Template · 10 min read
Board report template: the pack directors read before the meeting
Every founder learns the same lesson about board meetings, usually the hard way: the meeting is only as good as the pack, and the pack is only as good as its pre-read rate. Eighty slides sent the night before produces a board that spends the meeting being caught up instead of being useful. The alternative is a short, candid document — verdict on page one, cash where nobody can miss it, the decisions you need stated as the agenda — sent early enough to actually be read. Here's that structure with the reasoning, a prompt that drafts the pack from your real numbers, and a delivery pattern that tells you which directors walked in prepared.
What a board report is actually for
A board pack has one job that outranks the others: getting the decisions you need from people who spend two hours a quarter thinking about your company. Directors are not an audience to impress — they're a resource to deploy, and the pack is how you aim them. Every page that doesn't inform a decision or an ask is spending their two hours on theater.
The second job is trust, and it's built in an unexpected place: the lowlights. Experienced directors read a pack backwards from the bad news — they know every company has problems, so a pack without visible ones reads as either naivety or concealment, and both get priced into how much rope you're given. Three honest lowlights on page one buy more credibility than any highlight can. The corollary: no surprises in the room. A material problem gets a phone call to key directors before the pack lands, and the pack then shows what you're doing about it. The meeting is where bad news is discussed, never where it's revealed.
The third job is memory. Boards run on comparisons — this quarter against plan, this pack against the last one, your progress against what you said. Which means the format has to hold still: same sections, same KPI definitions, same baseline plan. A pack that re-invents itself each quarter (or quietly re-baselines the plan the numbers are judged against) destroys the trend, and directors notice the destruction more than the numbers.
The board report template, section by section
Seven sections. Page one stands alone; everything after it is support and depth.
- One-page summary. The verdict sentence, three highlights, three lowlights, and the cash/runway line — complete on a single page. This is the only page every director is guaranteed to read, so it must contain the true state of the company on its own: a director who reads nothing else should still be un-surprisable.
- KPIs vs plan. The handful of metrics the board agreed matter, each shown as plan / actual / delta with green and red. Against the same plan every quarter — if the plan genuinely must change, that's a decision for the board, made in the open, not a quiet re-baseline that flatters the deltas.
- Cash and runway. Cash on hand, net burn, months of runway, and any known cliffs — stated plainly even when (especially when) it's uncomfortable. This is the number a board most needs to never be surprised by, and the one they'll check your candor against first.
- Progress since last board. What was agreed last meeting, item by item, with status. Small section, outsized effect: it closes the loop that makes board decisions mean something, and directors notice which CEOs track their own commitments.
- Decisions needed. The real agenda. Each decision framed with the options, the tradeoffs, and your recommendation — because a board decides better between options than in the abstract, and because asking for a decision without a recommendation reads as asking the board to do your job.
- Asks. The non-decision help: introductions, candidate referrals, domain expertise, a hard conversation you need backup for. Boards routinely go underused because nobody aims them; this section is the aiming.
- Appendix. Departmental updates, the full financials, cohort detail. Everything a diligent director might drill into, kept out of the body so the body stays a fifteen-minute read.
And a rule about the meeting itself, because the pack's structure implies it: don't present the pack. If it went out early, open with ten minutes on the one-pager and go straight to the decisions. The pack read aloud is the pre-read culture failing in real time.
Send it early — and know who read it
one link every quarter — the board's standing address
- Thu
Pack sent — 5 days ahead
- Fri 07:40
Opened · 18m on page
- Mon 22:15
Two more viewers, one brief
- Tue
Your move: one director hasn’t opened it — nudge gently
The pack goes out at least three full days before the meeting — five is better. This single habit does more for board quality than anything inside the document, because it converts the meeting from a briefing into a working session. Directors are busy people reading in airport lounges; the night-before pack isn't disrespected, it's just physically unread.
Delivering it as a tracked link adds the piece founders usually fly blind on: who's prepared. A long read from your lead investor Sunday night, a director who hasn't opened it by Tuesday — both change how you run Thursday. The unopened pack gets a friendly nudge ("anything you want added to the agenda?" works without accusing anyone); the well-read board gets a meeting that starts at the interesting part. Password-protect the page — a board pack is among the most sensitive documents a company produces — and keep it at the same URL each quarter so the board has one standing address instead of a trail of attachments.
The honest caveats: opens aren't comprehension, a director's assistant may be the "read," and governance has a real claim here — minutes reference a fixed artifact, and some boards' practice requires an immutable pack on file. The clean pattern is both: the tracked link as the living pre-read, and a PDF snapshot exported at send time, filed with the minutes. The link is for reading; the PDF is for the record.
Build the pack with AI (prompt included)
A board pack is mostly assembly — numbers you already have, arranged into a structure that doesn't change. That's exactly what AI is good at, with one hard boundary: it must never touch the numbers. The prompt below encodes the full structure, forces the one-page summary to stand alone, and bans invented figures outright — a hallucinated runway number in a board pack is not a typo, it's a governance incident.
