The Board Pack Playbook

Founders of venture-backed companies spend a surprising share of each quarter assembling materials for people outside the company: the board deck, the monthly investor update, the budget, and, whenever the company raises, a fundraising deck and data room. In most companies these are built by hand from exports, and each one is built slightly differently. The result is familiar: hours of founder and finance time per cycle, numbers that drift between documents, and board meetings that open with reconciliation instead of strategy.

This playbook argues for a different operating model. Treat the board pack as a product with one data model underneath. Define a canonical set of ten metrics once, with formulas grounded in widely used investor conventions. Build every external document from those definitions. Reconcile them to the systems of record before anything is shared, freeze what you publish, and keep an audit trail of every change. Do that, and the board deck, the investor update and the data room become three views of the same truth.

The chapters that follow cover what boards and investors focus on at each stage, the metric set with precise formulas and sources, a slide-by-slide deck structure, how the same data serves investor updates and fundraising, the data-discipline practices that make numbers defensible, a data-room checklist, and how Kimo Business Intelligence automates the assembly. A methodology section states what this playbook does not claim, and an appendix provides reference SQL.

Boards and investors look at a growing set of metrics as a company matures, but the underlying questions stay the same: is there real demand, can the company capture it repeatably, and can it do so efficiently enough to reach the next milestone before the money runs out? What changes by stage is the evidence that counts as an answer.

There is no fixed revenue bar for any round. Discussing Y Combinator’s Series A guide, YC partner Aaron Harris noted that YC had seen companies funded with $200,000 of ARR and companies funded with $9 million, and that it is possible to raise on a great story with no metrics or on great metrics with no story.¹Source 1 · TechCrunch, 2020YC just published a 70-page Series A guide so founders don’t tank their own prospectstechcrunch.com The framework below is therefore a map of emphasis, not a set of thresholds.

Seed: learning velocity and cash

At seed, the board is small (often founders plus one or two investors) and the most valuable meetings are working sessions. Metrics matter less as targets than as evidence of learning: which customer segment retains, what usage pattern predicts conversion, how quickly the team ships and measures. Cash is the exception. Net burn and runway should be on every page you send, because at seed the main way a company dies is by running out of time. Andreessen Horowitz calls net burn the true measure of how much cash a company burns each month.²Source 2 · Andreessen Horowitz, 201516 Startup Metricsa16z.com

Series A: repeatability and retention quality

By Series A, investors want evidence that growth is not a founder-led one-off. The MRR bridge shows where growth comes from; retention shows whether it sticks. Net revenue retention captures expansion; gross revenue retention exposes the churn that expansion can hide. Andreessen Horowitz makes the point directly: gross churn estimates the actual loss to the business, while net revenue churn understates losses.²Source 2 · Andreessen Horowitz, 201516 Startup Metricsa16z.com Early unit economics (gross margin and CAC payback) show whether scaling the go-to-market motion will create or destroy value.

Series B: efficiency at scale

At Series B, the conversation shifts to efficiency and predictability. Burn multiple and CAC payback by segment show whether each additional dollar of spend still buys growth. Gross margin trends show whether the cost structure scales. Forecast accuracy (plan vs. actual over several quarters) shows whether management understands its own business. The Rule of 40 enters the conversation as a direction of travel, even though it was framed for much larger companies.³Source 3 · Brad Feld (feld.com), 2015The Rule of 40% For a Healthy SaaS Companyfeld.com

The benchmarks above come from Bessemer’s Scaling to $100 Million⁴Source 4 · Bessemer Venture Partners, 2021Scaling to $100 Millionbvp.com and David Sacks’s original burn multiple essay.⁵Source 5 · David Sacks (Craft Ventures), 2020The Burn Multiplesacks.substack.com Use them as context, not as targets: benchmark populations differ in segment, pricing model and vintage, and the most informative comparison is almost always your own trend.

Translating board questions into metrics

Directors rarely ask for a metric by name. They ask questions, and the board pack should map each recurring question to the metric that answers it, so the answer is on the page before the question is asked.

The canonical set is deliberately small. Every additional metric in a board pack is another definition to maintain and another place for numbers to disagree. These ten answer the questions in the previous chapter at every stage from seed to Series B. Full definitions, edge cases and Postgres queries are in the companion guide SaaS metrics, defined once; the essentials are below.

