# Impact Training Follow Up Survey

> Drafts the short survey sent some weeks after a programme, producing three to five items on what people used and what happened, open questions on what helped and what got in the way, a volunteer question for interviews, a send plan and a group-level summary template. Use for "run impact-training-follow-up-survey", "training follow-up survey", "did anyone use the training", "level 3 survey", "survey a month after the course", "training application survey", "behaviour change survey", "find people to interview after training", part of the Claude Guide for Training Impact Pack by 90n.

- Skill: `polar-bear-org/impact-training-follow-up-survey` (Agent Skill)
- Install (CLI): `npx get-skill add polar-bear-org/impact-training-follow-up-survey`
- Raw SKILL.md: https://api.skillmd.com/api/skills/polar-bear-org/impact-training-follow-up-survey/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: polar-bear-org (https://skillmd.com/u/polar-bear-org)
- Updated: 2026-11-14
- Page: https://skillmd.com/skills/polar-bear-org/impact-training-follow-up-survey

---


# Training Follow-Up Survey

## When To Use
The course ended a month ago and nobody has asked whether anyone used it. Use this when people have had time to meet the situations the programme prepared them for, and you need to know what they did with it, what followed and what got in the way.

## When Not To Use
On the last day of the programme nobody has used anything yet; run Post-Training Survey for readiness and expected barriers. If you need to know what shows up in the work as opposed to what people report, run Behaviour Observation Checklist.

## Inputs
- The critical behaviours from your Critical Behaviours Map, in the words learners would recognise
- The weeks after the programme set in your data plan, the number of people you will invite, and the minimum reporting group
If you have none of this, I start from the two or three things people were meant to do differently, as you describe them, and mark the output "no agreement on file: first draft".

## Approach
This is Level 3 application data, collected once people have had time to apply what they learned, as set out by Kirkpatrick Partners and in the ROI Institute's methodology. The items follow Pew Research Center's question-writing rules. The survey is also the third step of the Success Case Method, the approach developed by Robert Brinkerhoff: a survey to find cases. The judgement is in one distinction most follow-up forms miss: "I did not use it" and "I never had the occasion" are different facts, and the second is about the work, not the person. Blur them and a programme launched ahead of the situation it serves looks like a failure of the learners.

## Workflow
1. Ask three questions at once: which critical behaviours should the survey cover, how many weeks after the programme will it go out and to how many people, and what minimum reporting group do you set?
2. Write one closed item per behaviour, three to five in all. Each offers the same four concrete states: used it and it worked; used it and hit a problem; had an occasion and did not use it; no occasion yet. If there are more behaviours than five, you choose which matter most; I do not pick for you.
3. Add three open items: what happened as a result, what helped, what got in the way. Ask for what the person did and saw, not for an opinion of the course; that was the last-day form's job.
4. Put the volunteer question last: "Would you talk to us for [minutes] about how you used it?" The contact route sits apart from the answers (a separate form or a reply to a named person) and asks the volunteer for one line on whether their use worked or stalled, so interviews can be picked by use while the survey itself stays anonymous for everyone.
5. Set the send plan: the send date, one reminder, the close date, and who sends it. It goes to the learners themselves. It is never sent to managers to answer about their people.
6. Build the summary template: counts per state for each behaviour, with the number invited beside the number who answered, for groups at or above the minimum. Nothing is filled in for people who did not answer. The limit is written on the summary itself: this is self-report, so it is "suggested" evidence at most, and the people who answer a survey weeks later are not the whole group.

## Output Format
```markdown
# Training Follow-Up Survey
Programme: [name] | Sent: [weeks you set] weeks after | Anonymous | About how the programme is used, never about you
## Use items
| # | Behaviour (from the map) | Item as the learner reads it | Answer options |
|---|---|---|---|
| 1 | [critical behaviour] | Since the programme, [behaviour in plain words]? | Used it and it worked / Used it and hit a problem / Had an occasion and did not use it / No occasion yet |
## Open items
What happened as a result? | What helped you use it? | What got in the way?
## Volunteer question
Would you talk to us for [minutes] about how you used it? If yes, [separate contact route], with one line on whether it worked or stalled for you. Your answers above stay anonymous either way.
## Send plan
| Send date | One reminder | Close date | Who sends (role) | Channel | Who exports |
|---|---|---|---|---|---|
| [date] | [date] | [date] | [role] | [your survey tool] | [role] |
## Group summary template
| Behaviour | Invited | Answered | Worked | Hit a problem | Occasion, not used | No occasion yet | Reported or suppressed (minimum [you set]) |
|---|---|---|---|---|---|---|---|
| [behaviour] | [your figure] | [your figure] | [count] | [count] | [count] | [count] | [reported / suppressed] |

Limits: self-report from those who chose to answer: "suggested" at most. Volunteers: [count worked, count stalled, from their own line], held by [role], apart from the answers.
## Decision
[L&D lead name] approves the items and the send plan by [date]; [named person] sends on [date] and vouches for the invited and answered counts.
```

## Done When
- Every closed item names one behaviour and carries the four states, with "no occasion yet" kept separate
- The volunteer question is last and its contact route is apart from the answers
- The send plan has a send date, one reminder and a close date, and the summary shows invited beside answered with the self-report limit written on it

## Quality Bar
- No item asks one person to rate another, and no item asks about readiness or satisfaction
- No expected response rate is quoted anywhere; you record what came back
- Volunteers come only from people who tick the question; no manager nominates anyone
- How volunteers' contact details are stored, and for how long: check with a qualified adviser
- Answers describe use of the programme, never a person; results are reported for groups, and nothing is filled in for people who did not answer

## Next
Run impact-behaviour-observation-checklist (Behaviour Observation Checklist) to see whether the reported use shows up in the work.

## About the makers

This pack is made by 90n ([90n.ai](https://90n.ai)): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).

