Write a research report that supports a decision

Download the editable outline

Download the consumer research report template (Markdown). It opens in a text editor and can be adapted in a document. This is a writing aid, separate from the Excel workbook and CSV outputs described on the product page. It is not an example of an automatically generated application report.

Use six sections

1. Decision and scope: State what the research was meant to inform and what it cannot establish.
2. Method and source: Identify where the answers came from, audience, dates, questions, exclusions and analyzed bases.
3. Findings: Show the count or distribution before interpreting it. Keep a reference to the underlying output.
4. Recommendation: Explain the proposed action and the evidence against it as well as for it.
5. Next evidence: Name the remaining uncertainty, next test, owner and timing.
6. Review: Check labels, calculations, limitations and sharing permissions.

Keep the core report short enough to read, but retain the exact questionnaire and permitted supporting data. A chart pasted into a presentation still needs its source and base.

A completed demonstration, not a customer outcome

The following example uses the site’s constructed meal-kit demonstration. It is not a research run, not actual consumer feedback and not a business success story.

Decision: Illustrate how a team might assess whether a convenience benefit deserves follow-up.

Method: Constructed demonstration data with 250 records. The statement is “This meal kit would make weeknight dinners easier.” The full example CSV is available. No real participants were recruited for this demonstration.

Finding: 65 records strongly agree and 90 somewhat agree. Combined agreement is (65 + 90) ÷ 250 = 62%. The age-group illustration shows 90 of 125 agreeing among ages 25–34 and 65 of 125 among ages 35–44, or 72% and 52% respectively.

Interpretation: This demonstrates how an agreement distribution and group comparison can be summarized. Because the data were constructed, it supplies no evidence that a real population prefers the offer.

Recommendation: Do not make a market decision from these demonstration numbers. For a real project, define the offer and collect appropriate evidence about the convenience benefit from the intended audience.

Next step: Assign a research owner to prepare that brief, choose the evidence source and obtain any necessary recruitment quote. No test has been scheduled or commissioned by this example.

Separate observation, interpretation and action

“62% of these demonstration records agree” is a calculation. “Convenience matters to buyers” is an interpretation that this dataset cannot establish. “Launch the meal kit” is an action far beyond the evidence. Keeping those sentences separate makes overreach easier to spot.

For an actual simulated study, use the same discipline: describe the generated result as generated. For a human study, disclose the recruitment and design limits. Do not rewrite a close or contradictory result into a confident winner merely to simplify the executive summary.

Review before sharing

Recompute the headline numbers, inspect exclusions and keep full category labels. A percentage without a valid base can mislead when people compare groups or questions. See the crosstab guide for the worked calculation above and the naming-study article for a separate example with conflicting measures.

Check that the recommended commitment matches the strength and source of the evidence. Keep confidential briefs and identifiable response data out of public examples unless you have the appropriate permission. Assign responsibility for the next test rather than ending with an unsupported statement that more research is needed.