1–2. MRR and ARR

ARR means annual recurring revenue, not annual run rate. Andreessen Horowitz warns against multiplying a month of recognized bookings or revenue by 12⁶Source 6 · Andreessen Horowitz, 201516 More Startup Metricsa16z.com and says ARR should exclude one-time and professional services fees.²Source 2 · Andreessen Horowitz, 201516 Startup Metricsa16z.com Stripe’s Billing analytics define MRR as the monthly-normalized value of active and past-due subscriptions, excluding taxes, free plans and metered (usage-based) products,⁷Source 7 · Stripe DocumentationBilling analytics: metric definitionsdocs.stripe.com which is why variable usage stays out of the canonical definition. ARR is an operating metric: under ASC 606, revenue is recognized to depict the transfer of promised goods or services to customers,⁸Source 8 · Deloitte DARTRoadmap: Revenue Recognition, 3.1 Objective (ASC 606-10-10-2)dart.deloitte.com so ARR and GAAP revenue will differ, and the data room should explain how. Glossary: ARR, MRR.

3–4. Net and gross revenue retention

Disclosed definitions vary even among public companies. Snowflake’s annual report, for example, describes a net revenue retention rate computed over a two-year window of product revenue for a fixed cohort, with churned customers kept in at zero.⁹Source 9 · U.S. Securities and Exchange Commission (EDGAR), 2025Snowflake Inc. Annual Report on Form 10-K, fiscal year ended January 31, 2025sec.gov The lesson is not to copy any one formula but to state yours precisely. Glossary: net revenue retention, gross revenue retention.

5. Gross margin

Andreessen Horowitz recommends including all costs associated with the manufacturing, delivery and support of the product;²Source 2 · Andreessen Horowitz, 201516 Startup Metricsa16z.com for SaaS that typically means hosting, embedded third-party software, payment processing, support and onboarding. Bessemer puts the average gross margin for cloud businesses at about 65–70%, with the middle half between about 60% and 80%.⁴Source 4 · Bessemer Venture Partners, 2021Scaling to $100 Millionbvp.com

6. CAC payback

Bessemer measures payback against gross-margin-adjusted revenue and includes the renewal and upsell portion of customer success in acquisition cost.⁴Source 4 · Bessemer Venture Partners, 2021Scaling to $100 Millionbvp.com Glossary: CAC payback.

7. Burn multiple

David Sacks introduced the metric in 2020 to measure how much a startup burns to add each incremental dollar of ARR; lower is better, and it should approach zero over time.⁵Source 5 · David Sacks (Craft Ventures), 2020The Burn Multiplesacks.substack.com Use cash burn from the cash ledger, not P&L loss. Glossary: burn multiple.

8. Runway

Show runway with and without undrawn debt, and with the current plan’s burn rather than only the trailing average if a hiring plan is about to change it. Glossary: runway.

9. Rule of 40

Brad Feld wrote up the rule in 2015, crediting it to a late-stage investor: growth rate plus profit should add up to 40%. He framed it for SaaS companies at scale (at least $50 million in revenue) and favored EBITDA as a starting profit measure.³Source 3 · Brad Feld (feld.com), 2015The Rule of 40% For a Healthy SaaS Companyfeld.com Early-stage companies use it as a direction-of-travel indicator. Glossary: Rule of 40.

10. Pipeline coverage

Pipeline coverage is a convention, not a standard. Define which CRM stages count as qualified, whether amounts are probability-weighted and how slipped deals are treated, and keep those rules fixed.

A note on LTV:CAC. Bessemer recommends investing in customer acquisition when LTV/CAC is 3x or higher,⁴Source 4 · Bessemer Venture Partners, 2021Scaling to $100 Millionbvp.com and Stripe estimates LTV as ARPU divided by subscriber churn rate.⁷Source 7 · Stripe DocumentationBilling analytics: metric definitionsdocs.stripe.com Both are reasonable, but a company with eighteen months of history is extrapolating a lifetime from a short window. Report it, with the method, as a supplement to CAC payback rather than a replacement. Glossary: LTV:CAC ratio.

Worked example: NRR and GRR for one cohort

Consider a fictional cohort of four customers that together paid $40,000 of MRR twelve months ago. Customer A expanded from $10,000 to $16,000; B stayed flat at $10,000; C contracted from $10,000 to $7,000; D churned from $10,000 to zero.

The same cohort produces two very different headlines. If customer A had expanded to $24,000, NRR would rise to 102.5% while GRR would stay at 67.5%: the company would “retain” more than 100% of revenue while losing a quarter of its customers’ starting value. This is why a board pack shows both, and why a cohort chart belongs next to the blended rates. See cohort analysis.

