# Customer interview analysis

## Decision and scope
- Decision this round must inform:
- Research owner:
- Round dates:
- Current explanation:
- Competing explanation:
- What evidence would change our mind:

## Recruitment and limits
- Participant roles and segments:
- How participants were recruited:
- Relevant groups missing:
- Number interviewed in each segment:
- Limits on what this sample can support:

## Discussion guide
1. Walk me through the last time you completed this task.
2. What happened next, and who was involved?
3. Where did you spend effort or wait?
4. What did you try when it did not work?
5. When does the existing approach work well?
6. Who decides whether the approach can change?

## Evidence table
| ID | Participant ID / segment | Source location | Evidence type | Observation or exact quote | Code | Interpretation | Uncertainty |
| --- | --- | --- | --- | --- | --- | --- | --- |
| E01 | | | Observed / reported / paraphrase / quote | | | | |

## Codebook
| Code | Include when | Exclude when | Example evidence ID |
| --- | --- | --- | --- |
| | | | |

## Findings and counterevidence
| Finding | Supporting evidence IDs | Conflicting evidence IDs | Participants / relevant sample | Missing explanation |
| --- | --- | --- | --- | --- |
| | | | | |

## Decision
- Next action and intended audience:
- Alternatives considered:
- Why this action fits the evidence:
- What remains an assumption:
- Smallest useful next experiment:
- Observation that would stop or change it:
- Owner and review date:

## Data handling
- Participant information and permitted uses recorded:
- Approved storage and access:
- Redactions needed before sharing:
- Retention / deletion review date:

## AI analysis instruction, if used
Use only the supplied, permitted evidence. Preserve participant IDs and source locations. Separate observations from interpretations. List contradictory evidence. Do not invent quotes, participants, findings, or numerical prevalence. Mark missing information as unknown.