Survey results summary with chart suggestions

You are an analyst presenting survey results to [AUDIENCE, e.g. "the leadership team", "our volunteers"]. Summarize the results below.

Survey purpose: [WHAT DECISION THE SURVEY SHOULD INFORM]
How it was distributed and to whom: [e.g. "email to all customers, 9-23 October"]
Number of invitations (if known) and responses: [N INVITED, N RESPONDED]
Questions and answer options:
"""
[PASTE QUESTIONS]
"""
Results (counts or percentages per option, and segments if available):
"""
[PASTE TABLES OR EXPORT]
"""
Open-ended answers (sample or all):
"""
[PASTE TEXT ANSWERS]
"""

Produce:
1. A three-sentence executive summary.
2. Five to seven key findings, each with the exact number, the base (n) and a one-line interpretation.
3. Notable differences between segments, only where the subgroup has at least [MINIMUM N, e.g. 30] responses; label smaller groups as indicative.
4. Three to five themes from the open answers, with one short representative quote each.
5. For each finding, a chart suggestion (type, what goes on each axis, sort order) and a one-line chart title that states the takeaway.
6. Caveats: response rate, self-selection, question wording, anything that limits how far results generalize.
7. Suggested next steps tied to the survey purpose.

Use only the numbers provided. Do not compute statistics I did not give you unless you show the calculation.

Example (excerpt):
Finding 2: 62% of respondents (n=418) rate delivery speed as "very important", more than any other factor.
Chart: horizontal bar chart of "very important" share by factor, sorted descending. Title: "Delivery speed matters most to customers".
Segment note: first-time buyers (n=22) rate price higher - indicative only, the group is small.
Theme: "Packaging waste" (31 mentions) - "Love the food, but the box is half plastic."
Caveat: 418 responses from 5,000 invitations (8%); frequent buyers are likely over-represented.

When to use it

Use this prompt after closing a customer, employee, member or event survey, when you need to share results quickly without overstating them. It works with exported counts from most survey tools and with a sample of the open-text answers.

It is designed for descriptive summaries: what respondents said, how often, and how groups differ. If you need formal significance tests, weighting to match a population, or results that will be published as research, work with someone trained in survey statistics or use an assistant that can run code on the raw data and show its method. The summary from this prompt is still a useful first draft for those cases.

How to customize it

• [AUDIENCE] sets length and tone. Leaders want the summary and next steps; a project team may want all segments.
• [SURVEY PURPOSE] keeps the findings focused on the decision rather than on every question.
• [INVITATIONS AND RESPONSES] allow the model to state the response rate, the most important caveat.
• [RESULTS] should be counts if possible; percentages without a base hide small samples.
• [MINIMUM N] is your threshold for segment comparisons. Thirty is a common rule of thumb, not a statistical guarantee.

Why it works

Survey summaries go wrong when percentages are reported without bases, when a difference between two tiny subgroups becomes a headline, or when open answers are ignored. Every finding here must carry its number and n, and small segments are explicitly labelled.

Chart suggestions with a takeaway title follow a simple rule of data communication: the title says what the reader should notice. Restricting the model to the numbers you supplied, and asking it to show any calculation, keeps the summary checkable.

Tips and variations

• If your assistant can analyze uploaded files, give it the raw export and ask it to compute the tables first.
• For deeper work on open answers, use the thematic analysis prompt.
• Choosing between chart types? Try the chart recommendation prompt.
• Before your next survey, review the questionnaire with our survey design prompt.