Learner Interview Guide
When To Use
You have twenty minutes each with five volunteers and want more than "it was useful". Use this once the follow-up survey has closed and some people have ticked the volunteer question, to prepare a conversation that gets at what they did, what followed and what someone could see.
When Not To Use
If nobody has volunteered, there is no one to interview: a manager's nomination does not count, so go back to Training Follow-Up Survey and its volunteer question. For a conversation with a sponsor before a build, run Sponsor Discovery Questions. To turn finished notes into cases, run Success Case Method Write-Up.
Inputs
- The follow-up survey's answers about use, at group level, and the volunteers' own one-line replies on whether their use worked or stalled
- The critical behaviours from your Critical Behaviours Map, the minutes you have per interview, and who will interview
If you have none of this, I start from what the programme was meant to change and how many volunteers you expect, and mark the output "no agreement on file: first draft".
Approach
This is the interview stage of the Success Case Method, the approach developed by Robert Brinkerhoff: step four, interviewing and documenting cases. The original method looks at the extremes among participants; here that is reframed as uses that worked and uses that stalled, because a case is a use of the training, never a verdict on a person. The questions run in the same order every time so that five conversations can be read side by side. The failure it prevents: the friendly chat that produces "really valuable, I use it all the time" and not one thing anybody could go and look at.
Workflow
- Ask three questions at once: how many volunteers do you have and what did their survey answers say about use, how many minutes per interview, and who will see the notes?
- Set who is asked. Only people who ticked the volunteer question, chosen by what they said about use in their volunteer reply: some where it worked, some where it stalled. Never by performance, seniority, reputation, a manager's suggestion or any personal characteristic. If the stalled side has no volunteers, the guide says so; it does not go looking for them.
- Write the consent script, read aloud at the start: what this is for, that it is voluntary, that they can stop or withdraw, that the write-up is anonymous, who will see the notes, and that nothing is recorded without a yes. A "no" at any point ends it, with thanks.
- Lay out the five questions, in this order every time: what did you do; what followed; what could someone see; what helped; what blocked. Each is tied to a behaviour from the map, in the learner's words. A probe for evidence follows each: "what would I see if I looked?", "is there a record of it?". Opinion is noted as opinion, and a claim nobody could check is marked "reported only".
- Build the notes template keyed by a case letter: no name, team, location or identifying detail. Then the access rule: raw notes stay with the interviewer and the L&D lead, and never go to a manager or the sponsor. A case that cannot be anonymised is left out.
- Size it to your minutes and say which probe to drop first when time runs short. A person interviews. I never write an answer, and I never rehearse what a learner "would probably say" as if it were data.
Output Format
# Learner Interview Guide
Programme: [name] | [minutes] minutes per interview | Interviewer: [role] | Volunteers only
## Who is asked
[Count] volunteers from the follow-up survey: [count] whose answers describe use that worked, [count] use that stalled. Chosen by what they said about use, not by who they are.
## Consent script (read aloud)
[Purpose. Voluntary. You can stop or withdraw at any time. The write-up is anonymous. The notes are seen by [roles] and no one else. Nothing is recorded unless you say yes. Do you agree to go on?]
## Questions, in this order
| # | Question | Probe for evidence | Minutes | Drop first if short |
|---|---|---|---|---|
| 1 | What did you do? | [What would I see if I looked?] | [min] | [yes / no] |
| 2 | What followed? | [Is there a record of it?] | [min] | [yes / no] |
| 3 | What could someone see? | [Where, and how often?] | [min] | [yes / no] |
| 4 | What helped? | [What exactly, and when?] | [min] | [yes / no] |
| 5 | What blocked? | [A tool, time, a process?] | [min] | [yes / no] |
## Notes template, case [letter] (filled by the interviewer; consent given [yes / no]; use [worked / stalled])
| Question | What was said | Evidence (seen, a record, or "reported only") | Opinion |
|---|---|---|---|
| [#] | [your note, no name, team or place] | [evidence] | [opinion, marked as such] |
## Access rule
Raw notes: [interviewer] and [L&D lead] only. Never to a manager or the sponsor. Kept until [date you set], then [what happens].
## Decision
[L&D lead name] confirms who is approached and who may see the notes by [date]; [interviewer name] holds the interviews by [date].
Done When
- The consent script covers all six points and comes before the first question
- The five questions appear in the fixed order, each with a probe for evidence
- The notes template has a case letter and no field for a name, team or location; the access rule names who sees raw notes and says who never does
Quality Bar
- No question leads ("how did the course help you?" assumes it did), and none asks the volunteer to assess a colleague or a manager
- Stalled uses get the same time and the same questions as uses that worked
- Recording, storing notes and how long they are kept are privacy points: check with a qualified adviser
- Interviews are voluntary, with consent and anonymous; they describe uses of the programme, never a verdict on a person, and no answer is invented
Next
Run impact-business-data-request (Business Data Request) to get the figures that the stories cannot give.
About the makers
This pack is made by 90n (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).
