20 AI Prompts for Board Meeting Prep

TL;DR: Board meeting prep splits into four jobs: turning raw numbers into a narrative, building the deck outline around that narrative, rehearsing the hard questions before the board asks them, and writing minutes and follow-ups that people actually act on. The 20 board meeting prompts below cover all four. AI should never invent a metric or a decision — it drafts the structure and language around numbers you supply.

Most board meeting prep advice focuses on the deck template: slide count, section order, a logo in the corner. Templates matter less than most founders think. What separates a board meeting that builds trust from one that erodes it is the narrative underneath the slides, whether the story you're telling explains what changed and why, or just lists what the team was busy with. A board that reads ten activity bullets learns nothing it can act on. A board that reads three sentences about what moved and why can actually help.

This page is built around that distinction. The 20 prompts below are grouped into four stages, in the order an actual board cycle runs: build the narrative from your numbers, structure the deck around it, rehearse the questions a sharp board member will ask, and close the loop with minutes and follow-ups that don't evaporate the moment the call ends. Every prompt treats your real numbers, decisions, and context as something you supply. The model never invents a metric, a churn number, or a hiring decision.

The gap this page is trying to close is a specific one. Most founders are perfectly capable of writing a clear paragraph about their own business when they're not under time pressure. What actually breaks board prep is the combination of a hard deadline, a fresh set of numbers that just landed, and the blank-page problem of turning both into a coherent narrative in the same afternoon the deck is due. A structured prompt doesn't replace the thinking; it removes the friction of staring at an empty document while the thinking is still happening, so the actual judgment call, what matters this month and why, gets more of your attention instead of less.

It's also worth naming who this is for. A first-time founder with an all-VC board has a different job than an operator running a bootstrapped company with two advisors and no institutional investors at all. Both need a narrative, a deck, and a way to handle hard questions. Neither needs the other's tone. The prompts below are built to be adapted, not read verbatim into a board call, and the closing section says more about how to tailor them to your specific board.

Read that table once before you start. The order matters: writing slides before you have a narrative just produces a nicer-looking activity list.

How do you turn raw numbers into a board narrative?

By deciding what changed and why before you decide how to display it. A narrative answers "what should the board understand about this month" in two or three sentences; a metrics table answers "what happened" without saying which parts matter.

1. Month-in-three-sentences narrative

I'm prepping a board update. Here are this month's key numbers and events:
"[PASTE: REVENUE, GROWTH RATE, CHURN, RUNWAY, HEADCOUNT, ANY MAJOR EVENTS]"

Write a 3-sentence narrative summary that leads with what changed and why
it matters, not a list of what happened. If nothing meaningfully changed
this month, say so plainly rather than manufacturing significance.

2. Metric-to-story bridge

For each of these metrics, write one sentence explaining what changed and
the most likely reason why, based only on context I give you:
"[METRIC: OLD VALUE -> NEW VALUE. CONTEXT: WHAT WE DID OR WHAT HAPPENED]"

If you don't have enough context to explain a change, say "cause unclear"
rather than guessing at a plausible-sounding reason.

3. Good-news-without-spin check

Read this board update draft and flag any sentence that oversells a result
or buries a weak one under vague language.

Draft: "[PASTE DRAFT]"

For each flag, quote the sentence and suggest a more direct, specific
version. The goal is credibility with the board, not maximum positivity.

4. Bad-news framing, honestly

Help me write the paragraph explaining this setback to the board: "[WHAT
WENT WRONG, WHY, WHAT WE'RE DOING ABOUT IT]"

Direct and specific, no euphemisms, no burying the lead in a longer
paragraph about unrelated wins first. State the setback, the cause as we
understand it, and the concrete next step, in that order.

5. Trend-line context prompt

This metric has moved from "[OLD VALUE]" to "[NEW VALUE]" over "[TIME
PERIOD]". Here's relevant context: "[SEASONALITY, ONE-TIME EVENTS,
STRATEGY CHANGES]"

Write two sentences a board member with no other context would need to
correctly interpret this trend, without overstating confidence in the
cause.

Notice the instruction repeated across these five: use only what I give you, flag what's missing, don't manufacture confidence. A model asked to explain a metric will produce a plausible-sounding cause even when you haven't given it one, and a confident wrong explanation is harder for a board to catch than an honest "cause unclear."

