhere. Small, early, in use — and only then build further on what you’ve actually learned.

The worst outcome isn’t a project that runs differently from the roadmap. The worst outcome is a project that ran exactly along the roadmap — but where no one can say with certainty whether it’s really in use anywhere.