
From Monday to Friday: know what it's worth before you build it.
M2F puts a working prototype in your users' hands in one week. They test it on real work for two weeks, and you decide with evidence – go, rework or stop – before the real investment starts.

The costliest mistake is building the wrong thing well.
Most workflow and AI projects spend months on specifications and development before a single user tries anything. By the time people react, most of the budget is gone and stopping feels impossible.
Our experience is simple: whether a solution is really useful weighs far more on its return than what it costs to build. Technical problems can almost always be solved, at some extra cost. A solution nobody needs cannot.
So M2F turns the usual order around. We validate the value first, with a working prototype and the people who will use it. Only what proves useful gets built properly.
“The less you have invested, the easier it is to say stop.”
Five steps, two decisions, no surprises.
- 1
Discovery
2 weeksA 2-hour kick-off with the process owner who can decide, plus up to 6 hours of follow-ups. Together we scope the (sub)process, the data and where it comes from, an architecture that fits your rules, and the backlog of the first sprint.
- Decision 1Build, or fix it another way? Sometimes a clearer rule or a configuration change is enough – and that is a good result.
- 2
The M2F week
Monday to FridayWe build a working prototype with your key users. On Friday, they see it and use it.
- 3
Two weeks of testing
2 weeksYour users try the prototype on their own cases. Questions are welcome, small changes are made on the way.
- Decision 2Go, rework or stop. If the value is not there, we stop – after about five weeks, not five months.
- 4
Implementation
1 to 3 weeksOnly for what was validated: security, access rights, audit trail, connection to your real systems.
- 5
Validation, then handover
Real conditionsReal users, real work, measured against the starting point. Your team is trained to run, change or withdraw the solution without us.
One week. One prototype. Real feedback.
| Day | What happens | Your time |
|---|---|---|
| Monday morning | 1-hour kick-off with the key users who will test it and the person who decides | 1 hour |
| Monday to Thursday | We build. A 30-minute check-in each day with the decision maker keeps us on track | 30 minutes a day |
| Friday afternoon | 1-hour demo and Q&A: users see and try the working prototype | 1 hour |
| The next two weeks | Users test it on their own cases; we answer questions and adjust | Normal work |
About 12 hours of the decision maker's time for the whole cycle – and a decision based on something real, not on slides.
Built to decide, not to demo.
Value first
We test whether the solution is worth having before we spend on making it robust. Technical questions are assessed early and mocked up for the prototype, never ignored.
Stopping is a success
Two explicit decisions per project. A stop after one week saves months and budget for the next idea.
People decide, AI assists
For AI agents, a named person approves anything that leaves the organisation or cannot be undone. Even the prototype only prepares; a human acts.
Governance built in
Access rights, audit trail and a tested way to switch the solution off are designed from the first week and in place before real use – not added at the end.
It runs where you choose – in your environment, on a system we install on your premises, or hosted by us – with your security rules. What we build is yours, documented so your team can run it without us.
Proven on dozens of projects.
of projects delivered with M2F – many stopped after one week, by design
sessions a month on the largest solution born from an M2F week
of a global food & beverage leader's procurement volume processed
From humanitarian organisations to global corporations, M2F helps teams find out quickly what is worth building.
Questions we often hear
Can you really build something useful in five days?
Yes – because the scope is small and prepared. Discovery has already defined the process, the data and the architecture. The goal of the week is not a finished product but a prototype good enough for users to judge its value.
Is M2F only for AI projects?
No. We have used it for years on business workflows. It is especially useful for AI agents, where the value is often uncertain and the effort moves to review, data and governance.
What if the prototype isn't good enough?
Then we decide together: a second M2F week to rework it, or a stop. Both are fine. A stop after five weeks costs a fraction of a project that fails after six months.
Do you work with real data?
The prototype can use synthetic data, a copy or a live connection – decided in Discovery with your IT. Real use only starts once security, access rights and the audit trail are in place and tested.
How much of our time does it take?
About 12 hours for the decision maker over the whole cycle, a few hours for each key user, and short input from IT.
Where does it run, and who owns it?
Where you choose. Three options: in your own environment (your servers or your cloud); on a system we install for you, on your premises; or hosted by us, on our systems in Switzerland. In every case it runs under your rules, and what we build is yours.
Have a process you'd like to test?
Tell us about it. In one conversation we'll tell you whether it's a good candidate for an M2F week – and if it isn't, we'll say so.
