I Went Through All 236 "Co-Founder Wanted" Listings in France: What They Lack Isn't a CTO

A data analysis based on every live listing on Bpifrance's "Les Deeptech" platform Data captured: September 20, 2026 | Sample: 236 listings · 186 companies · 41 cities

The conclusion first

What French deeptech lacks most isn't technical talent — it's people who can turn technology into a business.

And in healthcare and AI — the two hottest tracks — the gaps point in opposite directions:

• Healthcare: short of sellers (29 CEO listings vs 11 CTO listings)

• AI / software: short of builders (17 CTO listings vs 6 CEO listings)

• Industrial hardware: short of people who can mass-produce (ops/manufacturing listings the most numerous, 11)

This isn't my speculation. It's public data on Bpifrance's own website. Below, the whole process and the data, laid out.

1. Where this data comes from

Bpifrance, France's state investment bank, has run a program called "Les Deeptech" since 2019, dedicated to supporting hard-tech companies that grow out of laboratories. Two services hang under it:

• Tandem — matching deeptech projects with co-founders. The site's own words: "France's first national service for assembling deeptech founding teams."

• Tango — matching deeptech companies with advisory boards. (This program's page URL is still /devenir-advisor, but the program itself has been renamed Tango.)

Both services post their listings publicly on the website for anyone to see. I scraped them all:

First, the limits of this data, so you don't over-weight the conclusions below:

1. These are listings on one Bpifrance platform, not the entire French job market. A 236-listing sample describes the structure of this pool; it does not represent all of France.

1. The "role classification" in the listings is my keyword-based automatic categorization, not an official field. Of the 236, 160 could be assigned to a specific role; the other 76 (32%) went into "other." Treat every number below as an order of magnitude, not precise statistics.

1. The most important point: this data shows "what the posting side wants," not "whether such people exist in the market." A company posting a CEO listing only proves it wants a CEO — it does not directly prove the market lacks CEOs. Proving a genuine supply-demand imbalance requires candidate-side data — which I don't have. I'll come back to this in the final section.

2. Finding 1: They look for a "business co-founder" nearly twice as often as a "technical co-founder"

The full picture first. The 236 listings, broken down by role:

Looking only at Tandem (the 162 listings actually doing "co-founder matching"), the business-vs-technical contrast is sharper:

Business side (CEO + commercialization): 79 = 48.8% Technical side (CTO): 45 = 27.8%

The business side is 1.8x the technical side.

I worried this conclusion was skewed by stale listings, so I cut the data again — only listings published in the last 90 days:

Of the last 90 days, 32 Tandem listings: business side 17 (53%) vs technical side 7 (22%)

The gap widens to 2.4x. The fresher the data, the more visible the business-side gap.

Why are CEO listings the most numerous?

The answer is hidden in how Tandem operates.

Tandem's project pipeline is not submitted by entrepreneurs — it comes from French public laboratories. The partner institutions listed on the official page include Réseau SATT (France's tech-transfer network), CNRS Innovation (the national research center), Inria Startup Studio (the national computing institute), INSERM Transfert (the national health institute), plus 30 more institutions spread across the country.

Look at how the official process describes step one:

"I discover projects: browse the startup projects and research teams of France's top laboratories and the founding partners they need."

The technology side is fixed; the seat being "matched" is the business side.

In other words: researchers stand there with patents and papers; what they lack isn't another technical genius — it's someone who can be CEO: someone who can translate the technology into a business story, go talk to customers, and raise money.

That also explains why the job title "CEO" appears so often in the data: it doesn't mean "a boss hiring"; it means "we bring the technology, you bring the business, and we partner on equity."

3. Finding 2: Three sectors, three opposite gaps (the most counter-intuitive part)

Stop at "they lack business people" and the conclusion is too coarse. Slice by sector and three very different structures appear:

Note: sector tags were assigned automatically by keywords, and one listing can hit two sectors, so the sector counts sum to more than 236 (another 62 listings matched no sector tag). This is a property of the classification method, not a data error.

The same structure holds in the last-90-days data: healthcare business 5 : tech 1; AI/software business 1 : tech 2.

Why do three sectors need completely different people? My read:

Healthcare — the technical barrier is in the lab; the commercial barrier is in the market. Once a biopharma project has experimental results and a patent, the bottleneck instantly shifts from "can it be built" to "who can push it into hospitals." Someone must handle clinical trials, regulatory approvals, reimbursement, physician channels — things researchers can't do, and more time-consuming than the technology itself. Hence healthcare's loudest cry is for the person who can open the market.

AI / software — the commercialization floor has been flattened while the technology bar has risen. The biggest change in AI products in recent years: selling has gotten easier; building something others can't build has gotten harder. Anyone can wrap a product around GPT; very few can train models or create real technical differentiation. Hence the AI track's desperate hunt for a CTO.

Industrial hardware — designing it ≠ being able to manufacture it. This is the classic valley of death. A prototype that works in the lab hits four walls at mass production: yield, supply chain, certification, cost. The most direct evidence in the data: 55.1% of listings mention "industrialization / mass production" — more than half are talking about how to make the thing. Industrial hardware has the most ops/manufacturing listings (11), which fits this logic perfectly.

