Fixed project or long-term team? One question decides
It is not about budget or trust. It comes down to whether the goal will still be the same in three months — and picking the wrong model costs months of work.
Every first conversation reaches the same fork: a fixed project with a defined end, or a long-term engagement where our engineers work inside your organization. The deciding question is not budget or trust — it is whether the goal will still be the same goal in three months.
If it will, take the fixed project. If it will not, go long-term. Everything else follows from that, because the two models run differently day to day.
A fixed project
The scope and the goal are written down before we start, and the estimate and the date follow from them.
- Scope agreed up front; changing it means a new estimate and a new date, not a message on Slack
- Priorities set once, at the start
- Fixed deliverable, fixed end date, documentation and a handover
- We leave when it ships — the system is yours to run
- The right choice when the deadline matters more than the roadmap
A long-term engagement
We work inside your process rather than alongside it, and the roadmap is allowed to move.
- Priorities reset every sprint, with no contract to renegotiate
- Your board, your standups, your definition of done
- Knowledge compounds — the same engineers keep working on the same system
- Problems that only appear months later get fixed, because the same people are still there
- The right choice when the roadmap matters more than any single date
How to choose
If the goal fits in one sentence and the date matters most, take the fixed project. If the roadmap matters more than any one date, go long-term. If neither fits in a sentence yet, start with discovery — it is the cheapest way to find out which one you need.
One trap is worth naming: do not go long-term because scoping feels like hard work. It is hard work, and that is exactly why it is worth doing. If nobody can write down what done means, you have a discovery problem, not a staffing question.