1---2name: impact-learner-interview-guide3description: Prepares the guide for short voluntary interviews with learners after a programme, producing a consent script, five questions in a fixed order with probes for evidence, a notes template keyed by case letter with no names, and an access rule for raw notes. Use for "run impact-learner-interview-guide", "learner interview guide", "interview questions after training", "success case interviews", "how to interview training participants", "consent script for learner interviews", "more than it was useful", part of the Claude Guide for Training Impact Pack by 90n.4metadata:5 source: https://skillmd.com/skills/polar-bear-org/impact-learner-interview-guide6---78# Learner Interview Guide910## When To Use11You have twenty minutes each with five volunteers and want more than "it was useful". Use this once the follow-up survey has closed and some people have ticked the volunteer question, to prepare a conversation that gets at what they did, what followed and what someone could see.1213## When Not To Use14If nobody has volunteered, there is no one to interview: a manager's nomination does not count, so go back to Training Follow-Up Survey and its volunteer question. For a conversation with a sponsor before a build, run Sponsor Discovery Questions. To turn finished notes into cases, run Success Case Method Write-Up.1516## Inputs17- The follow-up survey's answers about use, at group level, and the volunteers' own one-line replies on whether their use worked or stalled18- The critical behaviours from your Critical Behaviours Map, the minutes you have per interview, and who will interview19If you have none of this, I start from what the programme was meant to change and how many volunteers you expect, and mark the output "no agreement on file: first draft".2021## Approach22This is the interview stage of the Success Case Method, the approach developed by Robert Brinkerhoff: step four, interviewing and documenting cases. The original method looks at the extremes among participants; here that is reframed as uses that worked and uses that stalled, because a case is a use of the training, never a verdict on a person. The questions run in the same order every time so that five conversations can be read side by side. The failure it prevents: the friendly chat that produces "really valuable, I use it all the time" and not one thing anybody could go and look at.2324## Workflow251. Ask three questions at once: how many volunteers do you have and what did their survey answers say about use, how many minutes per interview, and who will see the notes?262. Set who is asked. Only people who ticked the volunteer question, chosen by what they said about use in their volunteer reply: some where it worked, some where it stalled. Never by performance, seniority, reputation, a manager's suggestion or any personal characteristic. If the stalled side has no volunteers, the guide says so; it does not go looking for them.273. Write the consent script, read aloud at the start: what this is for, that it is voluntary, that they can stop or withdraw, that the write-up is anonymous, who will see the notes, and that nothing is recorded without a yes. A "no" at any point ends it, with thanks.284. Lay out the five questions, in this order every time: what did you do; what followed; what could someone see; what helped; what blocked. Each is tied to a behaviour from the map, in the learner's words. A probe for evidence follows each: "what would I see if I looked?", "is there a record of it?". Opinion is noted as opinion, and a claim nobody could check is marked "reported only".295. Build the notes template keyed by a case letter: no name, team, location or identifying detail. Then the access rule: raw notes stay with the interviewer and the L&D lead, and never go to a manager or the sponsor. A case that cannot be anonymised is left out.306. Size it to your minutes and say which probe to drop first when time runs short. A person interviews. I never write an answer, and I never rehearse what a learner "would probably say" as if it were data.3132## Output Format33```markdown34# Learner Interview Guide35Programme: [name] | [minutes] minutes per interview | Interviewer: [role] | Volunteers only36## Who is asked37[Count] volunteers from the follow-up survey: [count] whose answers describe use that worked, [count] use that stalled. Chosen by what they said about use, not by who they are.38## Consent script (read aloud)39[Purpose. Voluntary. You can stop or withdraw at any time. The write-up is anonymous. The notes are seen by [roles] and no one else. Nothing is recorded unless you say yes. Do you agree to go on?]40## Questions, in this order41| # | Question | Probe for evidence | Minutes | Drop first if short |42|---|---|---|---|---|43| 1 | What did you do? | [What would I see if I looked?] | [min] | [yes / no] |44| 2 | What followed? | [Is there a record of it?] | [min] | [yes / no] |45| 3 | What could someone see? | [Where, and how often?] | [min] | [yes / no] |46| 4 | What helped? | [What exactly, and when?] | [min] | [yes / no] |47| 5 | What blocked? | [A tool, time, a process?] | [min] | [yes / no] |48## Notes template, case [letter] (filled by the interviewer; consent given [yes / no]; use [worked / stalled])49| Question | What was said | Evidence (seen, a record, or "reported only") | Opinion |50|---|---|---|---|51| [#] | [your note, no name, team or place] | [evidence] | [opinion, marked as such] |52## Access rule53Raw notes: [interviewer] and [L&D lead] only. Never to a manager or the sponsor. Kept until [date you set], then [what happens].54## Decision55[L&D lead name] confirms who is approached and who may see the notes by [date]; [interviewer name] holds the interviews by [date].56```5758## Done When59- The consent script covers all six points and comes before the first question60- The five questions appear in the fixed order, each with a probe for evidence61- The notes template has a case letter and no field for a name, team or location; the access rule names who sees raw notes and says who never does6263## Quality Bar64- No question leads ("how did the course help you?" assumes it did), and none asks the volunteer to assess a colleague or a manager65- Stalled uses get the same time and the same questions as uses that worked66- Recording, storing notes and how long they are kept are privacy points: check with a qualified adviser67- Interviews are voluntary, with consent and anonymous; they describe uses of the programme, never a verdict on a person, and no answer is invented6869## Next70Run impact-business-data-request (Business Data Request) to get the figures that the stories cannot give.7172## About the makers7374This 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).