This is useful for judging "should I go to Europe": If you're a pure AI technical team, the French market isn't friendly to you — their gap is technical people of their own; if you hold medical-device regulatory, market-access, and channel capabilities, the gap here is on open display.

4. Finding 3: Almost no one is hiring a CFO

Of the 236 listings, only 4 are "finance / fundraising" roles — 1.7%. Only 2 in the last 90 days.

For comparison: 56 CEOs, 46 CTOs.

In other words: money is not their core anxiety.

This observation matches external data. Per Bpifrance, French deeptech companies raised €4.1 billion in 2025 — about half of all French tech funding; and Paris became Europe's largest deeptech city by funding in 2025 ($3.5 billion, ahead of London's $3.1 billion).

Abundant capital on one side, almost no finance hires on the other — which means the real bottleneck is people, not money.

5. Finding 4: 65.6% of listings have been up for more than six months

The most surprising thing I found while scraping:

65.6% have been up for more than six months; 82 of them for more than a year.

Honesty requires laying out both interpretations:

Interpretation 1: the platform doesn't clean up expired listings. Pure dirty data; proves nothing. Interpretation 2: listings in this pool aren't being filled quickly.

I lean toward the second, but can't prove it. Listings still show "live" status and all the links open — whether a company is "still hiring" or "forgot to take it down" is indistinguishable from the outside.

Either way, the practical conclusion is the same:

This is not a "first come, first served" market. Listings staying up for months means they are waiting, continuously, for someone.

6. Finding 5: Only 10.6% of listings mention English

Of the 236 listings, only 25 mention English ability in the job description.

The other 211 never mention it at all.

That doesn't mean "English not required" — quite the opposite: it means French is the default.

This wall isn't written in the job requirements, but it is real: Tandem's own online matching events are entirely in French, the project pipeline is French public laboratories, and founders' day-to-day communication is in French.

For teams without French, this is the most underestimated barrier of all.

7. Finding 6: Half the listings talk about equity — but rarely with formal equity instruments

• 36.0% (85 listings) explicitly mention "partner / equity" terms (associé, parts du capital)

• But only 4.7% (11 listings) mention formal French employee-equity instruments such as BSPCE or stock-options

• Only 3 listings mention "unpaid / low pay"

• 43.6% (103 listings) mention fundraising-related arrangements

The essence of these listings is "equity for people": you aren't being hired — you're being invited to become a shareholder. Which also explains why they fill slowly: people willing to accept "a pay cut or no pay + uncertain equity" are inherently scarce.

By the way: McKinsey's October 2025 European deeptech report lists "competitive equity incentives" as one of six levers to accelerate European deeptech. This data matches exactly.

8. Finding 7: Bpifrance itself is the hardest evidence of the talent gap

Everything above is what I read out of the data. But one piece of evidence requires no inference —

With public resources, Bpifrance has deliberately built three things:

1. Launched Tandem, a national service to "help research projects find business co-founders"

1. Launched Tango, helping startups assemble advisory boards

1. Signed partnerships with 8 of France's top engineering schools to channel talent into deeptech faster

A state investment bank created a dedicated department and platform to help researchers find business partners. That fact alone is the most persuasive evidence — it needs no percentages behind it.

External research points the same way:

• McKinsey (July 2024, European deeptech report) describes deeptech founding teams as typically "comprising technical talent with academic or corporate R&D backgrounds and, ideally, combining both technical and business capabilities." Note the second half — this is McKinsey saying: technology alone isn't enough.

• McKinsey × Boardwave (June 2025), in "Europe's moonshot moment," lists the "shortage of commercial and GTM (go-to-market) expertise" as one of Europe's systemic bottlenecks. The report also gives one number: European software companies take 15 years on average to reach €100M ARR — 5 years longer than their US peers.

• Bpifrance innovation director Paul-François Fournier has said publicly: the need is to "broaden the entrepreneurial talent base" and to "combine scientific excellence with business capability."

9. What this means for Chinese teams (the sober version)

I have no intention of packaging this data as "Chinese teams' opportunity has arrived." It isn't.

The bad news first:

• French is a hard barrier. Only 10.6% of listings mention English. Tandem's pipeline is French public laboratories; the working language is French. This is the most underestimated wall.

• These listings want a "person," not a "service." 56 CEO listings means "please send a person to be my co-founder," not "please send a Chinese company to do my market entry." You cannot fill a founding-partner seat — someone who signs, holds shares, and bears responsibility — with a team back in China.

• You haven't seen the candidate side of the data. A CEO posting proves the company wants a CEO; how many business talents in France are actually looking for such roles, I don't know. If supply is actually ample, then the "gap" is only the posting side's preference, not a real window of opportunity.

