Marketing does not become part of the P&L because someone decides it should. It becomes part of the P&L when four layers are in place, in a fixed order — and the order is the whole point.
I am Anna Melnova, CMO at Lesta Games. I have built this function in games, on marketplaces, and across Russia, the CIS and MENA. P&L marketing, in my definition, is marketing accountable for contribution margin in the profit and loss statement rather than for spending a budget. The build order turned out to be the same everywhere, and I call it the four-layer rule: data quality → alignment with the data and finance functions → a model and metrics matched to the media mix → and only then accountability for the P&L. A layer skipped in that sequence always comes back as the bill for a wrongly optimised metric.
In practice, "putting marketing on the P&L" often amounts to a single conversation with the team: from now on, you own revenue. That is a transfer of accountability with no infrastructure underneath it. The connection to profit only holds once the system has been assembled layer by layer, with nothing skipped.
Why marketing deserves to be treated as an investment rather than a cost line is a separate argument, and I have made it elsewhere. This piece is about the how.
Layer one: data quality
Every attempt to connect marketing to a financial result starts with data — more precisely, with its quality. Volume stopped being the constraint years ago. Garbage in still means garbage out, and no dashboard repairs it downstream.
It comes down to details. Do all teams use one definition of a new user? Are ad networks claiming traffic they did not create? That is a question of how attribution is configured. Is the event counted as a conversion something the business actually cares about, or something shallow and early? How are cohorts formed, and how is payback calculated inside them? An error in any one of these settings distorts the entire picture, and the whole hypothesis–test–result–scale-up chain then moves in the wrong direction.
My conclusion from that number is blunt: no attribution or payback model repairs dirty data underneath it. The Gartner estimate dates back to 2020, and the same page carries a second figure that explains the first — 59% of organizations do not measure data quality.
How poor measurement simulates efficiency — buying your own organic traffic, one-and-done buyers, pseudo-metrics — I have covered in the piece on ROI, so I will not repeat it. For P&L purposes one point matters: fix the data first. The choice of more advanced metrics can wait, because without clean data no metric means anything.
Layer two: why the CMO's closest partner is the CDO
Here is the part the market underrates. Once marketing becomes a quantitative discipline, the question is no longer whether the team can do arithmetic. It is how good the collection and processing system is — and therefore whether that system has an owner.
On the data side that owner is the CDO, and the remit is foundational: data quality and availability, connected systems and identifiers, correct event logic, reproducible calculations. Attribution and reconciliation with the P&L are built by three functions together — data, marketing and finance. Without a strong partner on the data side, a correct system is close to impossible to build. The CDO supplies the ammunition: quality data and transparent measurement logic, without which strong decisions cannot exist. One caveat on that metaphor — this is not a service desk for marketing, it is the other half of a single decision loop.
In different operating models the owner of that foundation may be a CDAO, a CIO or product analytics rather than a CDO. The title matters less than the fact that the foundation has one named owner. Business definitions, however — what counts as a conversion, what counts as an incremental effect — stay with the CMO and are not delegated to the data function.
Who was surveyed matters as much as the headline number: the 2025 survey covered executives from 125 Fortune 1000 and leading global organizations, while the 2026 survey drew senior AI and data executives from nearly 110 leading companies. In companies of that size a data leader is usually already in place; the practical question is how well the alignment works.
The third pillar is the CFO, and here I will say the unpopular thing: the partnership splits by domain. The CDO is the primary partner on the measurement system — without one, a P&L conversation is technically impossible. On the P&L itself, though, the final word belongs to finance. Finance owns the methodology and decides which number counts as true. As long as marketing and finance hold different definitions of revenue, variable cost and payback window, any marketing dashboard remains marketing's internal reporting. A simple maturity test: three people should be willing to sign off on one campaign payback number — the CDO for data correctness, the CMO for interpretation, the CFO for methodology. If only marketing signs off, nothing has been built into anything.
