Brainstorming and strategy prompts

You are stuck on a problem and every idea in the room sounds like the last one. Use this to force variety before you judge anything.

Act as a creative strategist who generates unusual, practical options and then judges them like a skeptical operator.

The problem or opportunity: [THE PROBLEM OR OPPORTUNITY]
My business or team, in a few sentences: [YOUR BUSINESS OR TEAM]
Constraints (budget, time, people, brand, legal): [CONSTRAINTS]
What we have already tried and what happened: [WHAT HAS ALREADY BEEN TRIED]

Phase 1, no judging: produce 30 ideas, five in each of these buckets:
- Cheap and fast (doable this week)
- Bold (would make a competitor nervous)
- Borrowed from a different industry
- Subtract something (remove a step, a feature, a fee, a rule)
- Partner or piggyback on someone else's audience
- Automate or systematize something manual

One line each. No repeats of what we already tried unless you change it materially.

Phase 2: pick the five strongest. For each, give the reasoning, the biggest risk, the cheapest way to test it within two weeks, and a rough effort rating (low, medium, high).

End by asking me which two I want to develop further.

Tip: If the first 30 feel obvious, reply 'those are the ideas everyone has; give me 15 that a cautious competitor would never try' and watch the quality jump.

You need a quick strategic snapshot of a company, product, or team before a decision, and you want it honest about what is known versus guessed.

Act as a strategy consultant who refuses to write a SWOT full of platitudes.

What I am analyzing: [COMPANY, PRODUCT, OR TEAM]
Industry and market: [INDUSTRY]
What I know about customers, competitors, results, and trends (paste notes, numbers, feedback): [PASTE WHAT YOU KNOW]
The decision this analysis needs to inform: [DECISION YOU ARE TRYING TO MAKE]

First, ask me up to four questions whose answers would change the analysis most. Wait for my replies.

Then build the SWOT with these rules:
- Every item is specific and tied to something I told you or a clearly stated inference. Label each item Fact (from my input) or Assumption (your inference), and say what evidence would confirm the assumptions.
- No more than five items per quadrant, ranked by how much they matter to the decision.
- Strengths and weaknesses are relative to named competitors or alternatives, not absolute.
- Threats include what happens if we do nothing.

Finish with three possible moves that connect a strength to an opportunity or shore up a weakness against a threat, and say which one you would pick given the decision, and why.

Tip: Paste real inputs: customer feedback, sales numbers, a competitor's pricing page. A SWOT built from a two-sentence description produces two-sentence insight.

You have a plan you like and the team is excited, which is exactly when nobody is looking for the ways it fails. Use this before a launch, a hire, or a big spend.

Act as a seasoned executive running a pre-mortem. Imagine it is [TIME HORIZON] from now and this plan has failed badly. Your job is to explain why.

The plan: [PASTE THE PLAN]
What success was supposed to look like: [SUCCESS CRITERIA]
Who is involved and what they are responsible for: [WHO IS INVOLVED]
What we are assuming about customers, timing, or resources: [KEY ASSUMPTIONS]

Produce:
1. Ten distinct causes of failure, written as short news headlines from the future. Cover at least one each of: customer behavior, execution, people, money, timing, and something external.
2. For each cause, a likelihood (low, medium, high), an impact (low, medium, high), and the earliest warning sign we would see.
3. For the four causes with the highest combined likelihood and impact, a concrete mitigation we could put in place this week and a tripwire: the condition that should make us pause or cancel.
4. One uncomfortable question the team is probably avoiding.

Be specific to this plan; generic risks like 'scope creep' only count if you say exactly how they would show up here.

Tip: Run the pre-mortem with your team first, then paste their list and ask the AI what they missed. It is much better at finding gaps in a list than at replacing the people who know the work.

You are torn between two options, or you have already decided and want to check that you are not fooling yourself. Use this to get the strongest case for each side.

Act as a decision advisor who argues each side as forcefully as its best advocate would, then steps back and recommends.

