2026 is a contraction year for French apprenticeship, and the dominant explanation is financial: aid cut, budgets tight, companies retreating. The state's own survey data tells a different story about the companies that are missing.

Among companies that take no apprentices, only 6% cite insufficient aid or excessive cost as the reason (France Compétences, EFE 2024 survey, published February 2026, results still flagged provisional). Money is the sixth conversation. The first one never happened at all: 44% of companies under five employees that use no apprenticeship simply never considered the option. Even at 250 to 999 employees, 27% never considered it.

That claim needs precision, because there is a counter-figure you should know. In a 2021 DARES survey, 79% of companies not recruiting alternants said they did not wish to. So no, the market is not full of companies secretly longing for an apprentice. But look at what both surveys actually measured: an opinion about a concept, declared to a statistician. Nobody had put a concrete proposal in front of these companies, with a specific school, a specific program, a candidate, or a February re-placement carrying a 200 euro participation instead of 750. The companies in your territory that host no apprentice have, in their overwhelming majority, never received an offer they could say no to.

That is the hidden market. The data to size it now exists, and most of it comes from the state.

How small the visible market really is

Start with the recruiting side. About 367,000 companies took at least one new apprentice in 2024 (France Compétences, Deca data). That number grew 38% between 2020 and 2024, which is the quiet story of the boom: much of the growth came from companies entering the system, especially very small ones. Companies with one to four employees are the largest single group of recruiters, at 41%.

Now the non-recruiting side. In 2021, 17% of private-sector employers recruited at least one alternant, and among companies under ten employees the figure drops to 13% (DARES Analyses n° 77, December 2024). More than four out of five employers, and roughly seven out of eight small ones, were outside the system entirely in the middle of the biggest apprenticeship boom in French history.

Why do the ones inside participate? Their declared motives are worth reading, because financial aid is not on top. 56% recruit to train someone to their needs, rising to 75% at 250-plus employees. 31% cite the extra workforce. The hiring aid comes third, at 18% (EFE 2024, provisional). Companies come to apprenticeship for people, and stay for people. The aid moves timing, level and volume. It rarely creates the intent.

The honest counterweight, from the same survey: 17% of very small companies say they renounce an apprenticeship hire for lack of tasks to offer. That objection is real and almost three times more cited than cost. It is also exactly the kind of objection a conversation can test, and a job board never will.

The state already proved the hidden market is detectable

La Bonne Alternance, the public platform run by the DGEFP, describes its own core mechanism like this: facing the contraction of published offers, it identifies hidden-market companies by crossing their past hiring records, their financial data and the applications they already receive. Behind it, France Travail states that a predictive algorithm analyzing the last six years of hiring identifies around 300,000 companies with strong hiring potential every month. Companies that have posted no job offer (France Travail, service page, consulted 28 August 2026).

Put that next to the platform's own flow numbers for the first quarter of 2026: 5,950 offers deposited directly on the platform, against roughly 320,000 applications sent, of which 74% were spontaneous applications to companies with no posted offer (La Bonne Alternance, baromètre T1 2026).

Read those three numbers together. The visible market, the one your team scrolls through on job boards, is a few thousand offers a quarter. The state's own algorithm finds three hundred thousand plausible employers a month that published nothing. And three quarters of the candidate flow already goes around the offers entirely.

So the hidden market is a measured, state-documented population, refreshed monthly, rather than a sales metaphor.

Why none of this reaches your prospecting file

Every French student has heard of the hidden market: La Bonne Alternance, school career pages and student blogs all teach spontaneous applications. The entire apparatus is candidate-facing. It helps one student send one application.

Nothing equivalent exists on the school side, and the school side cannot simply borrow the candidate tool. La Bonne Alternance's terms of use restrict commercial exploitation of the platform's content, so its company list cannot become your prospecting file. And a list without program-level qualification, contacts and follow-up is not a prospecting file anyway.

Meanwhile, look at what a school's own sources cover:

We wrote previously about the quality of the visible flow, and why keyword-matched job board feeds waste your coordinators' time on false leads. This is the companion argument, and it is the bigger one: the visible flow is noisy, and above all it is small. The companies outside it never refused anything. Nobody asked.

What asking looks like at territory scale

One artisan looking for a first apprentice does not justify a process. Thirty thousand establishments in a school's catchment area do.

The signals are public: past apprenticeship hiring in the state registers, workforce growth, new establishments, activity codes matching your programs, the sector mix where apprenticeship actually concentrates. The method is what the France Travail algorithm already validated: past behavior and current capacity predict hiring intent better than the presence of a job ad. What no public tool provides is the school-side workflow on top: qualifying each company against your specific programs, finding the right person to talk to, running the outreach, and logging it all where your team works.

That is the part we built Alternel for. It watches the activity signals across a school's territory, qualifies companies program by program, enriches the employer-side contact, and runs the first conversation at a scale a placement team cannot reach by hand. The state finds the hidden market for students. Alternel works it for schools.

The pitch this changes

A contraction year punishes the schools whose prospecting list is a copy of the visible market, because the visible market is what contracts. Offers get rarer. Competing schools crowd the same postings harder.

The 44% who never considered apprenticeship are not waiting for you. Some will say what they told the surveyors: no wish, no tasks, no time. Survey answers to abstract questions are the softest form of no that exists. The concrete question, this program, this candidate, this financing, has mostly never been put to them.

In a shrinking market, the teams that grow are the ones asking questions nobody else is asking, to companies nobody else is calling.

If your prospecting still starts from job boards and your existing partner file, we can map what the hidden market looks like in your territory. Thirty minutes of conversation is enough to size it.

Sources

1. France Compétences - Les entreprises recruteuses d'apprentis entre 2020 et 2024: 44% / 27% never considered, 6% cost objection, recruiting motives, 367,000 recruiters, +38%. EFE 2024 figures flagged provisional by the source.
2. DARES Analyses n° 77 - Quelles entreprises recourent à l'alternance ?: 17% of private employers recruit an alternant (2021), 13% under 10 employees, the 79% counter-figure.
3. France Travail - La bonne alternance: the ~300,000 companies/month identified by the predictive algorithm (page undated, consulted 28 August 2026).
4. La Bonne Alternance - Baromètre T1 2026: 5,950 offers deposited, ~320,000 applications, 74% spontaneous.
5. La Bonne Alternance - Page recruteur: description of the hidden-market algorithm's inputs.
6. DARES - PoEm, tableau de bord des politiques de l'emploi: 2026 contraction: 955,000 under contract at end-June, higher-ed entries -16.2% year on year.
7. Alternel - The false trail: keyword matching on job boards: the companion piece on the quality of the visible flow.