This alignment always has two sides, and the second is underrated even more heavily: a strong CDO does not guarantee strong marketing. It is not enough for the data team to produce reports for marketing. A modern CMO cannot be someone who receives clean numbers and takes them at face value. The role requires understanding the system deeply enough to ask the right questions, to see where a metric lies, to know the limits of the attribution model, and to avoid becoming a hostage to false reporting.
That is where the new boundary of the profession runs. A strong CMO used to need strategy, brand leadership and a fine read on the market; today that is not enough — though the bar is not SQL fluency, only enough understanding of the data machinery to tell a real effect from noise. The quality of the answer is capped by the quality of the question: the CDO provides the environment (the right data, clean context, a working architecture), the CMO puts the right question to it, and a weak link on either side collapses the pair. What suffers is not the reporting but the quality of decisions.
Layer three: the model and the dashboard, with no universal metric set
Once the data holds, modelling follows. A company needs more than an assessment of past results; it needs to understand how changes in marketing will affect what comes next — channel valuation, budget planning, payback.
There is no universal metric set for a CMO dashboard, and anyone promising five essential metrics for every situation is bending the truth. The set depends on two things — the media mix and what attribution can actually see — and the choice between three measurement regimes comes down to one question: what share of media spend sits inside reliable attribution?
When almost all spend goes into trackable channels, cohort payback measures carry the dashboard: CAC payback window, LTV to CAC on a fixed horizon, contribution margin after marketing cost. When a meaningful share of the mix is structurally hard to attribute — brand placements, creators, out-of-home — the centre of gravity shifts to blended measures: blended CAC, marketing cost as a share of revenue, the organic-versus-paid trajectory. In the third case, where the unattributable share dominates, the only honest route is econometrics (marketing-mix modelling) and geo experiments: switch a channel off in some regions and read the difference. One clarification worth making: an "unmeasurable" channel almost always means "poorly attributed by click-based models" rather than "impossible to evaluate". Incremental brand effect is measurable by experiment — just not in a weekly report.
The second discipline is to keep two management views separate. ROAS expresses revenue per unit of advertising spend and answers whether media buying works. ROMI is calculated differently: subtract full marketing cost from incremental margin before marketing, then divide by that same cost — margin-based, incremental, and inclusive of the whole function's cost rather than media alone. Without incrementality in the numerator, the ratio will happily credit marketing with revenue the company would have earned anyway.
The difference is not academic. At a ROAS of 4 with 20% margin before ad spend, every $1 of advertising produces $0.80 of margin: the campaign loses $0.20 before any other cost of the function. What belongs in the report is not the ratio but contribution margin after marketing cost. ROMI remains a management measure inside the business case; ROAS remains the operational metric used to steer media buying. Why the industry so often presents ROAS as a business result I have covered in the piece on markets with no bottom of the funnel.
The third element is non-negotiable: the budget is derived from the financial goal, not from last year's budget plus indexation. The order runs from a contribution-margin target and a horizon, to the contribution-margin LTV of a new cohort on that horizon, to allowable CAC and a payback-window check, to the number of new customers required and the full acquisition budget, and finally to a feasibility check against marginal CAC and conversion rather than last period's average rate. Volume and CAC influence each other, so the scenario is always recalculated iteratively. Assembled that way, the conversation with finance is about the cost of the goal rather than a spending cut.
Then there is the balance between acquisition cost and customer value. On the surface it looks obvious: push cost down, push return up. In practice the task is to optimise the economics of the whole customer relationship — LTV to CAC on a horizon fixed in advance — rather than the price of a single install. That requires a model that accounts for user behaviour over time. Once marketing works in that logic, it inevitably starts influencing the product, because it becomes visible where value leaks: in acquisition, onboarding, retention or monetisation.
The 3:1 rule was born in SaaS, with subscription revenue and predictable churn. In mobile games and on marketplaces, where revenue is transactional and the LTV distribution is heavily skewed by a small group of payers, the number itself matters less than the horizon you calculate it on. In those businesses I always pair the ratio with a second measure: CAC payback in months. Bessemer's own cloud benchmarks hold both (the data covers its cloud portfolio from 2010 to the first half of 2021). The firm recommends investing in acquisition at a CLTV/CAC of 3x and up, and sets the payback window as a separate threshold: under 12 months for companies serving SMBs, under 18 for mid-market, under 24 for enterprise.
