What the meeting is for
A board meeting is not a status report. Your board has read the numbers — or should have, which is an argument for sending them in advance.
The meeting exists to get judgment on decisions you cannot make alone. Every slide should either establish shared context for one of those decisions or be cut.
The practical version of this rule: send the deck 72 hours ahead, spend the first ten minutes on numbers, and give the remaining time to two or three real questions.
The 72-hour rule
Decks that arrive the night before guarantee the first thirty minutes are spent reading. Sending early converts the meeting from presentation into discussion, which is the only thing that makes it worth the calendar time.
The financial section
Five slides is enough. More than that and you are reporting rather than deciding.
- Scorecard. Six to eight metrics, each with actual, plan, and prior period. One slide, no commentary.
- Revenue detail. New, expansion, contraction, and churn broken out. Net revenue retention alongside.
- Cash and runway. Cash on hand, gross and net burn, runway under base and downside cases.
- Plan variance. Every material miss against plan with the reason and the corrective action.
- Forecast. Updated projection with what changed since last quarter and why.
Metric definitions do not change
The fastest way to lose board confidence is a metric that means something different this quarter than last.
Write your definitions down once and put them in an appendix that appears in every deck. If a definition genuinely must change, show both the old and new basis for at least two quarters and say plainly why.
- What counts inside ARR — and what does not (pilots, services, one-time fees)
- When a customer is counted as churned
- Which costs sit in COGS versus operating expense
- Whether CAC includes fully loaded sales salaries or just program spend
- Whether burn is gross or net, and whether it is on collected cash
Report bad news first
Boards are far more tolerant of problems than of surprises. A miss you surface with a diagnosis and a plan is a competence signal. The same miss discovered by a board member reading footnotes is a trust problem.
The workable structure for any variance: what happened, why it happened, what you are doing about it, and what you need from the room. Four sentences, no hedging.
Never surprise your board in the meeting
Material bad news gets a call before the deck goes out. The meeting is for deciding what to do, not for the initial emotional reaction.
The appendix does the heavy lifting
Everything a board member might want to verify goes in the back, where it is available without consuming meeting time.
- Full P&L, balance sheet, and cash flow statement
- Cohort retention tables
- The complete metric definitions page
- Headcount by function with open roles
- Cap table summary and current option pool
- Detailed model assumptions
Common failures
- Forty slides, no decision requested
- Metrics with no comparison — a number with no plan or prior period next to it is not information
- Charts with no axis labels or a truncated y-axis that overstates a trend
- Different metric definitions than last quarter, undisclosed
- A forecast that has never once been revised downward
- Burying the runway number where it has to be hunted for