Strategy ·

How to present MMM results to your CFO

How to present MMM results to your CFO: line up revenue with finance, turn Modeled ROAS into profit, show uncertainty, and bring scenarios they'll approve.

Linnea Zielinski · 13 min read

Listen
0:00 / 0:00
AI-generated audio
How to present MMM results to your CFO

When a house goes on the market, the seller sets a listing price, and it's usually optimistic. The buyer's lender doesn't put much stock in that number. Before approving a loan, the bank sends an appraiser to figure out what the house is actually worth, and the loan gets sized to the appraisal.

Marketing budgets work a lot like that. Ad platforms hand you the listing price: their own reported revenue and ROAS, which tend to be generous and often count the same sale more than once. Your marketing mix model is the appraisal. And your CFO is the lender, deciding how much capital to put behind your plan based on which number they trust.

When budgets get tight, the marketing teams that keep (and grow) their spend are the ones that can walk finance through the appraisal in finance's own language: how much revenue marketing actually added, what that's worth after costs, and how confident anyone should be in the forecast.

Key takeaways

  • CFOs care less about which channel "won" and more about how much revenue marketing added, what it's worth after costs, and how much risk comes with spending more.
  • Line up your revenue and spend definitions with finance before the meeting, so the discussion is about decisions instead of whose numbers are right.
  • Expect Modeled ROAS to differ from platform-reported ROAS, and be ready to explain why: platforms overlap, reward the campaign that captured demand, and can't see halo effects.
  • Translate Modeled ROAS into profit using your contribution margin, and compare every channel to your break-even ROAS.
  • For omnichannel brands, include Amazon and retail revenue, since a website-only view can make your best demand-building campaigns look like your worst.
  • Show uncertainty in dollars and be upfront about where the model is less confident, especially at spend levels you've never tested.
  • Bring scenarios, a staged rollout, and a definition of success instead of a single budget request.

Your CFO reads MMM results differently than your team does

Your team opens the dashboard asking which campaigns are working. Your CFO opens the same report asking whether marketing is the best place for the company's money, compared with inventory, hiring, or paying down debt. It's the same data viewed through a different set of questions.

What your team asksWhat your CFO is really asking
Which channel has the best ROAS?Which channel can take on more budget and still pay for itself?
How much revenue did marketing drive?How much revenue would we lose if we cut this budget?
Should we spend more on Meta?Will more Meta spend cover its costs after margin?
Is our awareness campaign working?When does it show up in revenue, and how sure are we?
Can we hit our number?What will it cost, what are we assuming, and what happens if we're wrong?

None of the CFO's questions are about how the model works. Lead with decisions and dollars, and keep modeling details ready for when someone asks.

Get your numbers to line up with finance before the meeting

The fastest way to lose a CFO's attention is to show a revenue figure that doesn't match the one they already report to the board. Before you present, sit down with someone on the finance team and agree on three things.

Revenue definitions

Your MMM and your finance team may measure revenue differently. Check these before anyone else does:

  • Gross vs. net: Is the revenue in your MMM before or after discounts, returns, and refunds?
  • Timing: Are you counting orders when they're placed, while finance counts revenue when it's recognized?
  • Storefronts: Does your model include Amazon and retail sales, or only your website? If finance reports total revenue and your MMM only sees Shopify, the numbers will never match.

Small differences are fine as long as you can explain them. What hurts you is a gap you didn't know was there.

Spend definitions

Your MMM measures media spend, but finance tracks total marketing expense, which also includes agency fees, creative production, and software. Make that split clear, especially when you propose cutting a channel, since trimming its media budget doesn't make the agency retainer or creative costs go away.

Margin

Finally, ask finance which contribution margin they'd use to judge a marketing investment. That's revenue minus the costs that rise with each sale, like product, shipping, payment fees, and returns. You'll need it in the profit math below, and when your CFO helped choose the number, they're far less likely to argue with the conclusions built on it.

Explain why MMM revenue doesn't match platform reports

Your CFO has probably seen platform reports for years, so the first question is usually why the MMM's numbers look different. In Prescient's platform, you'll see these side by side: the revenue ad platforms report next to MMM Revenue, which is the revenue our marketing mix model estimates your campaigns actually drove. MMM Revenue is made up of two parts:

  • Base revenue: revenue from people who engaged directly with an ad
  • Halo effects: revenue a campaign drives indirectly through another channel, like a shopper who saw a Meta ad, kept scrolling, and later searched for your brand or bought on Amazon

The gap between platform-reported revenue and MMM Revenue usually comes from a handful of places. Here's how to explain each one in plain terms.

