Board reporting: the revenue section, done right

A founder sent me a board deck four days before the meeting and asked one question: "Can you check the revenue slides?"

The ARR number on slide 6 was 4.2 million. The ARR number in their HubSpot dashboard was 3.87 million. The pipeline number came from a Google Sheet the VP Sales updated on Sunday nights. Nobody in the company could say, out loud, what counted as a closed deal. Two people had different answers about whether an annual contract signed on the last day of the quarter went into Q2 or Q3.

The deck itself looked great. Clean charts, good design, a nice green line going up and to the right. It was also going to get taken apart in about eleven minutes by an investor who had seen the same slide last quarter and remembered a different number.

I have watched this happen at maybe fifteen companies now. The pattern is always the same. The revenue section of the board deck is rebuilt from scratch every quarter, by hand, from three or four sources that do not agree with each other. It takes a week of somebody's life, it produces numbers that shift between meetings, and it teaches the board to distrust the operating data instead of the operating plan.

This is a data problem wearing a presentation costume. Fix the data layer once and the revenue section takes an hour to produce, every quarter, forever.

Why the revenue section is the one that gets picked apart

Every other section of a board deck is qualitative. Product roadmap, hiring plan, strategic priorities: the board can push back on judgment, but there is no external check on the facts.

Revenue is different. Your board has your last four decks. They have the operating plan you committed to. Several of them sit on other boards and know what 110 percent net revenue retention looks like in your segment. They can do arithmetic in their heads while you talk.

So when the ARR bridge does not add up, or the pipeline coverage number moved from 3.4x to 2.1x with no explanation, the conversation stops being about your business and starts being about your reporting. That is the worst possible use of two hours with the people who decide whether you get a bridge round.

The failure is rarely dishonesty. It is that four different people pulled four different numbers from four different places, and nobody owns the definition.

The six numbers a board actually reads

I have sat in on enough of these to know what gets looked at and what gets skipped. Boards read six things in the revenue section. Everything else is appendix material.

1. The ARR bridge. Starting ARR, plus new, plus expansion, minus contraction, minus churn, equals ending ARR. One waterfall chart. If a board member only looks at one slide, it is this one, because it answers whether growth is coming from selling more or from finally stopping the leak. Get the difference between bookings, ARR and recognised revenue straight before you build it, because mixing them is the single most common error I find.

2. Net revenue retention and gross revenue retention, together. NRR alone hides a bad story. A company with 118 percent NRR and 82 percent GRR is buying its growth from a handful of accounts while the base burns. Show both or expect to be asked. NRR is the number that most predicts your next round, which is exactly why boards treat a lonely NRR figure with suspicion.

3. Pipeline coverage, with context. The ratio on its own is close to meaningless. More on that below.

4. CAC payback, in months. Not LTV:CAC. Payback period is harder to game and easier to sanity-check against the cash plan. Under twelve months is the current bar for most B2B SaaS. If yours is 19 months, say 19 months and explain the plan, because the way most teams calculate CAC leaves half the cost out and your board knows it.

5. Actual versus plan, by quarter, cumulative. Not just this quarter. Cumulative against the annual number, so the board can see whether you are behind or just lumpy.

6. Next quarter forecast, with a range. One number is a hostage to fortune. Commit, best case, and worst case, with the assumptions written next to each.

That is it. Six numbers, four or five slides. The rule of thumb I use is seven to ten KPIs total across the whole deck, and which seven depends on your stage, which I broke down in the eight SaaS metrics worth tracking. Below seven and you are hiding something. Above ten and you have buried the signal in noise, which reads exactly the same as hiding something.

Pipeline coverage is where most decks lie by accident

Coverage is the number I see abused more than any other. Total open pipeline divided by the quarter's target. If it clears 3x, everyone relaxes.

The problem is that identical ratios describe completely different businesses. A 3.5x coverage number made of 40 deals spread across stages, with an average age of 26 days, is healthy. A 3.5x number where 60 percent of the value sits in four deals, two of which have been in "negotiation" since March, is a disaster with good optics. Both slides say 3.5x.

