Qualiopi evidence: which proof for which indicator

The cardinal principle: what isn't recorded doesn't exist

The Référentiel National Qualité comes down to 7 criteria and 32 indicators. Each indicator describes a requirement the organisation must demonstrate on the day of the audit. The verb is the right one: demonstrate, not assert. A director may be certain that their pedagogical follow-up is beyond reproach, but if nothing attests to it (a procedure, a template, a review, proof of information), the indicator is marked non-conforming. The actual quality of the practice makes no difference.

This rule throws off well-meaning quality managers, and yet there's nothing bureaucratic about it. It's a rule of auditability: the auditor only validates what they can see, date and cross-check. An oral practice, however sound, remains unverifiable. The absence of a document cannot be argued away.

The auditor doesn't ask whether you do your job well. They ask you to prove it.

Two kinds of evidence: existence and implementation

Every piece of Qualiopi evidence falls into one of two categories. Proof of existence shows that the arrangement is planned: a contract template, a welcome booklet, a web page, a questionnaire template. Proof of implementation shows that it actually runs: signed attendance sheets, questionnaires completed by real beneficiaries, dated minutes.

The classic mistake: believing that a blank template covers an indicator. On its own it's worth nothing. It counts only when paired with copies genuinely filled in, on the engagements the auditor chooses, not the ones you would have chosen. Think in pairs, indicator by indicator: a document that proves the arrangement exists, and a record that proves it runs. It's this pair the auditor validates, never the document in isolation.

The date and consistency, the silent key

A piece of evidence can exist and still turn against you. A needs assessment dated after the first session. An attendance sheet with no date. A contract signed once training had already begun. These records don't reassure: they collapse the credibility of the entire file. The auditor cross-checks dates across indicators on the same engagement, systematically. The chronology must tell a story that holds together, from the needs analysis to the final evaluation.

That's why serious organisations fail on indicators they master day to day. The practice was sound, the record was silent on the date, or contradicted the document next to it. Date everything, and check the timeline logic before the audit, not in front of the auditor.

Anticipating sampling and the mechanics of the audit

The duration of the audit isn't negotiated with the certifier: it's set by the schedule in the order of 6 June 2019. A small organisation (less than €150,000 in training turnover, a single site, training action only) is entitled to a one-day initial audit and a half-day surveillance audit.

At the opening meeting, the auditor asks for the list of engagements carried out under Qualiopi and draws a sample from it (for example 2 engagements for a small organisation). The consequence is brutal: conformity isn't decided on a showcase engagement polished over three weeks, but on any randomly drawn file. Each engagement must therefore carry its complete set of dated evidence. Since V9 of the reading guide (in force on 8 March 2024), one blind spot keeps recurring: oversight of subcontractors on CPF, which the principal must now record in black and white.

A one-page evidence guide per indicator

A well-structured file always beats a bulky one. A thick, poorly organised binder, with undated and contradictory files, undermines your credibility and wastes precious time on the sample. A one-page guide per indicator (which document, where it is, its date, the engagement concerned) proves by itself that you have your quality system in hand, and it speeds up the audit.

This meta-document is the most cost-effective tool in the entire preparation. It saves you from rummaging under the auditor's eyes and signals, by its very structure, that you steer your quality rather than endure it. One final point, to put in perspective: a minor non-conformity does not block certification. You have 1 month to propose an action plan and up to 6 months to close the gap. In 2022, more than 98% of organisations audited by ICPF obtained or maintained their certification. The gap feeds continuous improvement, it isn't a guillotine.