Business · Planning & strategy
The Best Prompt for a SWOT Analysis
The best prompt for a SWOT analysis anchors the exercise to a decision and makes the model show where each point came from. This one sorts your facts into internal and external factors, tags every entry as given or inferred, ranks the entries by how much they bear on the decision, and then pairs the quadrants into concrete options. The result is a short grid plus a set of moves to consider, not four lists of generic adjectives.
Evidence-Based SWOT Analysis Prompt
Builds a SWOT around a specific decision, ties each point to evidence you supplied, and converts the grid into strategic options.
- Works with
- ChatGPT, Claude, Gemini
- Difficulty
- Beginner
- Complexity
- Multi-part reply
- Variables
- 5
- Version
- 1.0
Customize this prompt
Fill in what you know. Anything left blank stays as a [bracketed placeholder]. Nothing you type leaves your browser.
5 variables · fields marked * matter most
# Role You are a strategy analyst facilitating a SWOT session. Your job is to keep the analysis honest and specific, and to end with options the team could act on. # Subject - What is being analyzed: [What is being analyzed] - The decision this analysis should inform: [Decision the analysis should inform] - Time horizon: [Time horizon] # Evidence Facts about the organization, product or team: [Internal facts] What I know about the market, customers, competitors and outside conditions: [External facts] # Method Step 1 - Sort. Strengths and weaknesses are internal and within our control. Opportunities and threats are external and would exist whether or not we did. If something I listed is in the wrong group, move it and say why. Step 2 - Fill the grid. At most five entries per quadrant. Each entry must: - be one specific sentence, stating the factor and its consequence for the decision; - end with [GIVEN] if it comes directly from my evidence or [INFERRED] if you reasoned your way to it, with the reasoning in a few words. Step 3 - Rank. Within each quadrant, order the entries by how strongly they bear on the decision. Drop anything that does not bear on it. Step 4 - Combine. Produce a second table with four rows: - Strength + Opportunity: where to press an advantage - Strength + Threat: how to use a strength as a defense - Weakness + Opportunity: what to fix in order to compete - Weakness + Threat: where we are exposed and what to stop or avoid Give one or two options per row, each naming the grid entries it draws on. Step 5 - Conclude. In no more than 120 words, say what the analysis implies for the decision and what remains uncertain. # Rules - No entry may rest on a fact I did not supply unless it is tagged [INFERRED]. Never invent figures, competitor actions or customer opinions. - Reject vague entries such as 'strong brand' or 'economic uncertainty' unless the evidence makes them specific. - If a quadrant has thin evidence, leave it short and say so. Do not pad it to match the others. # Output format A 2x2 grid as a table, then the combinations table, then the conclusion, then 'Evidence I would want' - up to five facts that would most change the analysis.
Use it when you are
- Preparing for a strategy offsite or annual planning session
- Weighing whether to enter a new market, launch a product or drop a service line
- Giving structure to a pile of customer feedback, sales notes and competitor observations
- Checking a SWOT a team has already written for vague or misplaced entries
Why this prompt works
A decision gives the grid a filter
Most weak SWOTs fail because nothing tells the author what to leave out. By making the decision a required input and ranking entries by their bearing on it, the prompt gives the model a test for relevance, and anything that does not pass is dropped instead of decorating the grid.
Given versus inferred
Tagging each entry shows you at a glance which points rest on your own evidence and which are the model's reasoning. Inferred entries are often the interesting ones, but they are also where a model is most likely to be wrong, so they deserve a second look before a meeting.
Internal and external are enforced
Teams routinely file a market trend under strengths or a staffing problem under threats. Step 1 defines the boundary by control, asks the model to move anything misfiled and to explain the move, which is a small lesson in the framework each time you run it.
The grid is not the deliverable
A finished SWOT answers 'what is true' but not 'what should we do'. The combinations table pairs quadrants in the way the TOWS approach does, so the output ends with options that cite the entries behind them and can be argued about directly.
How to use it
- Write the decision as a real choice with a date or scope, such as 'whether to hire a second salesperson this quarter'.
- List internal and external facts separately, as bullet points. Include numbers where you have them and name your uncertainty where you do not.
- Paste the prompt into a fresh conversation and read the [INFERRED] entries first. Delete or correct any that are wrong and ask for the combinations table to be redone.
- Take the 'Evidence I would want' list as research homework, then rerun the prompt with what you find.
- Bring the grid and the options to your team as a starting draft and let them challenge the rankings.
AI model compatibility
| Model | Status | Notes |
|---|---|---|
| ChatGPT | Expected to work | OpenAI's guidance suggests using delimiters to mark distinct parts of the input. Keep the two evidence blocks clearly separated so internal and external facts are not blended. |
| Claude | Expected to work | If your evidence is long, wrap each block in tags such as <internal> and <external>; Anthropic's documentation recommends tags for keeping supplied material distinct from instructions. |
| Gemini | Expected to work | No model-specific adjustment needed. Paste the prompt as written. |
“Expected to work” means the prompt uses plain, model-independent structure and has been reviewed for that — not that it has been run on that model. A model is marked Tested only when a test record is published below.
Worked example
Scenario: An independent bakery with a growing catering business deciding whether to add a second kitchen. Illustrative inputs only.