Layer four: where P&L accountability is justified, and where it is not
Marketing should not be built into the P&L everywhere. It depends on how much of the revenue the function actually creates.
In my experience, cases where a CMO owns revenue or profit directly are still rare. Digital businesses and startups get there first, because the link between marketing investment and result is direct and visible. In mobile games the entire economy is built around acquisition: how many users came in, how they paid back, what revenue they produced. There, marketing effectively runs the top line.
Large companies with a strong product and heavy organic demand are a different case: too many factors beyond marketing move revenue, and marketing's contribution is often not decisive. Where marketing is the primary source of growth, connecting it to the P&L is logical. Where the contribution is modest, or the function controls neither pricing nor retention nor variable cost, handing over the whole P&L is wrong — but accountability for the contribution margin of a channel, segment or region it does control still makes sense.
I separate the two cases with three questions rather than instinct. First: what share of new revenue comes from paid acquisition? Second: does marketing have real levers on price, assortment and retention, or only on the top of the funnel? Third: would the attribution survive an experiment — are you prepared to switch a channel off in some regions and see the revenue difference? If any one of the three has no clear answer, a conversation about the whole P&L is premature.
People and culture: what the team has to be ready for
Building marketing into the P&L changes more than process. It changes what the team has to be, and that is the part most reorganisations underestimate.
Marketing becomes a far more analytical function. People need to read data, work with metrics and see cause and effect. Sensing a good insight and assembling an elegant media plan are no longer enough. The job requires understanding funnel logic, telling volume from quality, and arguing with data rather than merely looking at it.
None of which cancels creative work — if anything it puts a premium on it. The main risk in a P&L transition is over-correcting into pure analytics. A team pushed that way starts optimising whatever is easy to measure and loses the ability to find non-obvious answers. The paradox is that a poor data environment is exactly what makes marketing conservative: when the cost of a mistake is unclear, risk feels reckless. A good measurement environment does the opposite — it allows bolder bets, faster tests and more honest admissions of failure. The balance sits where data supports decisions without replacing thought.
On evaluating teams, the industry has no single system that works well, and I have not personally seen one laid out convincingly anywhere. Smaller teams can carry individual channel targets. Larger ones tend to run team-level metrics separately for performance, brand and CRM, which is what makes it possible to tie goals to the budget model. Some brand campaigns still run on nearer-term objectives such as reach or engagement, because tying each execution directly to revenue is not always operationally possible. That is fine: not every marketing action has to be tied to a revenue line straight away.
My working principle: one financial metric for the whole function — contribution margin after marketing cost — and operational targets that each team genuinely controls. A metric someone cannot influence contributes nothing to their objectives except cynicism.
What "built" actually means
Put together, marketing built into the P&L is not a team that has been told to own revenue. It is a function that knows where and how it moves profit, can see which of its actions produce a real business result, can calculate payback across the whole mix, and operates in an environment where data quality supports decisions at business level.
That system assembles in a fixed order — the four-layer rule: data and its quality first, then alignment with the data and finance functions, then modelling and metrics matched to the media mix, and only then accountability for the P&L. Order beats speed here. A transition started from the wrong end, with accountability but no infrastructure, usually ends with marketing optimising an elegant but incorrect metric. And the line marketing ultimately puts its name to is always the same one: contribution margin after marketing cost, not revenue, and not a ROAS number in an ad platform.
So what does marketing's P&L accountability rest on in your company: infrastructure, or an agreement that marketing now owns revenue?
Sources: ¹ Gartner, Data Quality: Why It Matters and How to Achieve It (estimate based on Gartner research from 2020) · ² Data & AI Leadership Exchange, 2026 AI & Data Leadership Executive Benchmark Survey and 2025 AI & Data Leadership Executive Benchmark Survey · ³ David Skok, Startup Killer: The Cost of Customer Acquisition; Bessemer Venture Partners, Scaling to $100 Million
Russian version of this article: Как построить маркетинг, встроенный в P&L