Answering Board Questions Live, Without 'I'll Get Back to You'

The moment arrives about forty minutes into the meeting. A director — usually the one who ran finance at a company two stages ahead of yours — sets down the deck and asks something that isn't on it. "What was your CAC payback in Q3 versus Q2, and does that hold if you strip out the two enterprise logos?" The room's temperature changes. Answering board questions live is the part of the job no rehearsal fully covers, because the question that matters is rarely the one you prepared a slide for.

There are two ways the next ten seconds go. Either you cut the number in front of them, or you say "I'll get back to you" — and something small but permanent shifts in how the board reads your command of the business.

The figures you carry cold

A handful of unit-economics numbers are not report line items. They are things a finance leader is expected to know the way a pilot knows airspeed. Directors probe them not to catch you out but to test whether you understand the machine you're running.

Burn multiple — net burn divided by net new ARR. David Sacks popularized the metric as a single-number read on efficiency: below 1x is elite, above 2x invites questions. Directors push on the denominator. Is that net new ARR gross of churn? Then it isn't burn multiple, it's a flattering cousin.

Net dollar retention — expansion minus churn and contraction within your existing base. Bessemer's State of the Cloud work has made 120%+ the number public SaaS benchmarks are judged against. The follow-up is always cohort-level: is retention holding across the base, or is one vintage carrying it?

LTV/CAC — lifetime value over fully loaded acquisition cost. The trap is the "fully loaded" part. Directors ask whether CAC includes SDR salaries and marketing overhead or just paid spend, because the ratio doubles when you quietly drop half the cost.

Payback period — months of gross margin to recover CAC. Under 12 months reads healthy; the probe is whether it's trending the right direction and whether it holds by segment. The blended number can look fine while your mid-market motion bleeds.

Default-alive date — Paul Graham's framing: on current revenue growth and spend, does the company reach profitability before the cash runs out? Directors want the date, and they want to know which two assumptions move it most.

Why the freeze isn't a knowledge problem

Here is the uncomfortable diagnosis. The finance leader who can't answer usually knows the concept perfectly. What they can't do is produce the cut the director just asked for — Q3 payback with two logos removed — because the model that holds it was refreshed on the last day of the quarter and hasn't moved since.

The blended payback number is on slide 14. The segmented version, ex-enterprise, isn't anywhere. To build it, you'd re-pull the CRM export, re-map the accounts, and re-run the formula. That's a forty-minute job, not a ten-second one. So you say you'll follow up, and the follow-up email goes out three days later when the moment has passed.

This is the same failure that turns a board deck into a data dump instead of a decision document. The static appendix is an attempt to pre-answer every question you can anticipate — twenty backup slides, most never shown. It works right up until the director asks the one cut you didn't foresee, which they reliably do. It's also, at root, a visibility problem: the answer exists in your systems, just not in a form you can reach in the room.

Anticipation versus retrieval

The conventional advice is to prepare harder: de-risk the meeting before it starts by pre-briefing directors and stacking the appendix. That's sound, and it lowers the odds of surprise. But anticipation has a ceiling. You cannot pre-build every segmentation of every metric, and the good directors know exactly where your prepared answers stop.

The alternative isn't more slides. It's the ability to retrieve — to pull a live segment on the spot, in the room, from data that reflects this week rather than last quarter. When the narrative you've architected holds up under a live cut, the board stops probing and starts trusting.

Where the tooling actually sits

Be honest about the range, because none of the options is free. The tradeoffs here mirror the broader state of the FP&A stack.

A well-built spreadsheet model is the most common answer and the most brittle. It's precise until someone asks for a cut you didn't wire, at which point it's a manual rebuild. Google Sheets and Excel remain the backbone of FP&A for a reason, but they refresh when a human refreshes them.

BI tools close part of the gap. A Looker or Tableau dashboard sitting on a live warehouse can slice by segment on demand — if the metric was modeled in advance and someone maintains the semantic layer. The failure mode is that the director's exact question falls outside the dimensions you built.

Modern finance platforms — the category a16z has tracked as the CFO stack — push toward connecting billing, CRM, and the general ledger so the unit-economics figures recalculate as the underlying data moves. The promise is that a mid-meeting cut is a filter, not a project. The reality is that it depends entirely on how clean your source systems are; garbage upstream still produces garbage on demand.

The through-line across all three is whether the number you're asked for is live or last-refreshed — worth auditing which of your board metrics could actually survive a follow-up question. →

The point isn't the tool. It's the posture. A finance leader who can answer the second and third question — not just the one on the slide — reads as someone who operates the business rather than reports on it. That distinction is most of what a board is measuring when it watches you handle a question you didn't see coming. The metrics you carry cold are table stakes. Being able to cut them live, in the room, is the part that builds the credibility everything else in the board relationship runs on.