Intermediate About 90 minutes 5 stages

Analyze customer feedback with AI

Organize anonymised feedback into themes, evidence and prioritised questions without hiding minority views.

Start the workflow See the prompts
Analyze customer feedback with AI

What you will produce

A reviewable result, not an automatic answer.

Follow the stages in order. Each stage ends with a tangible deliverable and a review checkpoint, so mistakes are caught before they become the next tool’s input.

Expected tool cost

Use free plans where suitable; confirm current provider limits and prices before starting. Compare the latest verified AI pricing before buying another subscription.

Suggested stack

Tools used in this workflow.

The process matters more than the brand. Use an approved equivalent when it offers the controls and output quality your organisation needs.

ChatGPTReview tool profile → ClaudeReview tool profile →

Inputs and ownership

Before you start

  • A lawful, minimised and anonymised feedback dataset
  • A clear decision the analysis should inform
  • A human reviewer who understands the collection method and sample limits

Interactive workflow

Complete each stage in order.

Tick a stage only after its deliverable and review checkpoint are complete.

  1. Remove unnecessary identifiers and document the source, sample and collection period.

    Deliverable

    A privacy-checked dataset note

    Review checkpoint

    Do not upload raw identifiers or account histories.

  2. Define a coding framework before asking the tool to group feedback.

    Deliverable

    A coding framework

    Review checkpoint

    Keep positive, negative and minority views visible.

  3. Group themes while preserving evidence, counts and meaningful exceptions.

    Deliverable

    A theme and evidence table

    Review checkpoint

    Do not imply statistical significance from a convenience sample.

  4. Check counts manually and inspect feedback outside the dominant themes.

    Deliverable

    A manual validation record

    Review checkpoint

    Verify every quoted example against the source.

  5. Present findings, uncertainty and recommended follow-up research separately.

    Deliverable

    A decision-focused insight report

    Review checkpoint

    Separate observed feedback from your interpretation.

Adapt, do not paste blindly

Working prompts for this recipe.

Replace bracketed placeholders, supply approved sources and remove unnecessary personal or confidential information.

Working prompt

Use this only after replacing every bracketed placeholder and removing unnecessary sensitive information.

Analyze the anonymised feedback below to support this decision: [DECISION]. Use these coding categories: [CATEGORIES]. Return a table with theme, count, supporting examples, exceptions and confidence. Preserve minority views. Do not infer demographics, intent or statistical significance. Flag any text that may still contain personal information.\n\nANONYMISED FEEDBACK:\n[PASTE DATA]

Final review prompt

Use as a second-pass checklist, not as approval to publish.

Review the draft against this intended outcome: Organize anonymised feedback into themes, evidence and prioritised questions without hiding minority views. List unsupported claims, missing evidence, ambiguous language, privacy concerns, accessibility issues and actions that still require human approval. Do not rewrite the draft until the issues are listed separately.

Human review required

Verification and cautions

  • Verify names, figures, quotations and source claims before use.
  • Remove confidential or personal information unless the selected service and account are approved for it.
  • Keep a named person responsible for the final decision and published output.

Keep the outcome, change the method

Practical alternatives

Code a small or sensitive dataset manually with two independent reviewers.
Use a governed analytics platform when scale, auditability or statistical analysis is required.

Continue with context

Guidance, tools and implementation help.

Related guideAI privacy basics for business users → Guided learningAI Advantage — Practical AI in 2 Hours → Webz Fusion serviceGet help implementing this workflow →

Related AI recipes

Research a new business idea using multiple AI toolsView recipe → Create a customer FAQ assistantView recipe → Create an evidence-led SEO content briefView recipe →