PoCs with startups: how to keep track of them and what to bring to the board
A PoC, a pilot project with a startup, reaches the end when the objective, the measure of success, the duration and the owner are written before it starts, and when the result enters the portfolio of initiatives and lands on the board’s table. In this guide: how to set one up, how to keep track of it, the most common mistakes, what to bring to the board and the tools compared.
30-minute demo, on your case · No commitment · We reply within one working day
- What has to be written before starting a PoC with a startup
- How to keep track of it without spreadsheets and slides that get lost
- The mistakes that leave PoCs half done
- What to bring to the board, and with which numbers
Large Italian organisations chose blendX to govern their innovation.




- What is a PoC with a startup?
- A small-scale test, with a written objective and a fixed duration, to decide whether a startup’s solution really works in your company.
- How long should it last?
- Usually eight to twelve weeks. Any longer and the PoC becomes a project with no budget and no decision.
- Why do so many PoCs never finish?
- Because an exit criterion written beforehand and a person who follows it afterwards are missing: nobody decides whether to scale, change or stop.
- What does the board want?
- A few numbers it can hold: how many PoCs, in what state, what they proved and what gets decided. A picture, with the history behind it.
What is a PoC with a startup, and how is it different from a demo?
A PoC (proof of concept) with a startup is a pilot project with a fixed duration and a reduced scope: the startup’s solution is tested on a real case inside the company, with a written objective, a measure of success decided beforehand and an end date. A demo shows that the product exists; a PoC proves that it works in your context, with your data and your people. It always closes with a decision: take the solution to a wider scale, change something and try again, or stop. Stopping is a result too, as long as the reason is written down.
Sources cited under each number
Why do so many PoCs with startups never reach the end?
Almost every company that meets startups starts PoCs. Few close with a decision. The reason shows when you look at the three phases of a pilot: before, during and after. In most cases the "before" phase is skipped, the "during" phase lives in a chat and a spreadsheet, and the "after" phase is a presentation nobody ever opens again.
| Phase | What gets decided | What has to stay written |
|---|---|---|
| Before | Objective, measure of success, duration, owner, what happens if it succeeds or fails | One page with these five things, signed by whoever has the need |
| During | Weekly progress, open issues, data collected against the starting point | The status of the PoC, updated in the same place as the others |
| After | Scale, change and try again, or stop | The decision, who took it, with which numbers, and what happens now |
What has to be written before you start
A PoC is decided before it begins, on a single page. You do not need a twenty-page contract; you need five lines everyone has read. If one is missing, the pilot starts anyway, but nobody will ever be able to say whether it went well.
- The objective, as whoever has the problem would say it: "cut invoice checking time by 20%", and the context where it is tested.
- The measure of success, with the starting point recorded beforehand: without the "before", no number from the "after" means anything.
- The duration, with an end date: eight or twelve weeks, and a meeting already scheduled to decide.
- Who follows it: one person in the company and one in the startup, with the time to do it, and who approves the final decision.
- The exit criterion: with which numbers you scale, with which you change, with which you stop. Decided beforehand, when nobody is attached yet.
The mistakes that leave PoCs half done
They are always the same four, and they show in the first week.
- Starting without an exit criterion: in the end everyone judges with their own yardstick, and the decision slips.
- No owner inside the company: the startup pushes, but nobody inside opens doors and collects the data.
- Measuring at the end: the starting-point data is gone and the result cannot be proven.
- Keeping each PoC on its own, in a different spreadsheet: the board sees separate presentations and no picture.
What to bring to the board: the picture of all PoCs, with the pilot’s numbers
The board does not ask for the detail of every experiment. It asks for a picture that holds all PoCs together: how many started, what state they are in, what they proved and what gets decided. The numbers that hold up are those of the pilot and of the portfolio: how many PoCs move to a wider scale, how long it takes from start to first result, what a pilot costs on average, how many are stopped in time instead of dragging on. For that picture to exist at the end of the quarter without rebuilding it by hand, every PoC has to live in the same place as the others, with its own card, status and decision. This is what blendX does: the PoC Management module holds the card and status of every pilot, the portfolio shows them together, and the board report is composed from data that is already there, with the history of every decision. All in a European cloud, priced by activated modules rather than by user.
