Qualiopi indicator 33: how a CFA builds its evaluation of teaching by apprentices

The decree n° 2026-728 of 1 August 2026 took the French National Quality Framework from 32 to 33 indicators, effective 1 November 2026. The new indicator 33, specific to CFAs and providers delivering apprenticeship training, was already introduced in broad terms: it requires a device for apprentices themselves to evaluate the content and teaching they receive, distinct from the general satisfaction survey. What remains is the most practical question: how do you actually build this tool, without waiting for the first audit under the new framework to discover what an auditor expects?

What indicator 33 requires — and what it does not

The text imposes three cumulative elements: a device to evaluate content and teaching, communication of the results to the teaching teams, and integration into the continuous-improvement process with a periodic measure of its effectiveness. It sets no mandatory format, no minimum frequency and no specific technical platform — as with most indicators in the framework, it is up to the CFA to demonstrate that the device is real and actually used, not to tick an administrative box.

The distinction with indicator 30 — collecting stakeholder feedback is the point auditors will check first. Indicator 30 covers overall satisfaction from all stakeholders (trainees, apprentices, funders, teams) with the service as a whole: welcome, organisation, premises, support. Indicator 33 covers only the pedagogical quality perceived by the apprentice: clarity of content, relevance of methods, and how well what is taught in the centre matches what is actually useful on the job. A single questionnaire cannot cover both without clearly separating which questions belong to which indicator.

Step 1 — Design a questionnaire dedicated to teaching

A credible device fits into around ten targeted questions, not a generic satisfaction form relabelled. Three families of questions cover the expected scope:

• Clarity and pedagogical progression: was the content understandable, was the sequence of notions logical, was the pace suited to the group’s level?
• Relevance of methods: did the exercises, role-plays or materials used help absorb the notion, or was the format unsuited to the subject?
• Fit with the target trade: was what was taught in the centre useful, recognisable or applicable during work placements?

Favour short rating scales (agreement from 1 to 4, with no neutral midpoint that invites dodging) paired with a free-text field per block: a number alone tells a teaching team nothing about what to adjust in a course, while the comment gives them something to act on.

Step 2 — Choose the right rhythm

An apprenticeship programme often runs one to three years: waiting until the end of the programme to evaluate teaching produces an off-topic survey, on content seen months earlier and already forgotten. The most robust practice is to evaluate at the end of each competency block or significant module, building on the breakdown already used to track pedagogical progress. This rhythm has a double advantage: the apprentice answers on fresh content, and the CFA can fix a method or a material before the next cohort hits the same snag — precisely the outcome the decree is aiming for.

Step 3 — Feed the results back to the teams, not just archive them

This is where a file is won or lost at audit. The text does not only require collecting data: it requires that it be communicated to the teaching teams and integrated into a continuous-improvement process, consistent with what indicator 32 on continuous improvement already requires. Concretely, a CFA should be able to show:

• a summary report per block or per cohort, even a brief one;
• evidence that this report was passed on to or discussed with the relevant trainer(s) (email, minutes of a teaching meeting, archived exchange);
• at least one documented example of a concrete adjustment decided as a result of an evaluation (rewritten content, replaced material, reorganised sequence).

A spreadsheet of scores that nobody outside the quality team ever opens does not document the indicator, even if the questionnaires were properly collected.

Step 4 — Measure the device’s effectiveness over time

The decree goes as far as requiring that the effectiveness of the approach be measured periodically: it is not enough to show that a device exists in a given year — you need to show that it produces an effect over time. A simple approach is to compare, from one cohort to the next, how scores evolve on the points that were the subject of a pedagogical adjustment: if rewritten content scores better the following year, the continuous-improvement loop is demonstrated by the figures themselves rather than by a statement of intent.

One caveat is worth flagging before leaning too mechanically on these scores. A meta-analysis by Bob Uttl, Carmela A. White and Daniela Wong Gonzalez, published in 2017 in Studies in Educational Evaluation and covering 97 multi-section studies, shows that ratings learners give their teachers are very weakly correlated with their actual learning outcomes. By contrast, the synthesis of research on feedback conducted by John Hattie and Helen Timperley, published in 2007 in Review of Educational Research, identifies feedback given to — and received by — the learner as one of the most effective levers for learning outcomes (see the study). The practical takeaway for a CFA: a numeric score mostly tells you about immediate impressions, while the qualitative comment and the feedback loop back to the trainer are what actually drive measurable pedagogical progress — that loop is what to document first for the auditor.

Step 5 — Gather the evidence expected at audit

On a recent indicator, an auditor will primarily look for consistency between what the CFA claims and what it can show. A solid file brings together:

• the teaching-evaluation questionnaire or tool, distinct from the general satisfaction questionnaire;
• a collection history covering at least one cohort or one cycle of blocks;
• the summary reports passed on to teaching teams;
• at least one documented example of a corrective action stemming from an evaluation;
• a simple indicator tracking how scores evolve over time.

A CFA with only a few weeks of history at the time of its audit is not at fault if the device is recent and consistent with the 1 November 2026 entry-into-force date: a young but real device beats a history fabricated after the fact for the audit.

Take action

Building a new indicator alongside the existing 32 is easier with an already-structured evidence file. The Complete Kit Certif (€297, 14-day guarantee, documents in French) provides templates and evidence tables for the 32 indicators of the current framework; it does not yet include a sheet for indicator 33, which this guide helps you prepare. If you are launching your CFA or training organisation, the ebook “Create your training organisation in 30 days” (€67) sets up the administrative basics from day one with these changes in mind, and the Complete pack (€347) bundles both resources. Browse all our blog articles to follow Qualiopi regulatory updates as they are published.