Immediate and delayed training evaluations: the complete guide

Why evaluate your training?

Evaluation is one of the pillars of Qualiopi certification. Indicators 11, 29, and 32 of the Referentiel National Qualite require you to assess the achievement of objectives, collect stakeholder feedback, and use this data to improve your services.

But beyond regulatory compliance, evaluation is a strategic tool. It allows you to measure the actual effectiveness of your training, identify what works and what needs improvement, and demonstrate the value of your services to clients and funders.

This guide details the two main categories of evaluation — immediate and delayed — with practical advice on designing, administering, and leveraging them.

The Kirkpatrick model: the reference framework

The Kirkpatrick model, developed in the 1950s and still relevant today, proposes four levels of training evaluation. It is the most widely used theoretical framework in professional training.

The 4 Kirkpatrick levels

What Qualiopi expects

Qualiopi does not explicitly require the Kirkpatrick model, but the RNQ indicators map directly to it:

• Indicator 11 corresponds to Levels 1 and 2 (assessment of learning and satisfaction)
• Indicator 29 corresponds to Level 1 (feedback collection)
• Indicator 32 corresponds to Levels 3 and 4 (continuous improvement based on results)

In practice, most organizations evaluate at levels 1 and 2 (immediate). Levels 3 and 4 (delayed) are less systematic but highly valued by auditors.

Immediate evaluation: satisfaction and immediate learning

The immediate evaluation takes place right after the training, or within 24 to 48 hours. It covers two dimensions:

Dimension 1: Satisfaction (Kirkpatrick Level 1)

The satisfaction questionnaire measures the learner's perception of the training they have just completed. It covers:

• Content: relevance, clarity, perceived usefulness
• Pedagogy: methods, pace, interactions
• Trainer: expertise, availability, clarity of explanations
• Organization: logistics, materials, physical conditions
• Overall satisfaction: recommendation, overall score

Sample satisfaction questions

Closed questions (Likert scale 1-5):

• "Did the training content meet your expectations?" (1 = Not at all to 5 = Completely)
• "Was the training pace appropriate?" (1 = Too slow to 5 = Too fast, 3 = Just right)
• "Was the trainer clear in their explanations?"
• "Were the practical exercises relevant?"
• "Would you recommend this training to a colleague?"

Open questions:

• "What did you appreciate most about this training?"
• "What could be improved?"
• "What topics would you have liked to explore further?"

Tip: limit yourself to 10-15 questions. Beyond that, the response rate drops and response quality deteriorates. Combine 8-10 closed questions (for statistics) and 2-3 open questions (for qualitative insights).

Dimension 2: Learning outcomes (Kirkpatrick Level 2)

The learning outcome assessment verifies that the learner actually learned what was planned. Unlike satisfaction (which measures perception), outcome assessment measures skills.

Learning assessment formats

Designing a good MCQ

A poorly designed MCQ measures nothing. Some rules:

• One correct answer per question (unless explicitly stated "multiple answers possible")
• 4 options minimum (reduces chance to 25% versus 50% with 2 options)
• Plausible distractors: wrong answers must be believable, not absurd
• No "all of the above": that is a hint, not an assessment
• No double negatives: "Which is not a reason to not..." is cognitive torture

Pre/post comparison

The most compelling method for assessing learning outcomes is the pre-test/post-test: the same quiz is administered before and after the training. The score difference objectively measures progress.

Example:

• Pre-test (placement): 45% correct answers
• Post-test (end of training): 85% correct answers
• Progress: +40 points — tangible proof of learning

Delayed evaluation: transfer and impact

The delayed evaluation takes place 30 to 90 days after the training. Its purpose is to measure what the learner has actually retained and applied in their professional context. It is the most valuable evaluation — and the most neglected.

Why 30 to 90 days?

• Before 30 days: too early. The learner may not yet have had the opportunity to apply what they learned.
• After 90 days: too late. The link between the training and observed results becomes unclear, and the response rate drops sharply.
• The sweet spot: 45 to 60 days after the training.

