← Guides

Survey Report Example and Template: What a Good Report Contains

Most advice about survey reports describes what to do. This page shows one. Below is a complete report, section by section, built from 144 real responses to a customer experience survey for a fictional coffee chain, with a note on why each part is written the way it is. At the end there is a fill-in template you can copy. For the reasoning behind the structure, see how to present survey results to leadership.

What sections does a survey report need?

Seven, in this order. The first five fit on one page, and that page should be enough to make the decision.

  1. Context: Who was asked, how many answered, and why the survey was run.
  2. Headline numbers: The one or two scores everyone will remember, with what they are made of.
  3. Executive summary: One paragraph a busy reader can stop after.
  4. Key findings: Three to five, each with a number, its base and a quote.
  5. Recommended actions: What to do, who owns it, and how you will know it worked.
  6. Method and limitations: What the numbers can and cannot support.
  7. Appendix: One chart per question, in survey order, for anyone who wants to check.

What does a good survey report look like?

Here is each section for the Coastal Brew Coffee Co. customer experience survey. The numbers are real responses to the demo survey behind our sample report; the business is made up.

1. Context

Example
Coastal Brew customer experience survey. 144 responses from customers of three locations and our delivery service, collected through a link on receipts in August. We ran it to decide where to spend this quarter's operations budget.

Three facts in two sentences: who answered, how many, and the decision the survey serves. If you know how many people were invited, add the response rate too. A reader who skips everything else still knows how much weight the report can bear.

2. Headline numbers

Example
Net Promoter Score: -33. 27 promoters, 43 passives and 74 detractors, from 144 responses.
Customer satisfaction: 71%. An average rating of 3.5 out of 5, from 144 responses.

A score on its own invites the question "out of what?" Showing the three NPS groups lets the reader see what -33 is made of: just over half of respondents are detractors. Two headline numbers is plenty. More than that and none of them stick. For how each score is calculated, see NPS vs CSAT.

3. Executive summary

Example
Customers like what Coastal Brew makes and dislike how long it takes to get it. 63% (90 of 144) say coffee quality is what they value most, and 68% (98 of 144) rate their drink or food 4 or 5 out of 5. But the Net Promoter Score is -33, with 74 of 144 respondents unlikely to recommend us. Speed of service is the weakest-rated part of the visit, and more seating is the improvement customers ask for most. We recommend adding peak-hour staff and seating before the next survey round, and measuring whether speed ratings move.

It opens with the whole story in one sentence, backs it with the two numbers that carry it, and ends with the decision. Every figure shows its base, because the questions behind them were answered by different numbers of people.

4. Key findings

Example

Speed of service is the weakest part of the visit. 34% (49 of 144) rate it poor or fair, compared with 19% (28 of 144) for staff friendliness. In their words: "Please fix the morning line, I've missed my train twice waiting."

Space is the improvement customers want most. Of the 78 who chose improvements, 49% (38 of 78) picked more seating, ahead of faster service at peak hours at 31% (24 of 78).

Customers are split on value for money. 25% (27 of 107) rate it 1 or 2 out of 5 and 30% (32 of 107) rate it 4 or 5. Its average, 3.0, is the lowest of any 1 to 5 question. In their words: "The prices felt high for what I got. I expected more for what I paid."

Staff are an asset, but stretched. 44% (64 of 144) rate staff friendliness very good or excellent. In their words: "Friendly, just stretched too thin."

Each finding is a full sentence someone could repeat in a meeting, followed by its evidence: a share with its base, a comparison that gives the share scale, and a quote that shows what it sounds like. The value finding reports both ends of the scale, because an average of 3.0 would have hidden a split. For turning written answers into findings like these, see how to analyze open-ended survey responses.

Example
  1. Add a second register on weekdays from 07:00 to 09:00 for a two-week trial. Owner: operations lead. Addresses finding 1. Measured by the speed rating in the next survey round.
  2. Add counter seating at the busiest location this quarter. Owner: store manager. Addresses finding 2. Measured by the share asking for more seating.
  3. Test a loyalty card for regulars before changing prices. Owner: marketing. Addresses finding 3. Measured by value-for-money ratings among weekly visitors.