Two additions fix most coverage slides. First, concentration: what share of the pipeline sits in your five biggest open deals. Second, age: median days in the current stage. A board that sees both stops asking whether the number is real and starts asking about the deals, which is the conversation you want.

I wrote a longer piece on why the 3x coverage rule breaks for most teams and how to work out your own ratio from win rate and cycle length. Do that math once and put your real ratio in the deck instead of the industry folklore one.

Where the numbers should come from

Here is the part that turns a week of work into an hour.

Every number in the revenue section should trace back to one system of record, with one owner and one written definition. For almost every company under 200 people, that system is the CRM. Not the finance spreadsheet, not the BI tool that reads from a warehouse that syncs twice a day, not the VP Sales' personal tracker.

The CRM is where deals are created, moved and closed. If the CRM's numbers are wrong, the answer is to fix the CRM, not to build a parallel set of books in Sheets. Every parallel set of books I have seen becomes the real one within two quarters, at which point the CRM is dead and forecasting goes with it.

Three things make CRM-sourced board numbers hold:

Closed-won hygiene. A deal is closed-won when the contract is signed and the amount matches the contract. Not when the rep is confident. Amount, close date and contract start date are required fields on the closed-won stage, and the deal cannot be moved without them. This one rule removes most of the drift between your CRM and your finance numbers.

Snapshots. This is the piece almost nobody sets up, and it is the reason numbers change between decks. Pipeline is a live object. If you pull coverage on the first of October and again on the eighth, you get different answers, because deals moved. So take a snapshot on day one of each quarter and store it. HubSpot does this natively with deal stage history and snapshot-based reports. Salesforce has opportunity history. In both cases you have to turn it on and build the report deliberately. Without snapshots, you cannot answer "what did we say last quarter and what actually happened", which is the question every good board member asks.

One definitions page. Written down, shared, boring. What counts as ARR. What counts as expansion versus new. Which currency and which FX rate. How you treat multi-year deals, pilots, and the deal that got signed at 11pm on the last day. Two paragraphs. It saves you an argument every single quarter, and it is the first thing a diligence process will ask for.

Building it in HubSpot

Most of our clients run HubSpot, so here is the concrete version.

Create a dashboard called "Board", separate from the sales team dashboards. Sales dashboards are built for pipeline management and change constantly. The board dashboard should be frozen and boring.

The ARR bridge is the hardest report to build natively, and I will be honest about that. HubSpot does not give you a waterfall out of the box. What works: a custom deal property for ARR type with values of new, expansion, contraction and churn, set on every closed deal. Then a single report grouped by that property, filtered to the quarter. You export four numbers and drop them into a waterfall chart in the deck. Five minutes, and it is reproducible because the underlying property is on the record, not in someone's head.

For retention, you need renewal deals as real deal records, not a note in the account. If your renewals live in a spreadsheet, the NRR number in your deck is an estimate and you should label it as one. Getting renewals into the CRM as deals is usually a two-week project and it fixes retention reporting, forecast accuracy and the renewal motion at the same time.

For coverage, use a snapshot report rather than a live pipeline report, and add a second report for concentration: open deals sorted by amount, top five, as a percentage of total. That one is a list view, not a chart.

If your data is not in a state where these reports can be trusted yet, that is the actual project, and it comes before the deck. Our CRM and RevOps work usually starts here, because poor CRM data quality shows up in board reporting before it shows up anywhere else. The board deck is the smoke. The data model is the fire.

For teams on Salesforce or Attio the mechanics differ but the shape is identical: one source, defined fields, historical snapshots, a locked dashboard.

Automating the assembly

Once the reports exist, the assembly is a scheduled job rather than a person's week.

We build this with n8n for clients who want it hands-off. On the first working day after quarter end, a workflow pulls the six reports through the CRM API, writes the values into a Google Sheet that feeds the deck's charts, and posts a summary into Slack with the quarter-over-quarter deltas. The CEO opens the deck and finds the numbers already in it. What is left is the part only a human can do: the commentary, the "here is what went wrong and what we are doing about it".

