Run your own AI board of directors with 7 prompts

Running a business on your own means every big call lands on one desk. You make the decision, you check the decision, and you live with it. There's nobody in the room to ask the awkward question before you sign.

A real advisory board costs money and favours. A pretend one costs nothing and is surprisingly good at one job: arguing with you from several angles at once. The AI Board of Directors series is seven linked prompts that set up five advisors in ChatGPT, Claude or Gemini, then put them to work on decisions, money, habits, growth, ideas and a 90-day plan. Start with step 1 or read how the whole thing works first.

How to use the series

Run all seven prompts in one chat. Step 1 creates the board, and every later step calls on the same five people, so the advice stays consistent. Fill in the [square brackets] with real numbers and real situations. The board can only be as honest as what you tell it.

A word on expectations. These advisors are a thinking tool. They're quick at spotting untested assumptions and gaps in a plan, and they'll say things a polite friend won't. They don't know your market better than you do, and the finance prompt tells you to check anything big with an accountant. Treat the board as a sparring partner and keep the final call yours.

Our worked example throughout: Sam runs a four-person web design studio, turns over about £320,000 a year and is wondering whether to launch a monthly care plan product.

Step 1: Set up the board

Step 1 builds five seats: an operator who has built and sold a company, a finance director, a strategist, a psychologist and a long-view mentor. Each gets a name, a way of thinking, a favourite question and a blind spot. You also set how blunt they should be.

The rules matter more than the characters. Advisors must disagree when they see it differently, name assumptions you haven't tested, say "I don't know" when the facts are missing, and the chair closes every session with one decision and the reasoning.

Step 2: Put a big decision on the table

Step 2 takes one decision and runs it through a set order: what each advisor thinks you're missing, a list of assumptions marked tested or untested, three outcomes for the main option and the best alternative, an open argument, then a straight recommendation.

The closing ask is the useful bit: the cheapest test you could run this week, and the sign that should make you reverse course later.

In Sam's case: the board flagged that "clients will pay £250 a month for care" was untested. The cheapest test was emailing ten past clients a one-paragraph offer before building anything. If you want a second pass on the same decision, our premortem prompt lists the likeliest ways it fails, and this stress test checks it against the worst realistic case.

Step 3: A finance director who reads the numbers

Step 3 hands your income, costs, debts and cash to the finance director. You get a one-page cash flow summary, months of runway, costs that look out of line, a debt payoff order, three profit moves for the next 90 days with the expected effect, and a spending limit for the quarter.

It labels every benchmark as rough and flags what needs a qualified adviser. For a forward view rather than a review, pair it with this 90-day cash flow forecast.

Step 4: The pattern you keep repeating

Step 4 is the one people skip, and often the one that helps most. You describe a habit that keeps costing you (saying yes to every scope change, avoiding sales calls, rewriting work at midnight) with three real examples.

The psychologist looks for triggers, the story you tell yourself in the moment and what the habit gives you in the short term. It offers other explanations to test and ends with a pause script, a rule and a two-week experiment. It doesn't diagnose anything and suggests professional help where that fits.

In Sam's case: the pattern was discounting on every proposal. The board traced it to dreading the silence after saying the price. The fix was a written rule: send the price, then wait 48 hours before any follow-up.

Step 5: Growth that's already in the building

Step 5 looks at what you already own: skills, customer list, content, data, partnerships and the things customers ask for that you don't sell. It lists ten opportunities, explains why each fits you, scores them on feasibility, growth, cost and time to first result, then gives a two-week test with a pass or fail number for the top three.

Step 6: Pitch your idea to a sceptical board

Step 6 treats a business idea like an investment pitch. Each advisor asks their three hardest questions. The board scores demand, market size, competition, pricing, unit economics, room to grow and risk, then lists the weakest assumptions in order of damage. The verdict is pursue, change and retest, or drop, followed by a 14-day validation plan inside your budget.

If the idea survives and you need to package it, our Offer Stack series turns it into a priced offer.

Step 7: The quarterly accountability meeting

Step 7 sets one 90-day goal and runs a review. Each advisor names what's slowing you down, one argues for a different priority, and the board settles on three priorities plus what you'll stop doing. You get a 13-week plan with weekly milestones, targets for weeks 4, 8 and 13, and a check-in script to paste back every Friday.

That weekly script is what makes the series stick. The board remembers what you promised and asks about it.

The seven steps at a glance

1. Set up a five-seat advisory board
2. Put a big decision in front of the board
3. Have the finance director review your money
4. Work out why you repeat the same pattern
5. Find growth opportunities in what you have
6. Pitch a business idea to a board that pushes back
7. Run a 90-day accountability meeting

Tips for better sessions

• Give real numbers. "About £8k a month in costs" beats "some overheads".
• Ask for the blunt setting at least once. The polite board tends to agree with you.
• When an advisor says "I don't know", go and find out, then paste the answer back.
• Save the step 1 profiles in a note so you can restart the board in a fresh chat.

Start today

Write down the decision that's been sitting on your desk longest, then run step 1 and bring it straight to step 2. You'll have five opinions and one clear next step in about 20 minutes.