Source: The innovation KPIs to present to the board, EconomyUp (in Italian)
How to manage a PoC with a startup in five steps
The method fits in one page written before and one decision written after. In between, the discipline of updating the status every week.
- 1Write the PoC on one pageObjective, measure of success, duration, owner, exit criterion. Have it signed by whoever has the need and whoever approves the budget.
- 2Record the starting pointMeasure the "before" with the same yardstick you will use for the "after". It is the step everyone skips, and the one that makes the result impossible to prove.
- 3Name who follows it, inside and outsideOne person in the company with the time to do it, and one in the startup. Without the internal owner the pilot stops at the first closed door.
- 4Update the status every week, in the same placeProgress, issues, data collected. All PoCs on the same platform, so the picture exists without having to rebuild it.
- 5Close with a decision and bring it to the boardOn the set date you decide: scale, change, stop. The decision stays written with the numbers, and enters the quarterly report.
Where to keep track of PoCs: the tools compared
Many companies keep PoCs in spreadsheets and slides, or in a project management tool built for software development. There are also tools built only for pilots with startups. An innovation management platform like blendX keeps the PoC connected to what comes before, the need and the startup found, and to what comes after, the portfolio and the board.
| RecommendedblendX | Spreadsheets and slides | Generic project management | |
|---|---|---|---|
| The PoC is linked to the need and the startup that generated it | |||
| Exit criterion written before and decision recorded after | by hand | ||
| All PoCs in one portfolio, each with a status | by hand | ||
| The board report is composed from data already there | to be built | ||
| History of decisions, with who and why | |||
| Data in a European cloud | depends |
Innovation results in front of the board
The hands-on checklist for presenting the results of innovation initiatives to the board: the metrics that hold up, the mistakes to avoid and the five-step method. In Italian.
Checklist · PDF, 6 pages · free · ItalianFrequently asked questions
What is a PoC with a startup?
It is a pilot project with a fixed duration and a reduced scope in which a startup’s solution is tested on a real case inside the company, with a written objective, a measure of success decided beforehand and an end date. It closes with a decision: scale, change and try again, or stop.
How long should a PoC last?
Usually eight to twelve weeks, with a decision meeting scheduled at the end. A PoC that lasts six months is a project that started with no budget and no exit criterion: better to close it and, if needed, open a new one with a clearer scope.
What has to be decided before starting a PoC?
Five things, on one page: the objective in the words of whoever has the problem, the measure of success with the starting point recorded, the duration with an end date, who follows it in the company and in the startup, and the exit criterion (with which numbers you scale, change or stop).
How do you judge the result of a PoC?
By comparing the measure decided beforehand with the starting point recorded beforehand, and applying the written exit criterion. The judgement is about the pilot: did it prove what it had to, in that context? A PoC that stops is a result too, as long as the reason is written down and whoever had the need agrees.
What should you bring to the board about PoCs?
A picture of all PoCs: how many started, what state they are in, what they proved and what gets decided. The numbers that hold up are those of the portfolio: how many pilots move to a wider scale, how long it takes from start to first result, what a PoC costs on average and how many are stopped in time. A few indicators, each with its history.
Does blendX help manage PoCs with startups?
Yes. The PoC Management module holds the card, status and decision of every pilot, linked to the need and the startup that generated them; the portfolio shows them together and the board report is composed from the data already entered. It is governance and traceability of projects, from the need to the decision, in a European cloud and priced by activated modules.
You might also like
Want to see how a PoC is followed from start to decision?
We show you blendX on a real pilot of your company: the PoC card, the portfolio and the board report.
- 30-minute demo, on your case
- No commitment
- We reply within one working day