What marketing KPIs should a CMO report to the board?

Report 8 to 10 metrics at most, each one traceable to the accounts and measured against a no-spend baseline. A workable page: incremental contribution from marketing, customer acquisition cost (CAC) payback, pipeline by opportunity stage, net revenue retention, blended CAC by market, share of branded demand, test coverage and forecast accuracy. Put the first 4 at the top, because a busy director may read only those. MQL counts, platform ROAS and click-through rate stay in the weekly marketing meeting, where they help the team steer spend.

How do you move a board off MQLs without looking evasive?

Run the old and new metrics side by side for 2 quarters before retiring anything. Show MQLs and platform ROAS in a clearly labelled legacy block, with pipeline by stage and tested incremental contribution next to them, and restate the last 12 months under the new definitions. Tell the board in advance why the page is changing and which numbers finance has reconciled. A switch the CFO co-signs reads as more rigour. A switch that appears without warning reads as hiding a bad number.

Should customer acquisition cost include sales costs?

In a sales-led B2B business, yes. A CAC figure built only from marketing's own costs understates the real cost of winning a customer, and a CFO will spot the missing sales salaries and commissions at once. Report fully loaded sales and marketing CAC as the board number. Illustrative: marketing cost of 400 and sales cost of 200 per new customer make the board CAC 600, not 400. If you also want a marketing-only CAC for internal steering, keep it in the exec pack and label it clearly. Consumer businesses without a sales team can use marketing cost, but should subtract launch discounts from the margin side.

How do you split centrally bought marketing spend across markets?

Agree one allocation rule with finance and apply it for a full 12 months. Common bases are each market's share of revenue, of new customers or of website sessions, applied to shared costs such as a global site, central tools or a regional ad buy. The choice matters less than consistency: a rule that changes mid-year makes quarter-on-quarter CAC meaningless. Write the rule into the scorecard's footnote so a director can see how the per-market acquisition cost was built and challenge it if needed.

How do you measure branded demand without counting existing customers?

Strip navigational searches before you count. Existing customers searching your brand plus login, support, invoice or app are not new demand, and in a subscription business they can dominate brand search volume. Filter those query patterns out of your own count in Google Search Console, then compare the remaining brand searches with competitors' brand terms from a keyword tool or Google Trends. Report the share quarterly and watch the trend over at least 12 months rather than any single month; a launch or a news spike distorts short periods.

How do you calculate CAC payback for a board report?

Divide fully loaded acquisition cost per new customer by the monthly gross margin that customer generates. For example, an illustrative CAC of 600 and monthly gross margin of 50 per customer gives a 12-month payback. Fully loaded means media, tools, agency fees and the salaries spent on acquisition, plus sales cost in a sales-led business, all agreed with finance in advance. Report it quarterly on a trailing 12-month basis, by market if you run several, so the board can see whether growth is funding itself.

What is test coverage and why should the board see it?

Test coverage is the share of media spend that sat under an incrementality test, such as a geo holdout, in the last 12 months. It tells the board how much of the reported impact is measured evidence and how much still rests on platform attribution or model assumptions. Showing low coverage honestly builds more trust than a board discovering it later. It also gives marketing a fundable goal for the year: raise coverage, starting with the largest spend lines that have the weakest evidence.

Should marketing mix modelling results go in a board report?

Yes, when the model is calibrated with experiments and shown with a confidence range. A marketing mix model estimates how revenue responds to spend across channels, including carryover effects that a same-quarter view misses. An uncalibrated model is correlational and easy for a sceptical director to pick apart, so pair it with holdout tests and label which figures came from which method. Present the model's output as a range, and say plainly which channels have been tested in the last 12 months and which rest on the model alone.

How often should marketing report KPIs to the board?

Quarterly for the board. The CEO and CFO get a monthly pack, and marketing works from weekly dashboards. The board page should change slowly: stable definitions, the same row order each quarter, and a restated history covering the prior 12 months whenever a definition has to change. That stability is part of what makes the numbers credible. Anything that needs weekly attention, such as platform ROAS or cost per lead, belongs in the operational tier, where the team can act on it quickly without the board reading noise as a trend.

Marketing KPIs for board reporting that CFOs trust

Which marketing KPIs for board reporting do boards trust?

Our test for marketing KPIs for board reporting: the number ties to cash or survives a counterfactual, and the headline rows do both. This guide is for the CMO or marketing lead who puts one page in front of a board and a CFO: by the end you can pick its 8 rows, decide what comes off, and score your current deck.

In plain terms, finance can trace it to the accounts, or it shows what would have happened without the spend. Marketing-qualified lead (MQL) counts and platform-reported return on ad spend (ROAS) fail both tests.