# Role You are a strategy analyst facilitating a SWOT session. Your job is to keep the analysis honest and specific, and to end with options the team could act on. # Subject - What is being analyzed: The catering arm of Harbor Lane Bakery - The decision this analysis should inform: Whether to lease and fit out a second production kitchen next year - Time horizon: Next 12 months # Evidence Facts about the organization, product or team: Catering is about 30% of revenue and growing; the kitchen is at capacity on Thursdays and Fridays; one part-time salesperson; most catering orders are repeat orders from six law and accounting firms; no delivery vehicle of our own What I know about the market, customers, competitors and outside conditions: Two office towers open nearby next spring; a national sandwich chain now offers boxed lunches with online ordering; flour and butter prices rose this year; several clients have asked for weekend event catering # Method Step 1 - Sort. Strengths and weaknesses are internal and within our control. Opportunities and threats are external and would exist whether or not we did. If something I listed is in the wrong group, move it and say why. Step 2 - Fill the grid. At most five entries per quadrant. Each entry must: - be one specific sentence, stating the factor and its consequence for the decision; - end with [GIVEN] if it comes directly from my evidence or [INFERRED] if you reasoned your way to it, with the reasoning in a few words. Step 3 - Rank. Within each quadrant, order the entries by how strongly they bear on the decision. Drop anything that does not bear on it. Step 4 - Combine. Produce a second table with four rows: - Strength + Opportunity: where to press an advantage - Strength + Threat: how to use a strength as a defense - Weakness + Opportunity: what to fix in order to compete - Weakness + Threat: where we are exposed and what to stop or avoid Give one or two options per row, each naming the grid entries it draws on. Step 5 - Conclude. In no more than 120 words, say what the analysis implies for the decision and what remains uncertain. # Rules - No entry may rest on a fact I did not supply unless it is tagged [INFERRED]. Never invent figures, competitor actions or customer opinions. - Reject vague entries such as 'strong brand' or 'economic uncertainty' unless the evidence makes them specific. - If a quadrant has thin evidence, leave it short and say so. Do not pad it to match the others. # Output format A 2x2 grid as a table, then the combinations table, then the conclusion, then 'Evidence I would want' - up to five facts that would most change the analysis.
What a good response should contain
- A grid with no more than five entries per quadrant, each tied to the kitchen decision and tagged [GIVEN] or [INFERRED]
- Customer concentration in six firms flagged as a weakness, with the reasoning shown if inferred
- A combinations table with options such as pre-selling to tenants of the new towers before committing to a lease
- A conclusion under 120 words that states what the evidence supports and what is still unknown
- An evidence list asking for things like margin per catering order and turned-away order volume
This is an illustrative scenario showing how to fill in the prompt and what to look for. It is not a recorded model output. Recorded outputs will be published here once this prompt has a test record.
Testing and editorial review
Status: Untested. Reviewed for structure and clarity by the editorial team. No recorded model test yet.
No test record has been published for this prompt yet. When it is tested, this section will show the model and version, the date, the inputs, the actual output and notes against our seven-point rubric. We do not publish numerical scores.
| Author | TheBestPromptFor Editorial Team — The editorial team writes, reviews and maintains the prompts and guides on TheBestPromptFor. Prompts are drafted with AI assistance and reviewed against published editorial standards. |
|---|---|
| Published | November 15, 2026 |
| Last reviewed | November 15, 2026 |
| Last tested | Not yet tested |
| Version history |
|
Limitations and things to check
- The analysis cannot be better than the facts you provide. Thin or one-sided evidence produces a thin or one-sided grid.
- Inferred entries are the model's reasoning, not findings. Check them against what you know.
- SWOT is a sorting tool. It does not weigh costs, estimate returns or tell you the probability of a threat.
- A single author's SWOT reflects that author's blind spots. Use the output to start a team discussion, not to replace one.
Alternative prompts
Business
Competitor Analysis Prompt (From Your Own Research)
Organizes the competitor information you have collected into a comparison, a positioning read and a list of what you still need to find out.
- ChatGPT
- Claude
- Gemini
Business
Business Plan Draft Prompt
Turns your idea, customer and known numbers into a structured plan draft, with every unverified figure flagged instead of invented.
- ChatGPT
- Claude
- Gemini
Productivity
Structured Decision-Making Prompt
A two-round prompt: the model questions you first, then weighs your options against your criteria and argues against the one you already favor.
- ChatGPT
- Claude
- Gemini
Marketing
Marketing Strategy Prompt
Turns your business facts into a focused strategy: a diagnosis, one priority segment, positioning, three channels at most, and a phased plan sized to your budget.
- ChatGPT
- Claude
- Gemini
Related resources
Guide
How to Improve AI Responses
A diagnostic approach to weak AI answers: name the fault, trace it to the gap in your prompt, then follow up precisely or rewrite and start again.
Guide
How to Write a Good AI Prompt
The five parts of a complete prompt (goal, context, task, constraints, output format), with a worked rewrite and a pre-send checklist.
Frequently asked questions
Why does the prompt insist on a decision?
Without one, any true statement about the business qualifies as an entry and the grid grows without getting more useful. A decision gives the model a way to rank and discard. If you have no decision yet, use 'where to focus over the next year' as a starting point.
Can the model research my competitors for the external section?
This prompt is built to work from what you supply, because a chat model without live search can state outdated or wrong competitor details with confidence. Gather the facts first, from your own sources, and paste them in. Anything the model adds will be tagged [INFERRED].
What is the difference between a weakness and a threat?
A weakness is internal: something about your organization that you could change, such as a skills gap or a dependency on one supplier. A threat is external: it would exist even if your organization did not, such as a new entrant or a regulatory change.
Sources and references
- Prompt engineering overview — Anthropic
- Prompt engineering guide — OpenAI
Was this prompt useful?
No account needed. Do not include personal information.