Copy, fill the brackets with the quarter's real numbers, paste into Claude or ChatGPT:
You are producing a board report (board pack) as a single, self-contained HTML file. CONTEXT — replace everything in brackets: - Company: [company name] - Board meeting date: [date] - Prepared by: [CEO name / fractional CFO name] - Period covered: [month or quarter] - Verdict on the period in one sentence: [your honest one-liner] - Three highlights: [list] - Three lowlights: [list — real ones, not humble-brags] - KPIs vs plan: [KPI — plan — actual, one per line] - Cash position: [cash on hand] · monthly net burn: [burn] · runway: [months] - Progress on what was agreed at the last board: [list, matched to last meeting's actions] - Decisions needed from the board this meeting: [each decision with the options and your recommendation] - Asks (non-decision help wanted — intros, hiring, expertise): [list] - Brand: [primary hex color] + [logo URL, or "text wordmark"] REQUIREMENTS: - Page one is a complete summary on its own: verdict sentence, three highlights, three lowlights, and the cash/runway line. A director who reads only page one must know the true state of the company. - Sections in this order: One-page summary · KPIs vs plan · Cash & runway · Progress since last board · Decisions needed · Asks · Appendix. - KPI table shows plan, actual, and delta with green/red — against the SAME plan every period, no silent re-baselining. - Keep every number exactly as I gave it. If something is missing, leave a clearly marked [TBD] — never invent or estimate a financial figure. - Make the "Decisions needed" section visually prominent — it is the agenda of the meeting. - Self-contained HTML: inline CSS, no external scripts, readable on a phone, printable to PDF for the minutes. Stamp "Data through [date]" under the title. - Tone: candid and unadorned. Boards read spin as risk.
- 1 Paste the prompt (with your details filled in) into Claude or ChatGPT. You get a finished, self-contained HTML document — review it, correct anything, iterate until it's right.
- 2 Publish it as a permanent tracked link: with the Pagelive connector in Claude, say "publish this page" and the link comes back in chat — next cycle, "update it" replaces the content at the same URL. Or paste the HTML at app.pagelive.io.
Before it ships, read the output once as your hardest director would: check every figure against your model, check that the lowlights are real ones, and check that the verdict sentence is something you'd stand behind if quoted back to you in six months. AI drafts the pack; the candor in it has to be yours.
The mistakes that kill board packs
- The eighty-slide pack. Length substituting for clarity. Directors don't read it, so the meeting becomes the reading, so the decisions get ten minutes at the end. Everything below one-page-summary quality goes to the appendix.
- Sent the night before. Guarantees an unprepared board and wastes the one resource directors actually bring — considered judgment, which needs time you didn't give them.
- Revealing bad news in the room. Directors ambushed in front of peers go defensive, and defensive boards make worse decisions. Bad news travels by phone before the pack; the pack carries the response.
- No decisions section. A pack that's all report and no ask produces a meeting that's all commentary and no output — and a board that slowly concludes it isn't needed.
- Quiet re-baselining. Sliding the plan so the deltas stay green. The single fastest way to teach a board to re-check everything you send them.
- Spin. Lowlights written as humble-brags ("we grew so fast hiring couldn't keep up"). Experienced directors have read a thousand of these; the candor discount gets applied to the whole pack.
What to build it in
Slides remain the default board pack format, and for a board that expects to be presented to, they're fine. But slides optimize for the read-aloud meeting this template is trying to kill — and they pre-read terribly on a phone in a taxi, which is where directors actually read.
Docs (Google Docs, Word) pre-read well and print cleanly for the record. The tradeoffs are access management across outside directors on mixed email domains, and no signal at all on who's read what.
Board portals (the governance platforms built for this) solve the record-keeping, voting, and minutes problem properly and make sense once a board is formal enough to need that machinery — typically with independent directors and committees. For an early-stage company they're often more governance overhead than the board itself wants; check current pricing and weight against what your board actually asks for.
An HTML pack at a password-protected tracked link is the pattern above, and it's what Pagelive is built for: publish the pack to the same private URL each quarter, see who opened it before the meeting, and export the PDF snapshot for the minutes at send time. The honest limits: no voting or minutes machinery — it's the pre-read layer, not a governance platform — and the numbers are as current as your last publish. For the wider cap table beyond the board, the same pattern carries the investor update; for the deck that raised the round in the first place, see how to send a pitch deck.
Frequently asked
Five parts, in order: a one-page summary (verdict, three highlights, three lowlights), KPIs against plan, cash position with runway and burn, the decisions you need from the board, and an appendix with departmental detail. The one-page summary and the decisions section carry the meeting; everything else exists to support them.
At least three full days before the meeting — five is better. Directors read packs in the gaps of other jobs; a pack sent the night before guarantees an unread board and a meeting spent presenting instead of deciding. Send it early, and the meeting starts at the discussion instead of at page one.
The board report is for the directors of the company: it carries governance weight, includes the decisions you're formally asking for, and feeds the minutes. An investor update goes to the wider cap table: lighter, no decisions, more narrative. Same underlying numbers, different duties — sending the board pack to all investors is usually a mistake, and vice versa.
Match your board meeting cadence — quarterly for most companies, monthly for early-stage boards that meet monthly. What matters more is consistency of format between packs: directors compare against the previous pack, so metric definitions and section order must hold still even as the numbers move.
For reading, no — a link the directors can open on a phone gets read more, and a tracked one tells you who's prepared. For the record, often yes: minutes typically reference a fixed version of the pack, and some boards' governance practice requires an immutable artifact on file. The clean answer is both — a link for the pre-read, a PDF snapshot filed with the minutes.
Related: how to send a pitch deck · QBR template · executive summary examples
One board link. Password-protected, pre-read, and you know by whom.
Draft the pack with AI, publish it to the board's standing link, and walk into the meeting knowing who's prepared. Free for 5 pages, no card.