How do you build the deck outline around that narrative?

By writing the outline before the slides, so the narrative decides the structure instead of the previous month's template deciding it for you.

The most common failure at this stage isn't bad writing, it's inherited structure. A deck built by copying last month's section order and swapping in new numbers quietly locks in whatever mattered last month, whether or not it's still what matters now. Writing the outline fresh from this month's narrative, even when several sections end up in the same order as before, forces an actual decision about ordering rather than a default.

6. Deck outline from narrative

Here's this month's board narrative: "[PASTE 3-SENTENCE NARRATIVE FROM
PROMPT 1]"

Here are the topics we typically cover: "[LIST: METRICS, PRODUCT,
HIRING, RUNWAY, ASKS, ETC.]"

Build a slide-by-slide outline that leads with the narrative, not the
topic list. Order sections by what matters most this month, not by our
usual template order.

7. Executive summary slide

Write a one-slide executive summary combining this narrative and these
top 3 numbers: "[NARRATIVE + 3 KEY METRICS]"

Under 60 words on the slide itself. This is the slide a board member reads
if they read nothing else — it needs to stand alone.

8. "Ask" slide framing

I need the board's help with this: "[WHAT YOU NEED: INTRO, ADVICE,
DECISION, MONEY]"

Write a clear, specific ask slide. State exactly what you need, by when,
and what you've already tried. Vague asks ("any intros welcome") get
vague responses — make this one specific enough to act on.

9. Product update in plain language

Turn these product/engineering updates into board-readable language:
"[PASTE TECHNICAL UPDATES/CHANGELOG]"

Assume the reader is smart but not technical. Explain what shipped and
why it matters to the business, not how it was built. Cut anything that's
interesting to engineers but irrelevant to the board.

10. Comparative context slide

Here's our number and, where I have it, a rough benchmark or prior period:
"[OUR METRIC], "[BENCHMARK OR PRIOR PERIOD, IF KNOWN]"

Write one sentence of honest comparative context. If I haven't given you
a real benchmark, don't invent an industry average — say the comparison
isn't available yet.

11. Risk register slide

Turn this list of current risks into a board-readable risk slide: "[LIST
OF RISKS: WHAT, LIKELIHOOD, WHAT WE'RE DOING]"

One line per risk: what it is, our mitigation, and who owns it. Order by
severity, not by how recently each one came up.

12. Slide tightening pass

Tighten this slide's text to fit on one slide without losing the point:
"[PASTE SLIDE TEXT]"

Cut to the fewest words that still make the point clearly. Flag anything
you cut that changes the meaning, rather than silently dropping it.

Run the tightening pass last, after every other slide is drafted. Editing for length before the content is settled just means editing the same slide twice.

One more thing worth deciding before this stage: how much of the deck is actually meant to be read live versus sent as a pre-read. Boards that get a deck 48 to 72 hours ahead of the call and actually read it can spend the live time on discussion instead of narration, which changes what the slides need to do. A deck built assuming nobody reads it ahead of time ends up over-explained on every slide, because it has to work as both the pre-read and the live narration at once. Decide which job your deck is doing before you tighten it, not after.

How do you rehearse the questions a sharp board member will ask?

By having a model play that role honestly against your actual deck, before the real board does it to you in the room.

This is the stage founders skip most often, usually because rereading your own deck one more time feels like enough preparation. It rarely is. The person who wrote the deck already believes the narrative in it, which makes it genuinely hard to spot the weak spot a skeptical outsider would find in thirty seconds. Handing the actual deck text to a model and asking it to be genuinely critical, not softened for your benefit, surfaces exactly the kind of question a rereading of your own work will not.

13. Skeptical-investor pass

You are a skeptical, experienced board member reviewing this deck for the
first time. Ask the 5 hardest questions you'd genuinely have.

Deck content: "[PASTE DECK TEXT OR OUTLINE]"

Be genuinely skeptical, not softened for my benefit. Focus on the weakest
parts of the narrative, not surface-level nitpicks about formatting.

14. Weak-metric interrogation

This metric looks weaker than the others in the deck: "[METRIC AND
CONTEXT]"

Write the follow-up question a board member would ask about it, and draft
an honest answer using only the context I've given you. If the honest
answer is "we don't fully know yet," write that rather than a
softer-sounding non-answer.