Sequoia’s guidance on board decks describes the job well: calibrate the board, because directors do not work in the company every day, then use their experience on the questions that matter. Its recommended agenda for early meetings is a big-picture CEO update with highlights, lowlights and where the company needs help; a calibration section that tells the story with a small set of accurate metrics; company-building topics; a working session; and a closed session. It also advises reusing materials you already run the company with and sending them ahead.¹⁰Source 10 · Sequoia CapitalPreparing a Board Deckarticles.sequoiacap.com

The nine-slide structure below implements that agenda with the canonical metric set. Keep the order stable; directors build a mental model of your deck over several meetings, and stability is what lets them spot change.

Narrative principles

• Headline first. The first slide states the one thing the board must know. Everything else supports it.
• Bad news early. A lowlight surfaced on slide 2 is a discussion; discovered on slide 14 it is a trust problem.
• Explain variance, not values. The board can read a chart. Tell them why the line moved and whether it will continue.
• Every number links to a definition. A footnote or appendix slide lists the definitions and their versions.
• One ask per slide that needs one. Decisions, introductions, hiring help: specific and owned.
• Memo or slides, consistently. Sequoia notes that some companies use written memos instead of slides.¹⁰Source 10 · Sequoia CapitalPreparing a Board Deckarticles.sequoiacap.com Pick one format and keep it.

The board deck template implements this structure in Kimo. For the hands-on build, see Build your board deck from live data.

Writing slide 1: an example

Slide 1 is the hardest to write and the most read. A useful pattern is one sentence for the result, one for the cause and one for the decision. For a fictional company it might read: “ARR reached $6.1M, 4% behind plan, because two enterprise deals slipped into October; expansion offset most of the gap. Net burn was in line and runway is 22 months. We want the board’s view on adding a second enterprise account executive now or after the slipped deals close.” Every number in that paragraph links to a certified measure, and the decision maps directly to the working session on slide 9.

Appendix slides worth keeping

• Metric definitions and versions, so any director can check what a number means.
• Customer concentration: share of ARR from the top 10 customers, trended.
• Cohort tables behind the retention chart.
• Hiring plan detail by function and quarter.
• Budget vs. actual at the account-group level, with variance commentary.

The board deck, the investor update and the fundraising deck answer different questions for different audiences, but they draw on the same metrics. The failure mode is building them separately and letting them drift. The operating model is to treat them as three templates over one dataset.

The investor update

A good update is short: a KPI table, highlights, lowlights and specific asks. It also serves a contractual purpose. Information rights in venture financings commonly include unaudited quarterly statements within 45 days of quarter end, annual statements within 90–180 days of year end, sometimes monthly statements, and a board-approved budget before each fiscal year.¹¹Source 11 · Morrison & Foerster (MoFo ScaleUp)Ask a MoFo: Common Provisions in Venture Capital Term Sheets — Information and Observer Rightsscaleup.mofo.com An update generated from the same certified data as the board deck turns those obligations into a by-product. The investor update template follows this format; the reasoning is in The monthly investor update, rebuilt on live data.

The fundraising deck

Sequoia’s classic outline for a business plan runs: company purpose, problem, solution, why now, market potential, competition and alternatives, business model, team, financials and vision.¹²Source 12 · Sequoia CapitalWriting a Business Plansequoiacap.com The metrics in such a deck should be the canonical ones, computed with the same definitions as the board pack, because the data room will expose any difference. Attention is scarce: DocSend data reported by TechCrunch found investors spent just under three minutes per deck on average in 2022, with product and business model sections drawing the most time.¹³Source 13 · TechCrunch, 2022Looking at 320 pitch decks, here’s what science tells us works besttechcrunch.com Make the traction slide unambiguous and footnote its definitions.

Data discipline is what makes the board pack defensible. Public companies already work under a clear standard: the SEC’s 2020 guidance on key performance indicators asks issuers to give a clear definition of each metric and how it is calculated, to disclose changes in methodology along with the reasons and effects, to consider recasting prior periods when methods change, and reminds them that effective controls matter when material metrics are built from internal data.¹⁴Source 14 · U.S. Securities and Exchange Commission, Federal Register, 2020Commission Guidance on Management’s Discussion and Analysis (Release 33-10751)govinfo.gov Private companies are not bound by it, but it is the bar diligence teams carry in their heads. Five practices meet it.

1. A single source of truth per fact

Assign each fact to one system of record. Subscriptions and MRR come from billing; revenue, costs and margin from the ledger; cash from the bank; pipeline from the CRM; usage from the product database. When systems disagree, the system of record wins and the other is corrected or the difference is explained. Never let a spreadsheet become a system of record by accident.

2. Definitions in a semantic layer, not in slides

A semantic layer stores metric definitions next to the data so every consumer (dashboard, deck, update, AI assistant) computes them the same way. Each definition records its formula, filters, owner and description. The companion post Metrics layers, explained with one MRR definition shows the idea on a single metric.

