Use ChatGPT to organise a SWOT, challenge your thinking and turn verified business notes into strategic options. Do not ask it to invent strengths, market opportunities or competitor weaknesses. Supply real operational evidence, research external factors separately, label assumptions, then use the final SWOT to choose a small number of actions with owners, deadlines and measures.

A SWOT matrix is easy to generate. A useful SWOT is harder.

The difference is evidence.

If you ask ChatGPT to list a company’s strengths, weaknesses, opportunities and threats from a short description, it will usually fill the empty spaces with plausible business language. That can look strategic while telling you very little about the actual company.

A better workflow is to use AI as a structured critic. Give it operating facts, customer evidence and researched market information. Then ask it to organise, challenge and prioritise what you already know.

This page restores an older IDJoy SWOT resource that was indexed through the previous query-string blog system. The 2026 version removes hard-coded tool prices and unsupported productivity claims so the prompts remain useful when products and market conditions change.

Start with evidence, not the four boxes

Before opening ChatGPT, collect notes from the business.

Useful inputs include:

  • Sales and enquiry patterns.
  • Customer feedback and objections.
  • Repeat business or churn.
  • Delivery delays and capacity constraints.
  • Skills, equipment, licences or relationships that competitors cannot easily copy.
  • Website and marketing performance.
  • Supplier dependencies.
  • Cash-flow pressure points.
  • Competitors you have actually researched.
  • Regulatory, economic or technology changes supported by current sources.

Separate internal evidence from external research. Strengths and weaknesses should mainly come from the business itself. Opportunities and threats usually need current market evidence.

Prompt 1: Build the evidence register

Act as a strategy facilitator.

I will give you notes about my business. Do not create a SWOT yet.

First sort every note into:
- verified internal fact,
- customer observation,
- external researched fact,
- assumption,
- missing information.

For each assumption or missing item, explain what evidence would make it stronger.

Business notes:
[Paste your notes]

This step catches weak inputs before they become strategic conclusions.

Prompt 2: Identify defensible strengths

Using only the verified internal facts and customer observations, identify potential strengths.

For each strength show:
1. The evidence supporting it.
2. Why it matters to a customer or to delivery.
3. Whether a competitor could copy it easily.
4. How the business currently uses it.
5. What would make the claim stronger.

Reject generic strengths such as "good service" unless the evidence makes them specific.

A strength should help explain why the business performs better, wins trust or delivers something difficult to replace.

Prompt 3: Find weaknesses without turning the exercise into criticism

Review the internal evidence for weaknesses that limit sales, delivery, cash flow, quality or growth.

For each weakness:
- describe the operational problem,
- cite the evidence from my notes,
- explain the business impact,
- rate urgency as high, medium or low,
- suggest one measurable corrective action.

Do not invent problems that are not supported by the information I supplied.

The goal is not to produce the longest list. It is to find constraints that change decisions.

Prompt 4: Research opportunities separately

Opportunities are where many AI-generated SWOTs become unreliable. A model can suggest attractive markets without proving demand exists.

Research first using current sources, then paste the findings into the prompt.

These are external market findings I have verified:

[Paste source-backed findings with links and dates]

Identify opportunities that connect these findings to capabilities the business already has.

For each opportunity show:
- the external evidence,
- the customer or market change,
- why this business could respond,
- what must be tested before investing,
- the smallest useful experiment.

Do not introduce additional market statistics.

A good opportunity is not simply “use AI” or “expand online”. It connects an observable change to a capability and a testable next move.

Prompt 5: Build a threat register

Using the external research I provide, identify threats that could materially affect this business.

Group them into:
- customer demand,
- competition,
- regulation,
- technology,
- suppliers and operations,
- financial exposure.

For each threat state:
1. Evidence.
2. Likelihood: low, medium or high.
3. Potential impact: low, medium or high.
4. Early warning sign to monitor.
5. One practical mitigation.

Do not treat general economic headlines as a business-specific threat unless the connection is explained.

This turns the “T” in SWOT into a monitoring list rather than a collection of scary headlines.

Prompt 6: Create the SWOT matrix

Only after the evidence work is complete should the four-box matrix be generated.

Create a concise SWOT matrix from the approved items below.

Rules:
- Maximum five items per quadrant.
- Every item must trace back to evidence already supplied.
- Keep strengths and weaknesses internal.
- Keep opportunities and threats external.
- Remove duplicate ideas.
- Add an evidence label after every item.

Then list the three strategic tensions that matter most.

Limiting each quadrant forces prioritisation. A matrix with 40 bullets is usually a notes archive, not a strategy.

Prompt 7: Turn SWOT into choices

The useful question is what to do with the matrix.

Use this SWOT to propose strategic choices, not generic recommendations.

Create four groups:
- Use a strength to pursue an opportunity.
- Use a strength to reduce a threat.
- Fix a weakness that blocks an opportunity.
- Reduce a weakness that makes a threat more dangerous.

For each proposed action include:
- the SWOT items it connects,
- the decision required,
- owner,
- first step,
- deadline,
- metric,
- evidence that would tell us to stop or change course.

Rank the final actions by impact and effort.

This is where SWOT starts earning the time spent on it.

Prompt 8: Challenge the final strategy

Act as an independent reviewer.

Challenge this SWOT and action plan for:
- confirmation bias,
- unsupported assumptions,
- strengths that are really basic expectations,
- opportunities that are only trends,
- threats with no clear business impact,
- actions with no owner or measure,
- contradictions between the financial plan and the strategy.

Do not rewrite the plan yet. Give me the questions I should answer before approving it.

A strategy tool should make decisions harder in the useful way. It should expose what the team has avoided answering.

A South African context check

Local context matters, but avoid inserting the same national issues into every SWOT automatically.

For example, electricity reliability, exchange rates, procurement requirements, skills shortages, transport costs or regulatory obligations may be material to one business and irrelevant to another.

Use this prompt:

Review my business model and list the South African external factors that could plausibly affect it.

For each factor:
- explain the direct connection to this business,
- state what evidence I need,
- do not include it in the SWOT until that evidence is supplied.

That keeps local context specific rather than decorative.

How SWOT connects to a business plan

The SWOT should feed decisions in the plan, not sit on a separate page that nobody uses.

For example:

  • A capacity weakness should appear in the operating plan.
  • A customer opportunity should affect the route to market.
  • A supplier threat should affect risk planning.
  • A defensible strength should influence positioning.

The restored ChatGPT business plan prompt guide shows how to carry those decisions into a full planning document without turning assumptions into facts.

What AI should not decide for you

Do not outsource these decisions to a prompt:

  • Whether a customer problem is real.
  • Whether a market is large enough.
  • Whether a competitor is weak.
  • Whether a financial assumption is realistic.
  • Whether a regulatory interpretation is correct.
  • Whether the team is willing and able to execute the plan.

AI can structure the questions and test the logic. The business still has to supply the evidence and make the call.

The result you are aiming for

A useful SWOT should leave the team with fewer priorities than it started with.

You should be able to point to the evidence behind each item, name the owner of the next action and explain what would cause the strategy to change.

If the output only gives you four neat boxes, the analysis is not finished.