15. Runway and burn question prep

Given this runway and burn context: "[MONTHS OF RUNWAY, MONTHLY BURN,
ANY PLANNED CHANGES]"

Draft answers to the three questions a board typically asks about runway:
what happens if fundraising takes longer than planned, what gets cut
first, and what triggers a hard decision. Use only the numbers and plans
I've given you.

16. Competitive question prep

A board member might ask how we compare to "[COMPETITOR NAME]" on
"[SPECIFIC DIMENSION]". Here's what I actually know: "[YOUR KNOWLEDGE OF
THE COMPARISON, INCLUDING GAPS]"

Draft an honest answer that states what we know, what we don't, and
avoids claiming a competitive advantage I haven't actually verified.

17. "What would you cut" stress test

If we had to cut 20% of spend next quarter, here's roughly where it would
come from: "[YOUR ROUGH PLAN]"

Play a board member asking follow-up questions about this plan — what
breaks, what's protected, and why. Help me find the gaps in my own
thinking before the real question comes.

Treat sections 13 through 17 as rehearsal, not scripting. The value isn't a memorized answer — it's finding the two or three places your own story is thinner than you'd assumed, while you still have time to either fix the underlying problem or prepare a specific, honest response.

How do you write minutes and follow-ups people actually act on?

By assigning an owner and a date to every action item in the same document the discussion produced, instead of a separate action-items list that quietly stops getting checked.

Minutes get written after the hard part is over, which is exactly why they're the part most likely to get rushed. A call ends, everyone's attention has already moved to the next thing, and the notes get typed up quickly with good intentions and no real structure. The action items buried inside a paragraph of minutes are the ones that quietly never happen, not because anyone decided to drop them, but because nothing forced them into a format where a missing owner or a missing date was visible.

18. Meeting minutes from raw notes

Turn these raw board meeting notes into clean minutes: "[PASTE NOTES OR
TRANSCRIPT]"

Structure: decisions made, action items with owner and due date, open
questions still unresolved. Do not add a decision or action item that
wasn't actually discussed — only clean up and organize what's here.

19. Action-item extraction with owners

Pull every action item mentioned in this discussion into a tracked list:
"[PASTE NOTES OR TRANSCRIPT]"

For each item: what it is, who owns it (if stated), and the deadline (if
stated). Mark "OWNER UNCLEAR" or "NO DEADLINE SET" rather than assigning
one that wasn't actually agreed.

20. Board follow-up email

Write a short follow-up email to the board after this meeting, summarizing
key decisions and next steps: "[PASTE MINUTES OR KEY POINTS]"

Under 200 words. Lead with decisions made, then open action items with
owners, then the date of the next update. No recap of the full discussion
— that's what the minutes are for.

The short version

A board meeting earns trust the same way a status update to anyone does: by leading with what changed and why, not with a list of what kept you busy. The 20 prompts above are built around that order deliberately, narrative first, deck structure second, honest rehearsal third, and a close-out that assigns real owners and dates fourth. None of them replace the judgment of knowing your own numbers, your own risks, and your own board. What they replace is the blank-page time between "the month closed" and "the deck exists," and the false confidence of walking into a board meeting having never once asked yourself the hard question first.

Save the prompts that work for your specific board. The tone that lands with an experienced VC board differs from one that works with a bootstrapped, founder-heavy board, and a template built for a five-person board of long-time advisors will feel over-formal for a two-person kitchen-cabinet check-in. Adjust the language, keep the underlying order, and reuse the structure every cycle instead of rebuilding it from scratch each month.

One habit worth building alongside these prompts: after each board meeting, note which questions actually came up that your rehearsal missed. Over two or three cycles, that running list becomes a better predictor of what your specific board asks than any generic list of "hard board questions" ever will, because it reflects what your actual board cares about rather than what boards care about on average. Feed that list back into the skeptical-investor prompt next time, and the rehearsal gets sharper with every cycle instead of staying generic.

None of this changes what actually earns a board's trust over a longer horizon, which is a track record of the narrative in the deck matching what was true at the time, quarter after quarter. A well-structured update built on an honest number beats a beautifully narrated one built on a number that later needed walking back. Structure and rehearsal make the update better. They don't substitute for the number being right.