Almost every large French group runs an intrapreneurship programme. A call for projects, a jury, a few months of coaching, a demo day in front of the executive committee. Far fewer have created a company. The question nobody asks: how many of these projects became businesses, with their own legal entity, their own accounts, their own CEO and their own cap table?
Since September, we have been counting those companies in the Observatoire of Corporate Venture Building. Of the 40 groups in the CAC 40, only 13 have at least one company created from zero that meets our criteria. 27 have none, and 21 do not even have a candidate. The programmes exist. The companies, much less.
The three-year test
A successful intrapreneurship project should pass a simple test after three years: is there a registered company, with its own accounts, its own CEO and shareholders who have something to lose? If not, it was a project, however well supported.
This is not about legal form. A company can raise money, hire, sign customers the group would never serve, and be sold. A project cannot. It lives as long as its sponsor protects it, and stops the day that sponsor changes jobs.
A POC is not a business
Launching a POC or an MVP and launching a new business are two different things, and programmes often confuse them. A POC answers a technical or usage question: does it work, does anyone use it? An MVP tests a value proposition with a handful of customers. A new business is a market you intend to win, customers who pay, a team that hires, funding over several years, and someone accountable for all of it.
Many programmes stop at the successful POC and call it a win. A successful POC proves that an idea works. It does not prove that a company can exist. In between, someone has to decide to create the company, fund it and hand it to a leader. That is exactly the step most groups never take.
Four reasons projects never become companies
1. An employee is not a founder
The best intrapreneurs are often excellent managers. It is not the same job. A founder decides with incomplete information, hires before the money is there, drops a direction in a week. A manager was promoted for the opposite: secure, consult, avoid mistakes. The first eighteen months of a company show the difference every time.
2. No equity
An intrapreneur keeps their salary, their status and their career path. If it fails, they get another role. If it works, they get a bonus. This is not a criticism, it is a structure. But it does not produce the level of commitment that building a company requires. When nobody has anything to lose, nobody really has anything to win.
3. The budget is renegotiated every year
Programmes fund drop by drop: one envelope for exploration, another for the pilot, a third to be argued for at the next budget round. That is the right way to fund a project. It is the wrong way to fund a company, which needs time to dig into its market and to invest before revenue follows. A team that has to convince its own shareholder every twelve months spends its energy on the group, not on its customers.
4. The project stays attached to a business unit
Until it is a company, the project has to follow the group's rules: IT choices, legal sign-off, pay and bonus policy, compliance, procurement, brand. Each of these rules has a good reason to exist in a group of 100,000 people. Together, they make a startup's pace impossible. The first customer contract takes six months, the first hire does not fit the salary grid, the chosen tool is not on the approved list.
Those who get there
The groups that produce companies have one thing in common: they built a structure whose job is to produce them. EDF Pulse Incubation has spun out seven companies, all of which meet our criteria. Crédit Agricole has six, mostly out of La Fabrique by CA. AXA, through Kamet, has six.
In these cases, the project does not stay a project. It is incorporated early, it has an identified leader, its own capital and funding committed over several years. That is no longer intrapreneurship in the strict sense: it is corporate venture building, the creation of a company with the group rather than by it, which the group controls but does not run.
From project to company
For an innovation director who wants their programme to produce companies, four decisions change everything, and they answer the four blockers one by one:
- Put an entrepreneur in charge, alongside the intrapreneur who carries the idea and knows the group.
- Open the cap table to the founding team, with an option for the group to go to 100% on terms set from the start.
- Commit the funding in two tranches at launch, the second released on criteria known in advance, not on a budget arbitration.
- Incorporate early, so the company has its own tools, its own contracts and its own pay policy, and draws on the group's rare assets without carrying its rules.
An intrapreneurship programme should be measured by the number of companies it creates, not the number of projects it supports.
On 10 December, the Observatoire publishes the first complete map of these companies in France, group by group. If your programme created a company that is not on the list yet, now is the time to tell us.