Every action names an owner, links back to the finding it answers, and says how you will know whether it worked, which is usually the same question in the next round of the survey. Not every finding needs an action. The staff finding is something to protect, not fix.

6. Method and limitations

Example
Most questions were optional, so the number of answers varies from question to question, and each figure shows its own base. Respondents chose to answer a link on their receipt, so very happy and very unhappy customers may be over-represented.

Stated plainly and briefly, limitations make the rest of the report more credible, not less. They also stop someone reading a 49% from 78 people as if it came from all 144. For how many responses a figure needs before it means much, see survey sample size.

7. Appendix

One chart per question, in the order the survey asked them, each with its number of responses. Nobody reads it front to back. It is there so anyone who doubts a finding can check it. You can see the full appendix for this survey, every chart and every written answer, in the live sample report.

SurveyService writes the executive summary, key findings and recommended actions for you, with every figure computed from your responses rather than estimated, and builds the appendix of charts automatically. See the sample report →

Survey report template

The same structure as a fill-in outline. Copy it into a document and replace the brackets.

Report outline
SURVEY REPORT: [Survey name]

1. CONTEXT
Who was asked: [audience], via [email / link / QR code]
Responses: [n] of [number invited] ([response rate]), collected [dates]
Why we ran it: [the decision this survey is meant to inform]

2. HEADLINE NUMBERS
[Metric, e.g. NPS]: [score] ([breakdown, e.g. promoters / passives / detractors], from [n] responses)
[Metric, e.g. CSAT]: [score] (from [n] responses)

3. EXECUTIVE SUMMARY
[One paragraph: what we learned, the biggest number behind it, and the decision this report asks for.]

4. KEY FINDINGS
Finding 1: [The finding as a full sentence]
Evidence: [x% (n of base)], compared with [a second figure that gives it scale]
In their words: "[a representative quote]"

[Repeat for 3 to 5 findings, most important first.]

5. RECOMMENDED ACTIONS
[Action] | Owner: [name] | By: [date] | Addresses: finding [n] | Measured by: [how]

6. METHOD AND LIMITATIONS
[Optional questions and smaller bases, who may be missing from the responses, anything that changed since the last round.]

APPENDIX
[One chart per question, in survey order, each with its number of responses.]

How should the report change for different kinds of survey?

The seven sections stay the same. What goes in them shifts with the audience.

  • Customer feedback: lead with NPS or CSAT and break it down by whatever you can act on separately, such as location, product or channel. The example above is this kind.
  • Employee engagement: compare with the last round rather than with other companies, and never report a group so small that answers could be traced to a person. Many organizations set a minimum group size, often five, below which results are combined. Be careful with quotes for the same reason. See the employee engagement survey template.
  • Market research: the context section matters most, because who answered decides who the findings describe. Say how respondents were found and how they differ from the market you care about before reporting any percentage. See the market research survey template.

Frequently asked questions

What should a survey report include?

Context (who was asked and how many answered), the headline numbers, a one-paragraph executive summary, three to five key findings with evidence, recommended actions with owners, a short note on method and limitations, and an appendix with a chart for every question.

How long should a survey report be?

Aim for one page that carries the decision: context, headline numbers, executive summary, findings and actions. Everything else, including the full set of charts, goes in an appendix. Most readers stop after the first page, so it has to stand on its own.

What is the difference between survey results and a survey report?

Survey results are the raw material: response counts, charts and answers for every question. A survey report is an argument built from them. It picks the findings that matter, explains them, and says what should happen next.

Should a survey report include individual responses?

Include a few short, representative quotes to show what a finding sounds like in respondents' own words, but not raw responses in bulk. Check that no quote identifies the person who wrote it, especially in employee surveys and small groups.

Keep reading

SurveyService supports scale, NPS, CSAT, matrix, multiple choice, checkbox and free-text questions, with an AI summary that reads them all together. Try it free.