That distinction matters. A board does not want your numbers, they want your read on your numbers. Every hour spent assembling data is an hour not spent thinking about what it means. This is the same argument I make about sales dashboards, and it applies double at board level, where the audience is expensive and the meeting happens four times a year.

If you want the automation side without rebuilding your whole stack, that is roughly what our AI and automation work covers.

What to cut from the revenue section

Things I remove from client decks almost every time:

Vanity funnel metrics. MQL counts, website traffic, form fills. No board member has ever made a decision based on your MQL number. If marketing performance matters, show pipeline created and pipeline converted by source, which is a real attribution question.

Win rate without a denominator definition. Win rate calculated on closed deals only tells a very different story than win rate on all opportunities created, including the ones still open or quietly deleted.

Logo counts as a headline. Logos matter at seed. By Series A the board is looking at ARR and retention, and a logo count next to a flat ARR number invites a bad question.

Anything with a chart type you invented. Waterfall, line, bar, table. That is the whole vocabulary.

Every quarter that you show a metric, you have committed to showing it again next quarter. Adding a metric because it looks good this quarter is a trap, because you will be asked for it in the quarter when it does not.

The uncomfortable part

Sometimes cleaning the data means the new number is worse than the old one. ARR drops from 4.2 million to 3.87 million because you finally excluded the pilots and the deal that never got countersigned.

Tell them. Put a slide in front of the revenue section that says: we changed how we calculate ARR, here is the new definition, here are the last four quarters restated on the new basis. Do it once, take the hit, and then never have the conversation again.

Every board I have watched handle this has responded well, because a founder who tightens their own definitions is a founder who can be trusted with the next number. The alternative is discovering the discrepancy during diligence, eighteen months later, with a term sheet on the table. That version is much more expensive. It is also how revenue leakage quietly compounds: nobody wants to restate, so the wrong number gets defended for another year.

FAQ

What revenue metrics belong in a Series A board deck?

Six: the ARR bridge, net revenue retention shown next to gross revenue retention, pipeline coverage with age and concentration, CAC payback in months, actual versus plan cumulatively, and a forecast range for next quarter. Seven to ten KPIs across the whole deck is the working range. Anything past that buries the signal, and any metric you add becomes a metric you owe them every quarter after.

How far in advance should the board deck go out?

Somewhere between two and seven days, depending on how your board reads. Sacks argues for one to two days so the data is fresh; Sequoia's guidance leans towards a week so directors have time to think. The variable that actually matters is whether your numbers can be produced quickly. If assembly takes five days, you send late and you send stale. Fix the assembly and the timing question answers itself.

Should board numbers come from the CRM or from finance?

Both, with a clear split. Recognised revenue, cash and burn come from finance. ARR, bookings, pipeline and retention come from the CRM. The two must reconcile, and the definitions page is what makes them reconcile. What does not work is a third set of numbers in a spreadsheet that nobody owns, which is what most companies actually have.

How do I stop the numbers changing between board meetings?

Snapshots. Store pipeline and ARR values on day one of each quarter so historical figures are frozen records rather than live queries. Without this, a report run today will always return a different past than the same report run last quarter, because deals get edited, close dates move and records get merged. This is the single most common cause of a board deck contradicting the one before it.

Is it worth buying board reporting software?

For most companies under 200 people, no. The tools are fine but they read from your CRM, so if the CRM data is wrong you have bought a faster way to publish wrong numbers. Get the definitions, field hygiene and snapshots in place first. If reporting is still painful after that, the tool will actually help. Nine times out of ten the pain disappears once the data layer is fixed.

Getting the reporting layer right

The revenue section of a board deck is a reporting problem, not a design problem. Six numbers, one source, written definitions, frozen history. Companies that do this spend an hour a quarter on the deck and the rest of the time on what the numbers mean. Companies that do not spend a week and then argue about arithmetic in front of their investors.

If your last board meeting turned into a debate about where a number came from, that is the signal. We fix the layer underneath: CRM and RevOps, CRM implementation, and go-to-market strategy that ties the plan to reporting people can actually run.

Book a 30-minute audit and we will tell you which of the six numbers your current setup can produce honestly today.