What the delayed evaluation measures

Sample delayed evaluation questionnaire

1. Since the training, have you had the opportunity to apply the skills you acquired? (Yes, regularly / Yes, occasionally / Not yet / Not applicable)

2. Which specific skills have you used most? (Open question)

3. On a scale of 1 to 5, to what extent has the training had a positive impact on your daily work?

4. Have you encountered obstacles in applying what you learned? (Lack of time / Lack of tools / Lack of management support / Content not adapted enough to my context / No obstacles)

5. In hindsight, would you recommend this training to a colleague? (Yes, without hesitation / Yes, with reservations / No)

6. Do you have any suggestions for improvement? (Open question)

The response rate challenge

The average response rate for delayed evaluations is 20 to 40%, compared to 70 to 90% for immediate evaluations. Here is how to improve it:

• Announce the delayed evaluation during the training: "In 6 weeks, you will receive a short questionnaire to measure the impact of this training. Your feedback is invaluable for our improvement."
• Keep the questionnaire short: 5-8 questions maximum, 3-5 minutes to complete
• Send a reminder: if no response after 5 days, follow up once
• Personalize the email: mention the training title and date
• Offer an incentive: a summary of collective results, a free supplementary resource

On OpenFormations, delayed evaluations are automatically scheduled when the course is created. The email is sent at the interval you choose (30, 45, 60, or 90 days), with an automatic reminder if there is no response.

Measuring training ROI (Kirkpatrick Level 4)

Level 4 — measuring the return on investment of training — is the holy grail of evaluation. It is rarely achieved rigorously, but even an approximation has value.

The simplified ROI formula

ROI = (Benefits attributable to training - Cost of training) / Cost of training x 100

Concrete examples

"Sales Negotiation" training (EUR 1,500/participant):

• Before: proposal conversion rate = 30%
• After: conversion rate = 38%
• Estimated impact: +8 percentage points of conversion = +EUR 80,000 annual revenue for the company
• ROI = (80,000 - 1,500) / 1,500 x 100 = 5,233%

"Time Management" training (EUR 800/participant):

• Before: 2 hours/day in unproductive meetings
• After: 1h15/day (45 min saved per day)
• Impact: 45 min x 220 days x EUR 50/hour (fully loaded hourly cost) = EUR 8,250/year
• ROI = (8,250 - 800) / 800 x 100 = 931%

Caveat: these calculations are estimates. Causal attribution (is the training the sole factor?) is always debatable. But funders and HR directors appreciate the effort to quantify.

The continuous improvement loop

Evaluation only has meaning if it feeds an improvement loop. Qualiopi (indicators 31 and 32) checks that you do not just collect data, but that you use it concretely to improve your training.

The 5-step process

1. Collect: administer immediate and delayed evaluations systematically
2. Analyze: extract trends, identify strengths and weaknesses
3. Decide: define priority improvement actions
4. Implement: apply changes in subsequent sessions
5. Verify: measure the impact of changes in following sessions

How to document continuous improvement

For the Qualiopi audit, you must be able to show an improvement history. Here is a simple, effective format:

This table is valuable audit evidence: it shows that your evaluations are not a formality but a management tool.

Digital vs. paper evaluations

Paper: on its way out

Paper questionnaires had their place when training was exclusively face-to-face. Today, they have too many drawbacks:

• Manual data entry (source of errors)
• Bulky physical archiving
• No automated analysis
• Impossible for remote training
• Long processing times

Digital: the standard

Digital evaluations offer decisive advantages:

• Results available in real time
• Automatic analysis (averages, trends, comparisons)
• Unlimited archiving, instantly searchable
• Automatic sending (immediate and delayed)
• Automatic reminders to improve response rates
• Data exportable for BPF and Qualiopi audits

The transition

If you still use paper questionnaires, the transition to digital is straightforward:

1. Reproduce your paper questionnaires in a digital tool (Google Forms at minimum, or directly in your LMS)
2. Test with one session
3. Compare response rates and data quality
4. Roll out across all courses

Mistakes to avoid

Mistake 1: Evaluating only satisfaction

Satisfaction (Kirkpatrick Level 1) is the easiest level to measure but the least informative. A learner can be very satisfied with a training and have learned nothing. Conversely, a demanding training can get average satisfaction scores while being very effective. Always combine satisfaction AND learning outcomes.

Mistake 2: Questionnaires that are too long

Beyond 15 questions, completion rates drop and responses become careless. Be concise. Every question must have a clear purpose.

Mistake 3: Not using the results

Collecting evaluations without analyzing them is worse than not collecting them. The Qualiopi auditor will look for evidence that you use results to improve your training. A folder full of unanalyzed questionnaires is a negative signal.

Mistake 4: Ignoring negative feedback

Negative feedback is the most valuable. Resist the temptation to dismiss or minimize it. A 2/5 score on a module is actionable information. Dig in, understand, correct.

Mistake 5: Forgetting the delayed evaluation

The delayed evaluation measures the real impact of your training. It is also the one Qualiopi values most (indicator 32). Not doing it means depriving yourself of the most powerful indicator of your effectiveness.

Summary: your complete evaluation system

Questions? Contact us at ${CONTACT_EMAIL}.