Board meetings do not punish bad numbers. They punish numbers that cannot be explained. A retention slide that triggers four follow-up questions you cannot answer does more damage than a churn figure you saw coming and can talk through.
The metrics below hold up. The ones after them usually do not.
What the board is actually asking
Behind every question about Customer Success is one of three real questions: is the revenue we already have safe, do we know about problems early enough to act, and is the team's effort going where the money is. Every metric you present should answer one of those directly.
The three that hold up
Net revenue retention, split by segment. One blended number hides everything interesting. Split it and you can say which part of the base is compounding and which is leaking, which is the beginning of a decision rather than the end of a slide.
Revenue at risk, with a date attached. Not a count of unhappy accounts — the dollar value of what is up for renewal in the next two quarters and currently off-track. This is the number that turns a CS conversation into a revenue conversation.
Warning time. How many days pass between an account being flagged and its renewal date. If that gap is shrinking, your team is finding out too late, and no amount of effort at the end will fix it.
Revenue at risk this quarter, the accounts behind it, and what is already in motion on each. Everything else is supporting material.
The ones that collapse under a follow-up
Average health score across the portfolio: it moves for reasons nobody can reconstruct, and a board member who asks "why did it drop two points" will not get an answer. Ticket volume: it measures activity, not outcome. NPS in isolation: it tells you how people felt when they answered, not whether they will renew.
None of these are useless internally. They just cannot carry a board conversation on their own.
Make the number defensible before the meeting, not during it
A metric is defensible when three things are true: it comes from the same system the team works in, anyone can drill from the number to the accounts behind it, and the definition has not changed since last quarter. If you rebuild the numbers in a spreadsheet the week before every board meeting, you will spend the meeting defending the spreadsheet.
Present risk as a decision, not a confession
"We have $1.4M at risk in Q3" invites panic. "We have $1.4M at risk in Q3, here are the six accounts, four have executive engagement scheduled and two need a commercial decision from this room" invites a decision. Same fact, entirely different meeting.
That framing only works if the underlying data is one click away. When your health data and your contract data live in the same place, it is. When they live in two systems that sync overnight, it is not.