The survey data shows the stakes. The CMO Survey 2026 (Duke University's Fuqua School of Business, with Deloitte and the American Marketing Association) surveyed 308 senior US marketers from 7 to 29 January 2026. Its report of 31 March 2026 found:

marketers rate the CFO 4.5 out of 7 as a partner in building the business case for marketing spend, up from 4.3 over four years;
56% report increasing CFO pressure to prove value and 45% board pressure, down from 63% and 50% in 2025;
when profits miss, 53.1% of companies focus on cutting expenses, and marketing is then cut 45.4% of the time, more often than other categories.

These are US marketers' own ratings: a benchmark for the problem, not a reading of your CFO.

What makes a marketing KPI board-grade?

A KPI is board-grade when it passes two tests. Forrester principal analyst Simon Daniels described the second one on 21 November 2023: "establishing a baseline and measuring the difference above this, a concept we call 'marketing lift.'"

Test one, the cash tie. Could a finance analyst trace the number to the management accounts without a marketer in the room? Customer acquisition cost (CAC) payback passes. Engagement rate does not.

Test two, the counterfactual. Platform conversions report what happened after someone saw an ad; a counterfactual asks what would have happened without it. A number passes if it comes from a controlled test or calibrated model (the core of marketing analytics and insights work), or is a cash measurement no ad platform grades.

Our sort, not an industry standard. Branded demand and test coverage are context rows.

The bottom-left box still has uses. Platform ROAS is a fine signal for a media buyer moving budget on a Tuesday. It fails on a board slide because the board cannot see what it leaves out.

What should a one-page board scorecard for marketing include?

Eight rows make a workable set of marketing KPIs for board reporting, and never more than ten. A director reading at 11pm gets through four rows, so those carry the argument.

The one-page marketing board scorecard (template)
Metric	Definition	Cadence	Why a CFO trusts it
1. Incremental contribution	Gross margin that would not exist without marketing, from holdout tests and a calibrated marketing mix model (MMM)	Quarterly, with a range and method note	A margin estimate with a stated baseline, labelled as a model, not a ledger line
2. CAC payback	Fully loaded acquisition cost divided by monthly gross margin per new customer. Sales-led B2B adds sales cost	Quarterly, trailing 12 months	It speaks cash conversion
3. Pipeline by stage	Sales-accepted opportunities, counted once per buying group, with stage conversion rates	Monthly to exec, quarterly to board	Reconciles to the CRM and the sales forecast
4. Net revenue retention (NRR)	Revenue this year from last year's customers, divided by their revenue last year	Quarterly	Straight from the ledger
5. Blended CAC by market	Marketing cost in a market divided by new customers won there, under a written rule for splitting shared spend	Quarterly	No attribution model; one allocation rule, agreed once
6. Share of branded demand	Your non-navigational brand searches as a share of the category's brand searches	Quarterly	A preference signal no ad platform grades
7. Test coverage	Share of media spend under an incrementality test in the last 12 months	Quarterly	Shows how much of the page is evidence
8. Forecast accuracy	Last quarter's marketing revenue forecast against actual, as a percentage error	Quarterly	Earns belief in everything else

Three rows need a word. "Fully loaded" CAC covers media, tools, agency fees and acquisition salaries; in sales-led B2B, a payback figure without sales cost fails finance's first question. For CAC by market, head-office spend on a global site or central ad buy must be split: agree the rule with finance once and keep it all year.

For branded demand, strip navigational searches first. Customers typing your brand plus "login" or "support" already bought from you.

Retention earns its row even if customer success owns it: The CMO Survey found retention grew 12.8% over the prior 12 months against 7.4% for acquisition, while acquisition budgets still run 26% larger.

CMO Survey 2026, US respondents: growth over the prior 12 months. Acquisition budgets still run 26% larger.

Which rows change with your business model
Row	Sales-led B2B	Subscription or SaaS	Consumer or ecommerce
2. CAC payback	Marketing plus sales cost	Marketing plus sales cost	Marketing cost, margin after discounts
3. Pipeline	Opportunities per buying group	Trials to paid, by cohort	First to second purchase rate
4. Retention	Renewal and expansion revenue	NRR	Repeat-purchase rate at 90 days

Do this first

Every board row is either a ledger figure finance can recompute within one working day, or a tested estimate labelled with its method, baseline and range. Anything that is neither goes to the appendix.

For the plumbing (event design and a BigQuery export finance can query), see our guide to GA4 reporting at a board-grade cadence. The marketing ROI calculators post has working versions of the payback maths.

Which marketing metrics cost CMOs credibility with the board?

The damaging ones look like revenue evidence without being evidence, and each invites the CFO to do the counterfactual maths less kindly than you would.

Metrics to move off the board page, and what replaces them
Metric	Why the board discounts it	Replace with
Platform-reported ROAS	Each platform credits itself, so the platforms together can claim more revenue than finance booked	Incremental ROAS from a holdout, turned into margin
MQL volume	Counts people, not deals, against a threshold marketing sets	Pipeline by stage
Cost per lead	Falls when lead quality falls	Cost per sales-accepted opportunity
Reach, followers, sessions	No stated path to cash; sessions can drop when AI answers remove the click	Share of branded demand
"Influenced" revenue	If every touch counts, nearly all revenue qualifies	Lift above a baseline finance agreed
"Sourced" revenue	Credits marketing alone for deals sales and renewals also won	Lift above a baseline finance agreed

Ad platforms are not lying. They count conversions in their own attribution window, including a returning customer who would have bought anyway. Add three platforms together and the total can exceed what finance booked.

