Survey results analysis and insights report

Analyze survey responses, from rating scores to open-text comments, into themes, key findings and recommended actions.

Change note: Initial version

Prompt

Analyze the results of this survey and give me findings I can act on.

Survey purpose: {{purpose}}
Respondents: {{respondents}}
Questions and results (scores, counts, and/or open-text comments):
{{results}}

Decision this survey should inform: {{decision}}

Do the following:
1. Data quality first: note sample size, likely response bias, margin of error where I gave enough information, and anything that limits what we can conclude. Do not overstate results from small groups; say "too few responses to say" when appropriate.
2. Quantitative findings: summarize each closed question in one line (use percentages and counts), highlight the 3 strongest patterns, and compare any subgroups I provided. Avoid calling small differences meaningful.
3. Open-text themes: group comments into 4-6 themes, with a count or rough share, a short description, and one representative quote that I gave you. Do not invent quotes.
4. Tensions: where the numbers and comments disagree, or where subgroups want opposite things.
5. Recommendations: 3-5 actions ranked by impact and confidence, each tied to evidence above.
6. Next time: 3 improvements to the survey itself.
7. One-paragraph executive summary.

If a key figure is missing or ambiguous, list the questions at the end rather than assuming.

Variables

• {{purpose}}Survey purposetext

  Default: Customer satisfaction survey after a new onboarding flow launched

• {{respondents}}Who respondedtext

  Default: 212 of 1,400 invited new customers (15% response rate), surveyed 14 days after signup

• {{results}}Resultslong

  Default: Q1 Overall satisfaction (1-5): mean 3.9; 5=38%, 4=31%, 3=15%, 2=9%, 1=7% Q2 'Setup was easy' (agree): 62% Q3 'I got value in the first week' (agree): 48% Open text 'What was most frustrating?': 'importing contacts took forever', 'couldn't find where to invite my team', 'great support but slow import', 'too many emails'

  Paste scores, counts and comments.

• {{decision}}Decision to informtext

  Default: Which part of onboarding to fix first this quarter

Example output

**Data quality**: 212 responses (15%) is a modest, self-selected sample; the margin of error on a 62% result is about +/-6.5 pts, and unhappy or very happy customers respond more. Treat results as directional.

**Quantitative findings**
- Satisfaction is positive: 69% rated 4-5, but 16% rated 1-2.
- Setup ease (62%) outperforms first-week value (48%): people finish setup but do not reach value quickly.
- 14-point gap between ease and value is the strongest pattern.

**Themes (4 comments provided; counts indicative only)**
1. Slow contact import (2 of 4): "great support but slow import".
2. Team invites hard to find (1 of 4).
3. Email volume (1 of 4).

**Tension**: Support is praised while import is criticized, suggesting support is compensating for a product problem.

**Recommendations**
1. Fix import speed or show progress and an estimated time. High impact, medium confidence.
2. Make team invites visible in the first session. Medium impact, low confidence (one mention).
3. Review onboarding email cadence.

**Next time**: Add a "time to first value" question and segment by plan.

**Executive summary**: Onboarding is liked but slow to deliver value. Import is the likely bottleneck; fix it first, and gather more comments before acting on the other themes.

Notes

It separates sample limitations from findings and refuses to invent quotes, which keeps the report honest. Paste more raw comments for better theme counts.