Is the French innovation ecosystem mature enough, are companies transforming, and is funding keeping up? At Tech For Industry Show 2026 (Paris Expo Porte de Versailles), a panel brought together Arthur (France Digitale), Pascal (Crédit Agricole group, aerospace-defense expert) and Raphaël Didier (BPI France) for an unvarnished assessment of industrial innovation funding.

Industry has always innovated heavily in France, BPI France recalls, with major groups filing many patents. But over the past decade, new structures have emerged — industrial startups — carrying projects once reserved for established companies. There are now around 3,500 active players, with a few hundred created each year, mostly in greentech, but also health, electronics, New Space, quantum and defense.

For France Digitale, the ecosystem numbers around 16,000 startups, with a growing share of industrial applications, often dual-use (civilian that can shift to defense or industry). France's assets are real: talent and labs envied worldwide, robust infrastructure, decarbonized and accessible energy, and public support at several levels (BPI, State, Europe, regions). An encouraging sign: talent that had gone abroad is returning to build in France (Mistral's teams are cited).

The picture is more nuanced on digital transformation. France has around 6,500 mid-cap companies but ranks around 24th of 27 in Europe for their transformation — a huge untapped potential. Large groups (CAC40, SBF120) have accelerated, pushed by geopolitics to seek sovereign solutions. The hard part: industrial SMEs (25,000 to 30,000 with over 10 employees), with more limited resources.

Pascal (Crédit Agricole) recalls that aerospace-defense has weathered successive shocks: Covid, the war in Ukraine and supply uncertainty, then an abrupt ramp-up (full order books). The result: leaders "already very busy" as the world transforms (drones, electric aviation, new launchers). BPI France leverages its proximity to support these SMEs through diagnostics and accelerators, combining co-financing, guarantees and innovation tools.

A major challenge looms: generational renewal and the transmission of many SMEs. Often at a second generation viscerally attached to its business, with a third hesitant to commit. It's also an opportunity: the successor will often want to do things differently and invest. France Digitale and BPI France see it as a modernization lever, provided lean-structured firms are supported and sector solidarity is fostered (GIFAS is cited).

BPI France's answer is nuanced but positive: the 3,500 industrial startups have raised over €15 billion since 2022, plus subsidies (France 2030, ADEME, regions, Europe). Last year's rounds alone reached €3.7 billion. France is "rather good," behind the UK and the US — the latter driven by massive amounts heavily concentrated on AI. A range of financing covers each stage of a startup's life, from non-dilutive to the factory.

The point of concern: an acceleration of large rounds but a deceleration of small (seed) rounds. A bad signal, as it dries up the source of future industrial projects. France Digitale welcomes the third Tibi plan fund (€13 billion announced) but warns: investment funds mechanically shift toward "growth" (large rounds), a French and European market gap. How to ensure a substantial share reaches seed stage? On the ground, a community of over 50 incubators is starting to suffer from the disengagement of private funds, corporates and public actors, worsened by local election cycles.

Considerable volumes circulate worldwide and invest in France massively and selectively — foreign pension funds, family capital with "considerable" valuations — but "not always to keep the innovations here." For a bank, Crédit Agricole recalls, financing requires mastering risk: you lend others' savings while having to be able to return them. Hence the difficulty of forming an authoritative judgment on highly specialized projects.

The panel's consensus fits in one phrase: "the best money remains the customer's." Beyond grants and PoCs, it's contracts that grow innovative players. In the US, large defense contracts have gone to players created less than five years ago — still rare in France.

Arthur (France Digitale) quantifies the stakes: public procurement represents €2,500 billion a year in Europe (some 250,000 public buyers). Channeling even 5% toward European tech startups and SMEs would create enormous leverage. BPI France identifies three types of clients to activate: public authorities (the Brittany region, buying massively from its local ecosystem, is cited as virtuous), large groups and mid-caps (still too reluctant to integrate French innovation — hence the DAPI innovative-procurement program), and citizens (preference for competitive French solutions).

BPI France notes 240 new industrial sites created since 2022 thanks to France 2030, including about 140 scale factories. But "when you have the factory, you're not done": it must run, win contracts to cover fixed costs, and survive two years of ramp-up — with companies rarely meeting their business plans. The real challenge is no longer building factories (that's mastered), but sustaining them, which requires everyone's commitment, not just shareholders or the State.

All converge: funding isn't only about money, but also about culture. An innovation culture and risk-taking "can't be decreed": they feed on examples (Alice&Bob in quantum, Fairmat in composite materials are cited). The two worlds must talk: large groups must express real needs, entrepreneurs must understand internal trade-offs (financial, compliance, procurement), and dialogue must extend to research labs to help financiers judge a project's innovativeness. A shared warning: beware the PoC that "only commits those who listen to it" and doesn't lead to scale.

Sovereignty, understood as keeping jobs and companies on national soil, is consensual. France Digitale recalls the value-chain resilience stake ("if a US player cuts cloud access…"): not shifting everything to French solutions, but having this reflection on strategic topics. BPI France distinguishes two aspects: defensive sovereignty (barriers, protection, where Europe must help) and offensive sovereignty — creating robust players offering alternatives to Chinese and American solutions, like Mistral in AI.

What happens if things don't move fast enough? The speakers point to a risk of falling behind in global competition, of technological dependence and of a drying up of seed funding. But all remain combative: a market of 450 million consumers, top schools, global groups, decarbonized energy. The final word goes to BPI France: industry is a long game — "it took time to deindustrialize, it will take time to reindustrialize." Reaching 12% industrial GDP within 10 years would require a very consistent long-term plan, without overly abrupt political reversals.

What is an industrial startup's "valley of death"? It's the critical phase between developing the project and building the factory, requiring far heavier investment over an industrial timeframe of 5 to 7 years — often incompatible with the horizons some investors expect.

Why is seed funding a concern in France? Because funds concentrate on large rounds ("growth"), while small seed rounds slow down. This risks drying up the source of future industrial projects.

Why say "the best money is the customer's"? Because beyond subsidies and fundraising, contracts are what truly grow innovative players. Public procurement (€2,500 billion a year in Europe) and integration by large groups are major, still under-used levers.

What's the difference between defensive and offensive sovereignty? Defensive sovereignty means protecting (barriers, European regulation); offensive sovereignty means creating robust players offering credible alternatives to foreign solutions, like Mistral in AI.

This panel was recorded at Tech For Industry Show 2026, the B2B trade show for industrial transformation (Smart Manufacturing, AI, Supply Chain) at Paris Expo Porte de Versailles. Watch the full discussion above, and join us on May 19–20, 2027, Hall 7, for the next edition.