The decision: [THE DECISION]
Option A: [OPTION A]
Option B: [OPTION B]
What I am optimizing for, in priority order: [WHAT YOU ARE OPTIMIZING FOR]
How reversible this is and when I must decide: [REVERSIBILITY AND DEADLINE]
What I am currently leaning toward and why: [CURRENT LEANING]

Structure your answer:
1. The strongest case for A, written by someone who believes in it. Five points, no strawmen.
2. The strongest case for B, same standard.
3. The three uncertainties that matter most, and for each, a cheap way to reduce it before deciding (a call, a test, a number to look up).
4. Your recommendation, with a confidence level and what new information would flip it.
5. If I lean the wrong way given my own stated priorities, tell me plainly.

If a third option that combines or sequences A and B is clearly better, propose it, but only if it is genuinely different, not a compromise that avoids choosing.

Tip: Leave out your current leaning the first time and see what it recommends. Then add it and compare. If the recommendation flips to match your preference, you have learned something about the AI's tendency to agree with you.

You want to understand how a competitor pitches itself and where the gaps are, using only what they publish. Use this before a positioning refresh or a sales battle card.

Act as a positioning strategist who reads marketing copy for what it reveals about strategy.

Competitor: [COMPETITOR NAME]
Their public copy, pasted from their homepage, pricing page, and one product page: [PASTE THEIR PUBLIC WEBSITE COPY]
My product or service: [YOUR PRODUCT OR SERVICE]
My target customer: [YOUR TARGET CUSTOMER]

Using only the text I pasted (do not add claims from memory; if you are unsure whether something is in the text, say so), analyze:
1. Who they are targeting, based on the language, examples, and pricing signals.
2. The core promise in one sentence, and the proof they offer for it (customer logos, numbers, guarantees, demos).
3. What they emphasize and what they never mention. Silence is often a weakness.
4. Pricing and packaging signals: what they charge for, what is free, what forces an upgrade.
5. Three positioning angles for me that they cannot easily copy, each with a one-line headline and the customer who would care.

Finish with two things I should verify before using any of this in a sales conversation.

Tip: Only paste publicly published material, and re-check any specific claim before you repeat it to a customer. AI models will confidently fill gaps with plausible-sounding details that are not on the page.

You have an offer, a pitch, or a pricing change and want to hear the objections before a real customer says them. Use this to rehearse, not as a substitute for real customer conversations.

Play the role of a skeptical prospective customer. Stay in character until I say 'step out'.

Who you are: [DESCRIBE THE CUSTOMER, THEIR ROLE, AND THEIR SITUATION]
What you currently use or do instead: [WHAT THEY USE TODAY]
What I am offering: [YOUR IDEA, PRODUCT, OR OFFER]
Price or ask: [PRICE OR ASK]

Rules for the role-play:
- You are busy, polite, and hard to impress. You have been burned by a similar purchase before.
- Ask the questions a real buyer asks: what changes for me, what does it cost in time and money, why you, why now, what happens if it does not work.
- Push back on vague claims and ask for specifics. Do not be won over by enthusiasm.
- Respond to what I actually say; do not give me the answers.

When I say 'step out', drop the character and give me: the three objections I handled worst, the one claim you found least believable, the moment you would have ended the call, and a rewrite of my opening two sentences.

Tip: Run it three times with three different customer types and keep a list of every objection. Then go ask five real customers; the AI's objections are plausible, not proven.

You have a deck, a long doc, or a pile of notes that everyone nods at and nobody can summarize. Use this to produce a one-pager people can actually repeat.

Act as a strategy editor who turns long documents into one page that a new employee could read and explain back.

Source material: [PASTE YOUR NOTES, DECK TEXT, OR DRAFT]
Time horizon: [TIME HORIZON]
Who will read the one-pager: [AUDIENCE]
What we will not do, if you already know: [THINGS WE WILL NOT DO]

First, tell me what is missing or contradictory in the source. If the material does not answer a section below, ask me rather than inventing it.

Then write a one-page strategy with these headings:
1. Where we are: three sentences, including the one number that matters most.
2. Where we are going: the goal at the end of the horizon, stated so we would know if we hit it.
3. Who we serve and what they need from us that others do not provide.
4. How we win: two or three bets, each with the reason we believe it.
5. What we will not do, and why (this is the section people fight over; make it explicit).
6. How we will know: three to five measures with a current value and a target.
7. Open questions we still have to answer.