3. Reconciliation before publication

Before every board meeting, not only before a raise, run the checks that tie the pack to the systems of record. The seven checks in chapter 7 take minutes when automated and save hours of discussion when they catch something.

4. Versioned definitions and recasts

Definitions will change: you will decide to exclude a usage component, reclassify support into COGS, or count reactivations separately. Each change gets a version, a date, an owner and a reason. When a definition changes, recast prior periods so charts compare like with like, and show both methods for one period so the board sees the effect.

5. Frozen snapshots and an audit trail

Live data keeps moving after a quarter closes: late credit notes, reclassified expenses, corrected CRM records. Each published deck should be frozen with its period, the definition versions used and the publication time. If a material correction is needed, publish an erratum showing the old value, the new value and the reason. The audit trail (who changed which definition, who published which deck, who viewed it) answers most governance questions before they are asked.

Ownership: who signs off on what

Discipline needs owners. In a seed or Series A company, one person may hold several of these roles, but each line should still have a name next to it.

A data room is where investors verify the story told in the deck. Law-firm request lists such as Cooley GO’s sample VC due diligence request list show what investors will look for before closing and help founders organize records early.¹⁵Source 15 · Cooley GOSample VC Due Diligence Request Listcooleygo.com Y Combinator’s Series A guidance warns that diligence requests and reference checks can sink a round, and recommends understanding what the investor is trying to learn and preparing customers before calls.¹Source 1 · TechCrunch, 2020YC just published a 70-page Series A guide so founders don’t tank their own prospectstechcrunch.com The checklist below covers the seven folders; the step-by-step build is in Prepare your data room for a fundraise.

Everything in this playbook can be done with a warehouse, careful SQL and discipline. Kimo Business Intelligence exists to make the discipline the default and the assembly automatic, so the team’s time goes into judgment and narrative.

1. 1
   Step 1:

   Connect the systems of record

   Billing (Stripe, Paddle), ledger (QuickBooks, Xero), CRM (HubSpot, Salesforce) and the product database (Postgres and others) connect with read-only credentials. Sensitive databases can stay on your network behind Kimo Bridge: queries run through an outbound-only, mutually authenticated tunnel, and in Bridge mode nothing is stored on Kimo’s side beyond optional short-lived caches you can disable.

2. 2
   Step 2:

   Certify the canonical metrics

   The SaaS metrics template ships the ten definitions as certified measures. Each has an owner, a description, a version history and review-gated changes (measures and dimensions).

3. 3
   Step 3:

   Generate the pack

   In Board decks, the board deck and investor update templates build each slide or section from certified measures and draft commentary on material variances.

4. 4
   Step 4:

   Reconcile and review

   The seven checks run automatically. Publication is blocked until each passes or carries a written explanation.

5. 5
   Step 5:

   Freeze, share, audit

   Published decks become dated snapshots with definition versions attached. Links are scoped to named people, expire by default and are logged; every published deck is filed into the data room.

Between meetings, the same definitions power Ask Kimo: a director’s question (“what was gross retention for customers acquired in 2025?”) is answered from certified measures with the query shown. That consistency is what lets a founder say, with confidence, that the number in the chat, the deck and the data room is the same number.

What stays human

Automation handles assembly, consistency and checks. It should not write your judgment. The headline on slide 1, the explanation of a lowlight, the choice of which decision to bring to the board and the tone of a difficult update are the founder’s job. Kimo drafts commentary on material variances so you start from a page rather than a blank slide, but every published deck carries the name of the person who approved it.

This playbook synthesizes published investor guidance, public-company disclosure practice and vendor documentation with Kimo’s product design. It is not accounting, legal or investment advice. Its limits are worth stating plainly.

• Benchmarks are context, not targets. The Bessemer figures cited come from a specific population and period;⁴Source 4 · Bessemer Venture Partners, 2021Scaling to $100 Millionbvp.com other benchmark surveys use different samples and definitions. Compare your own trend first.
• Definitions are conventions. Apart from GAAP revenue, none of the canonical metrics has a single authoritative formula. This paper picks defensible defaults and stresses documenting them; your investors may prefer variants.
• Some sources are dated. The Rule of 40 (2015), the a16z metric lists (2015) and the burn multiple (2020) remain widely used, but market expectations shift with funding conditions.
• The stage framework is a synthesis. The seed / Series A / Series B emphasis table reflects the cited guidance and common practice, not a survey of investors.
• Illustrative data is fictional. All charts in this paper use illustrative data for fictional companies and are labelled as such.
• Legal documents need counsel. Corporate, equity and contract folders in the data room should be prepared with your lawyers and accountants.

