Operating
Why we stopped billing hours
A timesheet rewards the wrong thing. Moving to outcome-based pricing changed what our engineers optimised for, and made the incentive honest.
The hourly model has a quiet problem that everyone in the industry knows and nobody says at the pitch: the slower the work goes, the more the supplier earns.
Nobody sets out to exploit that. But incentives do not need intent to shape behaviour. Over a year, a team billing hours will make a thousand small decisions: whether to automate a mechanical task, whether to push back on a scope creep, whether to write the tooling that makes the next engagement faster. The pull is always in one direction.
What we do instead
We scope a cycle: a defined set of features with acceptance criteria the client has agreed to in writing. We quote a fixed monthly subscription for the squad delivering it. If the work takes longer than we estimated, that is our cost to absorb.
Three things changed once we made that switch.
Automation stopped being a luxury. When speed is our margin, the internal tooling that saves two days per cycle pays for itself immediately. Under hourly billing that tooling was always next quarter’s project.
Scope conversations became honest and early. Under hourly, scope creep is revenue. Under fixed output, it is a cost, so we surface it in week two rather than discovering it in week five.
Estimation got better fast. When you eat your own estimation errors, you learn quickly. Our variance on cycle scoping is a fraction of what it was.
The objection
The obvious pushback: does fixed pricing incentivise cutting corners? It would, if the only thing measured were delivery. So the acceptance criteria are written before the cycle starts and reviewed by the client at the end, and the code goes through the same review and CI gates regardless of schedule pressure.
The remaining protection is structural. We want the next cycle. A squad that ships something fragile to hit a date has traded a year of work for six weeks of it, and everyone on the team knows the arithmetic.
Where it does not fit
Outcome pricing needs a scopeable outcome. Genuine research, where the answer might be “this does not work”, does not fit. We run that as a fixed-fee exploration with an explicit possibility of a negative result. Trying to force it into a delivery subscription would just push the risk somewhere it does not belong.