Message-Market Fit Observatory France:
61% of startups hide their differentiator.
744 French post-funding startups audited across two editions, in June and then November 2026. The second, published in partnership with Paris Business Angels, covers 375 startups. Here are its three most telling figures, then what changed from one edition to the next.
Q4 2026 edition published in partnership with 
61% have a real differentiator, but off their home page.
A prospect lands on the home page and leaves without knowing what sets them apart from a competitor.
55% see another player cited in their place by Google’s AI (205 of the 374 startups measured).
The researching buyer gets a competitor’s name, and the shortlist is drawn up without them.
R² = 0.001 between amount raised and message quality, across 367 startups.
The money raised does not fix the message. It funds campaigns built on an argument the prospect does not remember.
Which sectors communicate best?
From one sector to another, the average message score ranges from 35.4 to 50.4 out of 75. Sectors that sell to business decision-makers lead. Biotech, space and deep tech come last.
In a sector at the bottom of the table, the first startup to say what its customer gains takes the spot.
| Sector | Startups | Score | |
|---|---|---|---|
| Cybersecurity | 10 | 50.4 | |
| HR | 12 | 49.3 | |
| Marketing / Retail | 19 | 48.0 | |
| FinTech / Insurance | 33 | 47.5 | |
| AI | 51 | 45.9 | |
| Software | 21 | 44.0 | |
| Mobility / Logistics | 15 | 42.7 | |
| Industrial | 9 | 42.3 | |
| Energy / Climate | 43 | 40.9 | |
| Health / MedTech | 39 | 40.4 | |
| Agri / Food | 9 | 40.4 | |
| DeepTech | 39 | 38.1 | |
| Space | 12 | 36.9 | |
| Biotech | 31 | 35.4 |
Your startup sits somewhere in this table.
Is a good message enough to be cited by AI?
No. The ranking almost flips. Second to last on message, space is the most cited sector: 83% of its startups. AI, fifth on message, drops to 25%.
In a crowded category, Google’s AI cites the established players. A startup that answers no specific buyer question stays off the list.
| Sector | Score out of 75 | Cited | |
|---|---|---|---|
| Space | 36.9 | 83% | |
| Agri / Food | 40.4 | 78% | |
| Industrial | 42.3 | 67% | |
| HR | 49.3 | 50% | |
| FinTech / Insurance | 47.5 | 45% | |
| Health / MedTech | 40.4 | 44% | |
| Energy / Climate | 40.9 | 42% | |
| Cybersecurity | 50.4 | 40% | |
| Mobility / Logistics | 42.7 | 40% | |
| Biotech | 35.4 | 39% | |
| Marketing / Retail | 48.0 | 32% | |
| DeepTech | 38.1 | 31% | |
| AI | 45.9 | 25% | |
| Software | 44.0 | 24% |
What do these figures tell a founder and an investor?
The same audits can be read two ways.
If you run a startup
Your home page says what you do. In 87% of cases, it does not say what the customer gains, and in 77% it does not name who it is for. A prospect who does not recognise themselves leaves no contact details.
If you invest
Neither the amount raised nor the stage fixes the message: three Series B startups out of four still need to consolidate it, just like Seed startups. Of 42 funds, 2 stand out from the pack, which is what chance alone produces. The message remains a value lever that almost nobody works on.
What changed between the two editions?
The panel is new, and the level has not moved. Five points still set the two editions apart.
A new panel, the same level
Re-audited with the second edition’s grid, 30 startups from the first score 43.5 out of 75, against 42.8 for the 375 new ones. The gap is not significant.
Your competitors therefore share your weakness, and fixing it gives you a lead.
The gap between funds does not survive a chance test
The Q2 edition found a 32% gap between the best and worst placed funds. Run again on 42 funds, the exercise gives a gap that randomly drawn groups also produce. At this portfolio size, no fund stands out.
For an investor, the fund’s name guarantees nothing, and each portfolio company has to be looked at on its own.
The AI sector moves above average
Average in Q2, it scores 45.9 out of 75 in Q4. Yet Google’s AI cites only 25% of its startups.
A good message is therefore not enough to get into the answers buyers read.
AI citation is now measured
Google’s AI received 1,125 buyer questions, three per startup. It never cites 59% of the startups. Being readable by bots is not enough: cited and uncited startups have the same GEO score.
As a result, some buyers draw up their vendor list without ever reading their name.
A co-published edition
The Q4 edition is written with Olivier Baroux, President of Paris Business Angels. Its Investors part takes up a whole chapter.
“I want to commend the quality of the work done by Hervé Dhélin and Fast Growth Advisors. Their expertise sheds a particularly relevant and practical light on what is at stake in message-market fit. This Observatory is a reminder of an essential reality: having excellent technology or a differentiating value proposition is not enough. You still have to know how to put it into words and make it clear, understandable and attractive to the market.”
Olivier Baroux, President of Paris Business Angels (translated from French)
Can the scores of the two editions be compared?
No, and the Q4 report says so itself.
The first edition’s scores rest on a different scale and do not map onto the second’s. A higher average in November than in June therefore signals no progress. The only valid comparison is the 30 startups re-audited with the same grid.
Where can the two reports be read?
Online, in full, with no sign-up.
Q4 2026 edition
It covers 375 startups, 15 criteria and 1,125 questions asked to Google’s AI. Published in November 2026 in partnership with Paris Business Angels. The full report is in French.

Q2 2026 edition
The Observatory’s first edition: 369 startups, data as of 14 May 2026.
How do I get the study as a PDF?
Leave your work email: you will receive the Q4 2026 edition as a PDF (in French) once your address is confirmed.
Double opt-in. No spam. Data processed in line with the GDPR.
What method is the Observatory France based on?
- Authors: Fast Growth Advisors, in partnership with Paris Business Angels for the Q4 2026 edition.
- Panel: 375 French post-funding startups, all new to the Observatory.
- Dates: audits carried out between 29 October and 5 November 2026.
- Method: 15 criteria scored out of 5, a five-layer GEO audit scored out of 100, and 1,125 buyer questions asked to Google’s AI. Public data, nothing else.
How to cite this study
Reference: “Fast Growth Advisors and Paris Business Angels, Message-Market Fit Observatory France, Q4 2026 edition, November 2026, fast-growth.fr/en/observatory/france/”
The figures may be reused freely, with the source credited and a link to this page.
Frequently asked questions about the Observatory France
How many French startups has the Observatory audited?
744 in 2026, across two editions.
The Q2 2026 edition covers 369, the Q4 2026 edition 375. All 375 are new to the Observatory.
What is the main messaging problem of French startups?
They do not show what they have.
61% of them have a real differentiator, kept off the home page. And 42% leave their quantified results in the press or on inside pages.
Does raising money improve the message?
No.
Across 367 startups, the correlation between amount raised and message quality is zero (R² = 0.001). Three Series B startups out of four still need to consolidate their message.
Who publishes the Observatory France?
Fast Growth Advisors, with Paris Business Angels for the Q4 2026 edition.
The Q4 edition is written by Hervé Dhélin and Michel Crestin (Fast Growth Advisors) and Olivier Baroux, President of Paris Business Angels.