Home · Business · IT, Data & Security · Data analysis & reporting
As of 13 August 2026, AI can analyse your NPS survey results.
This still needs a person who signs their name to it.
Can you do it?
5 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0
Skill neededchat-fluent
Who has to check ityou
What the alternative costsA spreadsheet or data analyst can produce the alternative report; no price is supplied in the available tool data.
If this goes wrong: you act on a misleading segment or theme and spend time and money fixing a customer problem that the survey did not establish.
What to actually do
Hand it to a person
The route this page recommends
A person who owns the outcome does this end to end, worth it when the failure is dear.
Use a tool built for this
Second choiceDo it yourself
The distant thirdA chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.
How to actually do it
- Open the survey export in Excel or Google Sheets and remove names, email addresses, free-text personal details and any columns that are not needed for analysis.
- Check that each row represents one response, the recommendation rating is on the 0 to 10 scale, and segment labels such as product, date or region use consistent spelling.
- Add or retain a response identifier, response date, NPS rating, segment columns and comment text, then save the cleaned file as a CSV or spreadsheet.
- Open ChatGPT, Claude or Gemini and paste the prompt, then upload the cleaned survey file.
- Ask the chatbot to show the overall NPS calculation and compare its response count, promoter count, passive count and detractor count with a pivot table or manual count in your spreadsheet.
- Compare each segment table and chart with spreadsheet filters or pivot tables, and inspect the quoted comments to confirm that the themes and examples match the source responses.
- Edit the report to remove unsupported recommendations, disclose sampling and response limitations, and send the checked version to the colleague responsible for customer insight or service improvement.
Prompt
Analyse the attached NPS survey data for a UK business. Use only the data supplied and do not invent figures, respondents, segments or explanations. Treat ratings from 0 to 6 as detractors, 7 to 8 as passives, and 9 to 10 as promoters, then calculate NPS as the percentage of promoters minus the percentage of detractors. Show the response count, missing or invalid ratings, counts and percentages for each rating group, and the full calculation so I can check it. Break the results down by each available segment, such as date, product, region or customer type, but do not draw conclusions from a segment with too few responses without saying why. Analyse open-text comments separately: identify recurring themes, give the number of comments supporting each theme, quote only short anonymised examples, and distinguish clearly between what respondents said and your interpretation. Flag duplicate rows, inconsistent labels, leading questions, sampling gaps and other limitations you can see. Create a concise report with: executive summary, data-quality checks, overall NPS, segment findings, comment themes, limitations and practical next steps. Keep recommendations tied to evidence in the file. End with a list of every figure and conclusion that a human should verify before sharing the report. Do not include names, email addresses or other personal data in the output.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot know whether the respondents represent your wider customer base or whether non-response has biased the result.
- AI groups comments by wording and may miss the operational meaning of a complaint that is obvious to your customer-facing team.
- AI cannot decide which NPS movement matters commercially without your targets, survey design, customer economics and business context.
- AI may produce plausible segment conclusions from small or uneven groups unless you check the underlying counts.
- AI cannot establish that a change in NPS was caused by a particular product, policy or service change.
Even on a YES, the friction has a name: judgement under ambiguity, context depth and verification cost.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 2 |
| Verification | 1 |
| Liability | 2 |
| Effort delta | 2 |
| Total | 9 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can AI calculate my NPS?
- Yes. Give it the 0 to 10 ratings and ask it to show the promoter, passive and detractor counts, percentages and full calculation. Check those figures against a pivot table or formula in your spreadsheet.
- Can AI analyse NPS comments?
- Yes, it can group open-text comments into recurring themes and draft a summary. It cannot reliably judge the importance or business meaning of a theme without your customer and operational context.
- Is it safe to upload customer survey data to AI?
- Remove names, email addresses, contact details and identifying free text before uploading anything. Check your employer's data policy and the tool's handling terms, and use an approved workspace for business data.
- Can AI write an NPS report for me?
- Yes, it can turn checked calculations, segment findings and comment themes into a report with charts and recommendations. You still need to verify every figure, confirm that the sample supports the conclusions and approve what is shared.
Nearby answers
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.