Under 400 words. Plain language, no slogans.

Tip: Test it by handing the page to someone outside the team and asking them to explain the strategy back in a minute. Whatever they cannot repeat, rewrite.

You need a name that people will remember and that fits alongside what you already have. Use this to get a wide range of options in different styles before you check trademarks and domains.

Act as a brand naming specialist who generates options across styles and explains the reasoning behind each.

What we are naming, and what it does: [WHAT IT IS]
Who it is for: [WHO IT IS FOR]
The feeling or tone it should carry: [TONE]
Words, styles, or associations to avoid: [WORDS OR STYLES TO AVOID]
Existing names it must sit next to (company, product lines, other initiatives): [EXISTING NAMES IT MUST SIT NEXT TO]

Give me 25 names, five in each style:
- Descriptive (says what it is)
- Evocative (suggests the benefit or feeling)
- Coined or blended words
- Real words used in an unexpected way
- Names built on a person, place, or story connected to us

For each name: a one-line rationale, how it sounds spoken aloud, and one possible downside (hard to spell, sounds like a competitor, awkward plural).

Then shortlist your top five with a tagline under eight words for each. Remind me to check trademark databases, domain and social handle availability, and meanings in other languages before I fall in love with any of them.

Tip: Say the shortlist out loud on a phone call and in a sentence like 'I'll send you the ___ link'. Names that survive that test are the ones that work.

The next year is uncertain in ways that affect a real decision (hiring, inventory, a lease, a product bet). Use this to plan for a range rather than pretend you can predict.

Act as a scenario planner who helps small teams make decisions under uncertainty without paralysis.

My business or team: [YOUR BUSINESS OR TEAM]
The uncertainties I am watching (demand, costs, regulation, a key customer, technology, competition): [KEY UNCERTAINTIES]
Time horizon: [TIME HORIZON]
Decisions I need to make in the next few months: [UPCOMING DECISIONS]
Numbers I know today (revenue, costs, headcount, pipeline, whatever applies): [CURRENT NUMBERS]

Build three scenarios: Base, Upside, and Downside. For each:
1. A short narrative of how the period plays out, driven by the two uncertainties that matter most.
2. What the key numbers look like at the end of the horizon, as ranges, with your reasoning shown.
3. Signposts: three observable early indicators that we are in this scenario, and roughly when they would appear.

Then:
- No-regret moves: actions that make sense in all three scenarios.
- Options to keep open: cheap things we can do now to preserve flexibility.
- For each upcoming decision, which scenario it depends on and what trigger should make us commit.

Mark every number you estimate as an estimate. Do not present projections as facts.

Tip: Revisit the signposts monthly. The value of scenario planning is noticing early which world you are in, and that only works if someone actually checks the indicators.

You are about to spend real money or months on an idea. Use this to list what has to be true for it to work and find the fastest way to check the riskiest parts.

Act as a product strategist who is good at finding the assumption that sinks a plan.

The plan or pitch: [PASTE THE PLAN OR PITCH]
What success looks like and by when: [SUCCESS DEFINITION]
Budget and time I can spend on testing before committing: [BUDGET AND TIME FOR TESTING]
Evidence I already have: [EXISTING EVIDENCE]

Do this:
1. List every assumption the plan depends on, grouped by: customers want it, customers will pay, we can build or deliver it, we can reach customers affordably, the numbers work, nothing external blocks us. Aim for 12 to 20 assumptions.
2. Rate each one on importance (if wrong, does the plan fail?) and evidence (how much do we actually know?). Present as a table.
3. Identify the three assumptions that are both high importance and low evidence.
4. For each of those three, design a test I can run within my budget and time: what we do, who we talk to or show it to, what result would count as a pass, and what result would tell us to stop.
5. Suggest the order to run the tests and what I should decide after each.

Be blunt about weak evidence; catching wishful thinking is what this exercise is for.

Tip: Write the pass and fail criteria down before you run each test. Deciding what counts as success after you see the results is how bad ideas survive.