How it works
The satisfaction survey is an activity like any other: you add it to a course, at the point where it should appear. The learner answers it, and the platform derives two indicators from it that it can then display everywhere: the satisfaction rate and the NPS.
The fact that it is an activity, and not a separate module, has immediate practical consequences: the survey follows the same rules as the rest of the course — its position, its prerequisites, its access window, and the tracking of who has completed it.
Immediate or delayed: the same activity
It is the access window that makes the difference, not a dedicated setting.
- Immediate survey — place the activity at the end of the course and leave the window open. The learner reaches it naturally after the last activity.
- Delayed survey — add a second survey activity, and give it a later opening date (three months later, for example). It will stay invisible until then, and then open on its own.
Both surveys coexist in the same course. Their answers are combined in the course indicators — which is what you expect from a "course" figure — while remaining viewable separately on the results screen of each one.
Write the questionnaire
When you create it, five questions are suggested: overall rating, achievement of objectives, quality of delivery, recommendation, free comment. These are not mandatory questions — you can edit them, delete them, add more. They are there because a blank page in front of a questionnaire editor is the best way never to have a survey at all, and because these five produce the two indicators and cover Qualiopi's expectations.
Five answer types, deliberately few:
| Type | What the learner sees | What it feeds |
|---|---|---|
| Scale from 1 to 5 | Five named levels, from "Very poor" to "Very good" | The satisfaction rate |
| Recommendation from 0 to 10 | A row from 0 to 10 | The NPS |
| Yes / No | Two buttons | A breakdown |
| Choice from a list | Your options, one per line | A breakdown |
| Free text | A text box | Verbatim comments, never aggregated |
The levels of the scale are named and not just numbered: "2" means nothing without its scale, and two learners do not put the same meaning behind the same number.
Edit a survey that is already running
You can correct the wording of a question, change its type, insert a new one in the middle or reorder the whole set: answers already collected stay attached to their question. Each question has its own identity, independent of its position.
However, deleting a question deletes its answers — there would be no point in keeping figures without the question text needed to read them. The editor asks you first.
Answers not shown with names
A single box, ticked or not, with a concrete consequence: when it is ticked, no name appears on the results screens, not even on the free comments.
Let us be precise about what this means, because the learner reads the same thing: the platform still records who answered. This is what makes it possible to calculate the response rate, prevent a double submission and pass the activity. So it is not complete anonymity, and we would rather write that than promise the opposite.
Both settings have their uses:
- names hidden for a satisfaction survey — you want honest answers, and a learner who knows they can be identified goes easy on their instructor;
- names shown for a needs assessment or a request for support — here, knowing who answered what is the whole point.
Read the results
Open the activity as an instructor: you go straight to the results, never to the form. (Your own answers would skew the session statistics: they are not collected.)
Three indicators at the top of the screen:
Response rate — how many enrolled learners have answered. This is the indicator to look at first: a satisfaction rate calculated on three people says nothing about a session of forty.
Satisfaction rate — the share of scale answers at "Good" or "Very good". It is not the average turned into a percentage, and the difference matters: a survey where everyone answers "Fair" is worth 50% on average and 0% in satisfaction. The second reading is the one that tells the truth at a steering committee. The average is still shown alongside, question by question.
NPS — the percentage of promoters (9–10) minus the percentage of detractors (0–6). It ranges from −100 to +100, and it is not a percentage: that is why it is shown without a "%" sign. An NPS of 30 is good, whereas 30% satisfaction would be poor. The two therefore have different scales, shown under each figure.
Each indicator carries its rating — "72% — Good" — with the scale applied just below it. An indicator that cannot be challenged is not an indicator.
The detail follows, question by question: a breakdown bar, the average, the number of answers, and for free-text questions the list of comments.
Export the answers
An "Export answers (CSV)" button at the bottom of the results screen produces a file that opens directly in a spreadsheet: one row per respondent, one column per question. Values are written out in full — "Very good" rather than "5", because the file travels without its scale.
What the file contains depends on the survey setting:
- survey with names — last name, first name, email address and date of answer, then the answers; rows are sorted by name;
- answers not shown with names — no name, no date, and the order of the rows is randomised.
This randomisation deserves a word. Without it, the rows would come out in the order of the account IDs, which roughly follows the order in which the accounts were created: someone who knows how long each member of the group has been there could match the answers to the people. A promise of anonymity that the file silently undoes would be worthless.
In reports
The same indicators are available as variables in your PDF report templates, at both session and course level:
| Variable | Outputs |
|---|---|
@m_session_satisfactionRate@ | The satisfaction rate, for example "72%" |
@m_session_satisfactionLevel@ | Its rating: Low / Average / Good |
@m_session_npsScore@ | The NPS, for example "+34" (without %) |
@m_session_npsLevel@ | Its rating |
@m_session_surveyResponseRate@ | The response rate |
@m_session_surveyRespondents@ | The number of respondents |
The same variables exist as @m_course_…@ for a course-level report.
> With no survey in the course, these variables output a dash, never 0. A course without a survey does not have zero satisfaction: it has no measurement. Displaying 0 would suggest a failure where there is nothing — and a report handed to a funder would carry a disastrous score that nobody gave.
For your quality review, the ready-made template dgtconcept - Satisfaction consolidée (CSAT et NPS) brings all your courses together: overall CSAT and NPS, month-by-month trend, then the detail by course and by session with the response rate.
What else you need to know
- The learner can come back and change their answer as long as the activity remains accessible. Their submission replaces the previous one rather than adding to it: one person counts as one vote. Freezing the answer at the first submission would not have protected any statistic — an average built on acknowledged mistakes is worthless.
- Answering passes the activity. There is nothing to mark and nothing to wait for from anyone.
- Mandatory questions are marked with an asterisk, and submission is refused while they are unanswered. Use them sparingly: one mandatory question too many makes people abandon the whole questionnaire.
- The introduction text is optional but worthwhile: saying what the answers will be used for noticeably increases the participation rate.