Now the part I think is genuinely useful:

Read differently, this data is a checklist of "what capabilities French deeptech lacks" — and these capability gaps happen to be deliverable as services:

Note: this checklist has two layers of credibility. "These listings exist" is hard data; "Chinese teams can fill them" is my speculation, unverified. There is only one way to verify it — go talk. Send an email, book a meeting, and see whether they reply "we already have someone" or "can you solve this for us?"

Finally, if one sentence must be distilled from this data for anyone doing China-Europe business, it's this:

French deeptech doesn't lack technology, doesn't lack money, doesn't lack policy. It lacks people who can turn lab work into a business. And that job happens to cross languages, cross cultures — and can be outsourced.

10. Three things you might want to challenge me on

Let me preempt the most likely objections myself:

1. "236 listings is too small a sample." Fair. It's a cross-section of one platform, not all of France. I cross-validated with the "last 90 days" subset (same conclusion, wider gap), but that still doesn't replace a full-market statistic.

2. "Your role classification is automatic — couldn't it be wrong?" Yes. Of the 236, 160 could be assigned to a specific role; 76 went into "other." I manually audited and fixed four categories of misclassification (including classifying an EV-battery company as healthcare and environmental monitoring as defense), but the remaining tags are still an "order-of-magnitude reference," not precise statistics. I cross-checked every conclusion across two time windows precisely to reduce the impact of single-point error.

3. "A posted listing ≠ can't hire." Entirely correct — this is the analysis's biggest weakness. As I said clearly in Section 5: this data only proves "what the posting side wants"; it cannot prove "the market lacks supply." Proving the latter requires candidate-side data — e.g., "how many business talents are looking for deeptech co-founder roles" and "how long does an average match take." Bpifrance doesn't publish that data, and I can't obtain it.

So read this piece as "a demand-side portrait from the people posting the listings," not as a "supply-demand report on the French deeptech talent market."

Data & sources

Primary data in this article

• Source: Bpifrance "Les Deeptech" platform public interface (lesdeeptech.fr)

• Capture date: September 20, 2026

• Sample: 236 live listings (deduplicated by primary business key)

• The site's robots.txt declares Allow: /

• Fully queryable version: all 236 listings searchable, filterable, and exportable by city/type/institution

External citations

• Bpifrance 2025 annual deeptech data (2,830 active companies, 410 new per year, €4.1B raised, about half of French tech funding, 200+ industrial sites, 15 unicorns): Bpifrance annual summary

• Paris as Europe's largest deeptech city by funding in 2025 ($3.5 billion, vs London's $3.1 billion): Dealroom "French Deep Tech Report 2026"

• Les Deeptech program launched January 2019; Bpifrance official figure €2.5B over 5 years

• Tandem's official positioning and partner institutions (Réseau SATT, CNRS Innovation, Inria Startup Studio, INSERM Transfert, etc.): lesdeeptech.fr

• Deeptech founding teams "combining technical and business capabilities," Europe's shortage of commercial and GTM expertise, European software companies averaging 15 years to €100M ARR: McKinsey "European Deep Tech" (2024-07), McKinsey × Boardwave "Europe's moonshot moment" (2025-06)

• "Competitive equity incentives" listed among six levers; deeptech unicorns averaging 6 years (vs nearly 8 for traditional tech): McKinsey "Europe's deep-tech engine could spur $1 trillion in economic growth" (2025-10)

• Bpifrance innovation director Paul-François Fournier on "combining scientific excellence with business capability": Bpifrance official press release

(End)

Appendix: Social-media short version (~400 words, for Xiaohongshu / X / LinkedIn)

I went through all 236 "co-founder wanted" listings in France. The conclusion runs against intuition.

France's state investment bank Bpifrance runs a platform that matches lab-born hard-tech projects with partners. The listings are all public — I scraped 236 of them, across 186 companies.

What they lack most isn't a CTO.

Of the 162 "co-founder matching" listings:

• business co-founder (CEO / commercialization): 79, 48.8%

• technical co-founder (CTO): 45, 27.8%

Last 90 days only: the gap widens to 2.4x.

The reason is in the project pipeline: the projects come from French public laboratories (CNRS, Inria, INSERM) — researchers are the initiators; they hold the patents and lack the person who can turn technology into a business.

More interesting: the three sectors need opposite profiles:

• Healthcare (86 listings): 29 CEO vs 11 CTO → needs sellers

• AI / software (51): 17 CTO vs 6 CEO → needs builders

• Industrial hardware (37): ops/manufacturing the most numerous → needs people who can mass-produce

A few more details:

• Only 4 finance listings (1.7%) — money isn't their anxiety

• Only 10.6% of listings mention English — French is a wall that goes unwritten

• 65.6% of listings have been up for over six months — this is not a first-come-first-served market

One-line summary: French deeptech doesn't lack technology, money, or policy. It lacks people who can turn lab work into a business.

And that job crosses languages and cultures — and can be outsourced.

(Data captured 2026-09-20; sample of 236 listings. Role classification is keyword-based and for reference only.)

English edition draft · Data as of 2026-09-20 · Source: Europe Startup Playbook (Chinese edition)