Boards and investors do not want more data; they want numbers they can trust and a founder’s honest reading of them. The practices in this playbook deliver both: a small canonical metric set defined once, a stable narrative that leads with what matters, and data discipline that makes every figure reproducible. Once they are in place, the board deck, the investor update and the data room stop being three projects and become three views of one well-kept model of the business.

Start small. Write the ten definitions this week, build the customer-level MRR table, and run the seven checks before your next board meeting. When you are ready to automate the assembly, start with the board deck template or open Board decks in Kimo. For an example of the outcome, browse our customer stories.

Reference queries for Postgres. They assume a subscriptions table from billing, ledger_lines from the accounting system, cash_movements and cash_balances from the cash ledger, and crm_opportunities. The full set with explanations is in SaaS metrics, defined once.

1234567891011121314151617create or replace view mrr_monthly as
with months as (
    select generate_series(date '2024-01-01',
                           date_trunc('month', current_date)::date,
                           interval '1 month')::date as month_start
)
select m.month_start, s.customer_id,
       sum(case s.billing_interval
               when 'month' then s.unit_amount_cents * s.quantity
               when 'year'  then s.unit_amount_cents * s.quantity / 12.0
           end) / 100.0 as mrr
from months m
join subscriptions s
  on s.started_at < m.month_start + interval '1 month'
 and (s.ended_at is null or s.ended_at >= m.month_start + interval '1 month')
where s.is_trial = false and s.pricing_model = 'licensed'
group by m.month_start, s.customer_id;

123456789101112131415161718create or replace view mrr_bridge as
with pairs as (
    select coalesce(cur.month_start, (prev.month_start + interval '1 month')::date) as month_start,
           coalesce(prev.mrr, 0) as prev_mrr,
           coalesce(cur.mrr, 0)  as cur_mrr
    from mrr_monthly cur
    full outer join mrr_monthly prev
      on prev.customer_id = cur.customer_id
     and prev.month_start = (cur.month_start - interval '1 month')::date
)
select month_start,
       sum(case when prev_mrr = 0 and cur_mrr > 0 then cur_mrr else 0 end)                as new_mrr,
       sum(case when prev_mrr > 0 and cur_mrr > prev_mrr then cur_mrr - prev_mrr else 0 end) as expansion_mrr,
       sum(case when cur_mrr > 0 and cur_mrr < prev_mrr then prev_mrr - cur_mrr else 0 end) as contraction_mrr,
       sum(case when prev_mrr > 0 and cur_mrr = 0 then prev_mrr else 0 end)                as churned_mrr
from pairs
where month_start <= date_trunc('month', current_date)::date
group by month_start;

12345678910with base as (
    select customer_id, mrr from mrr_monthly
    where month_start = (date_trunc('month', current_date) - interval '13 months')::date
), today as (
    select customer_id, mrr from mrr_monthly
    where month_start = (date_trunc('month', current_date) - interval '1 month')::date
)
select round(sum(coalesce(t.mrr, 0)) / sum(b.mrr), 4)               as nrr,
       round(sum(least(coalesce(t.mrr, 0), b.mrr)) / sum(b.mrr), 4) as grr
from base b left join today t using (customer_id);

123456789101112131415161718with burn as (
    select month, -sum(amount) as net_burn
    from cash_movements
    where category in ('operating', 'capex')
    group by month
)
select date_trunc('quarter', b.month)::date as quarter,
       round(sum(b.net_burn)
             / nullif(sum(12 * (r.new_mrr + r.expansion_mrr - r.contraction_mrr - r.churned_mrr)), 0), 2)
           as burn_multiple
from burn b
join mrr_bridge r on r.month_start = b.month
group by 1 order by 1;

select round(max(cb.balance) / nullif(avg(b.net_burn), 0), 1) as runway_months
from cash_balances cb
join burn b on b.month > cb.month_end - interval '3 months' and b.month <= cb.month_end
where cb.month_end = (select max(month_end) from cash_balances);

123456789select b.month_start,
       b.billing_mrr,
       l.ledger_sub_revenue,
       round((l.ledger_sub_revenue - b.billing_mrr) / nullif(b.billing_mrr, 0), 4) as gap_pct
from (select month_start, sum(mrr) as billing_mrr from mrr_monthly group by 1) b
join (select period_month as month_start, sum(amount) as ledger_sub_revenue
      from ledger_lines where category = 'subscription_revenue' group by 1) l
  using (month_start)
order by b.month_start;