Intermediate About 120 minutes 5 stages

Research a new business idea using multiple AI tools

Create a source-led first assessment of a business problem, audience, alternatives and questions requiring real validation.

Research a new business idea using multiple AI tools

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.

Inputs and ownership

Before you start

  • A clearly stated problem, customer and geographic market
  • A dated assumptions list
  • Time to read original sources and speak with potential customers

Interactive workflow

Complete each stage in order.

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

  1. Write the assumptions that would make the idea succeed or fail.

    Deliverable

    A prioritised assumptions register

    Review checkpoint

    Prioritise assumptions by impact and uncertainty.

  2. Locate recent primary and authoritative sources for the market, problem and alternatives.

    Deliverable

    A source pack with publication dates

    Review checkpoint

    Open every source instead of trusting generated summaries.

  3. Build an evidence table that separates facts, estimates, opinions and unknowns.

    Deliverable

    An evidence and uncertainty table

    Review checkpoint

    Do not combine incomparable market figures.

  4. Ask a second assistant to challenge the evidence and expose missing perspectives.

    Deliverable

    A challenge memo

    Review checkpoint

    Look deliberately for disconfirming evidence.

  5. Create a real-world validation plan using interviews, behaviour and willingness to pay.

    Deliverable

    A two-week validation plan with decision thresholds

    Review checkpoint

    Define what evidence would stop or change the idea.

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.

Act as a critical research assistant for this business idea: [IDEA]. The target customer and market are [CUSTOMER/MARKET]. Start by listing the ten highest-impact assumptions. For each, specify evidence needed, suitable primary sources and what would disprove it. Do not estimate market size without a cited method and dated source.

Final review prompt

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

Review the draft against this intended outcome: Create a source-led first assessment of a business problem, audience, alternatives and questions requiring real validation. 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

Use a librarian, industry analyst or specialist researcher for a high-stakes market decision.
Run customer interviews and a small paid demand test before purchasing additional research tools.