Swapping "influenced" for "sourced" does not fix this. Daniels is blunter in the same 21 November 2023 Forrester post: "We often say that marketing-sourced metrics are the fastest way for a CMO to get fired." A sourced number invites sales to dispute every deal; lift against a baseline both teams agreed leaves little to dispute. The same logic applies to AI tooling spend, covered in our piece on AI marketing ROI proof a board will accept.

The warning

Do not swap board metrics overnight. A board used to MQLs and ROAS reads a sudden switch as dodging. Show the old and new rows side by side for two quarters, with a restated history, then retire the old ones.

Is the MQL dead? Demand generation vs lead generation for the board

As a board metric, yes. Forrester's "Saying Goodbye To MQLs" (21 November 2023) states: "Our research has shown that fewer than 1% of leads convert to closed deals."

The reason is buying groups. The same post reports that 46% of B2B buyers bought in a group of two to three people and 28% in a group of four to nine. If four people in one purchase each become an MQL, the count rises fourfold while revenue stays flat.

Lead generation optimises for captured contacts. Demand generation optimises for buyers who already prefer you when they arrive, and preference is where the board case sits. In a 25 September 2026 post, Forrester analysts Kelvin Gee and Ian Bruce cite the firm's buyer research: "the initial preferred vendor wins the business 55% of the time on average". A dashboard that starts at the form fill joins that race halfway through.

Keep the MQL in the CRM to route sales follow-up; stop reporting it upward. Forrester also relays a client-reported result (not a controlled study): delivering three or more buying-group members to sales brought "a 50% increase in their conversion rates from meetings to closed-won."

How do you prove marketing's impact with a counterfactual?

Measure against a no-spend baseline: holdout tests prove single channels, and an MMM calibrated by them covers the rest. The board needs to know which method produced each number.

Counterfactual methods and where each belongs
Method	What it answers	Weak spot	Board use
Holdout or geo test	What one channel added against a matched group that did not see it	One channel per test; needs volume	Headline evidence for the biggest spend lines
Marketing mix model	How revenue responds to spend across channels, with lagged effects	Correlational unless calibrated by tests	The incremental contribution row
Platform lift study	Whether one platform's ads changed behaviour	Run by the party being measured	Supporting evidence, labelled as such

Lag matters too. The CMO Survey puts the median duration of marketing's impact on customers at six months, longer than in 2022, so a metric that counts only same-quarter revenue can undercount brand and retention work. Our incrementality testing guide covers test design, and the Google Meridian MMM guide shows a model that takes test results as priors.

Illustrative example (made-up round numbers)

A retailer's paid social account reports ROAS of 5.0. A four-week geo holdout switches the channel off in matched regions, and sales there fall by less than the platform implied: incremental ROAS is 2.1. At a 40% gross margin, each unit of spend returns 0.84 units of margin.

The board page does not say "5.0x". It says "incremental contribution: minus 0.16 per unit of spend, tested in Q3", with a plan to cut retargeting frequency and re-test. Uncomfortable, and it gets the next positive number believed.

Is your board deck ready? Score it in two minutes

Tick what is true of your current board page. 8 or more: send it; 5 to 7 means fix the gaps with finance first; 4 or fewer means use the 30-day rebuild below.

How to rebuild your marketing KPIs for board reporting in 30 days

Start with finance, not a data platform, and be ready to show a smaller, truer number.

Week 1: audit, then book the CFO. Sort every metric in the last deck with the two tests. Get finance to sign written definitions of CAC and pipeline, plus the allocation rule for shared spend.
Week 2: reconcile to the ledger. Pull four quarters of cost and new-customer counts from finance systems. If platform revenue exceeds booked revenue, the gap is your case for testing. With sales operations, switch pipeline to opportunities per buying group.
Week 3: design the first holdout. Pick the biggest spend line with the weakest evidence, often Google Ads branded search or retargeting.
Week 4: draft the page and a forecast. Fill the eight rows, add a next-quarter forecast you will be graded on, and put the old metrics beside the new for the two-quarter changeover. Send it to the CFO before the board sees it.

Teams skip the last step, and it matters most: a CFO who has already challenged the numbers sits beside you in the meeting, not across from you.

Leapbuzz analytics and measurement work in Singapore and our four other markets follows this sequence. Near budget season, pair it with marketing budget planning when AI answers first.

This week, book 45 minutes with your CFO and bring one thing: the demotion table above, filled in for your current deck. Any row you cannot tie to the ledger or to a named test by the end of that meeting comes off the board page next quarter.