Why the numbers differWhat's happening
Double countingMeta and Google can both claim the same purchase, so adding up platform revenue overstates the total.
Capturing vs. creating demandRetargeting and branded search get credit for buyers whose interest started somewhere else.
Attribution windowsEach platform decides for itself how long after a click or view a sale still counts.
Limited visibilityPlatforms only see activity on their own turf, so they miss halo effects on other channels and storefronts.

Some channels will look better in the MMM and some worse, so present both. It also helps to frame platforms as answering a different question rather than being wrong: they report the sales their ads touched, while the MMM estimates the sales your ads actually added.

Turn Modeled ROAS into a profit conversation

Revenue is where most marketing reports stop, and it's exactly where your CFO's analysis starts. Prescient reports Modeled ROAS, which is MMM Revenue divided by spend. But a 3x Modeled ROAS can be a great investment for one brand and a money-loser for another, and the difference is margin.

Find your break-even ROAS

Using the contribution margin you agreed on with finance, break-even ROAS is simply 1 divided by your contribution margin. Any channel below that line costs more than it brings in, at least before you count the longer-term value of new customers.

Contribution marginBreak-even ROAS
30%3.33x
40%2.50x
50%2.00x
60%1.67x
70%1.43x

This is why ROAS benchmarks from other brands don't mean much to a CFO. A 2x Modeled ROAS clears the bar for a brand with 60% margins and loses money for a brand at 40%.

Show the profit math

A worked example turns this from theory into something finance can check line by line. Say a channel spent $100,000, it has a 3.0x Modeled ROAS, and your contribution margin is 45%:

  • MMM Revenue: $300,000
  • Contribution before ad spend: $300,000 × 45% = $135,000
  • Profit after ad spend: $135,000 − $100,000 = $35,000

That's a 35% return on the ad spend, a very different headline than "3x." Run this math for your biggest channels and put it next to Modeled ROAS in your deck. If a channel brings in new customers who tend to buy again, make that case separately using finance's retention data.

Show the full revenue picture, not just your website

For omnichannel brands, this is where MMM results surprise CFOs most. In an analysis of 25 retail brands, Prescient found retail accounted for more than half of total revenue on average, yet a lot of marketing measurement only looks at website conversions. If your reporting only sees Shopify, your demand-building campaigns will look weaker than they are, because much of what they drive shows up at your retail partners or on Amazon.

When you present, make sure your CFO can see:

  • MMM Revenue by storefront, so they can match it against the revenue lines they already track
  • Retail halo effects from campaigns that never sent anyone to your website
  • Amazon revenue driven by ads that ran elsewhere, like Meta or CTV

Revenue through retail partners often carries a different margin than a direct-to-consumer sale, so apply the right one to each storefront.

Be ready for the retargeting question

Retargeting often looks less efficient in the MMM than in platform reports. That's because Prescient credits the campaign that created the demand, not just the one that captured it. Retargeting still drives sales, and prospecting drives its own conversions, both directly and through halo effects. The point for your CFO is that cutting prospecting to protect "high-ROAS" retargeting would shrink the pool of shoppers retargeting depends on.

Show where more spend will (and won't) pay off

Historical ROAS tells your CFO how a channel performed at the budget you already used. It doesn't tell them what happens if you scale that channel or move budget into it, and that's usually the decision on the table. Saturation curves answer that question by showing how revenue changes as spend goes up or down for each campaign.

A campaign that performs well at $50,000 a month may flatten out at $150,000, while another still has room to grow. Our Media Forecasting feature uses what the model has learned about each campaign to forecast what happens when you scale a channel or shift budget between channels.

Ask whether the curve is real or assumed

Many MMMs assume every channel eventually hits diminishing returns, and they build that shape in whether or not the data supports it. Prescient's research on saturation found that in campaign-level data, a simple straight line described how revenue grew with spend more accurately than standard diminishing-returns curves. When a model forces curves to flatten too early, it tells you to cap spend on channels that still had room to grow.

That's worth checking before your CFO approves a budget cut based on a curve. Ask these questions about any saturation curve you plan to show:

  • Has the brand ever actually spent at the level where the curve flattens?
  • Does the curve's shape come from your data or from a default setting?
  • How confident is the model at the spend level you're proposing?

Be upfront about how confident the model is

Marketers sometimes worry that showing uncertainty will make the model look weak. But finance teams work with ranges every day, so a model that's honest about what it doesn't know usually earns more trust. A few habits make uncertainty useful:

  • Put ranges in dollars. "We expect $250,000 to $400,000 in added revenue" lands better than statistical terms.
  • Use Prescient's confidence tools for what they are. The confidence band is the shaded range around a saturation curve forecast. The confidence score is a separate rating (like Strong or Moderate) for how much to trust a result.
  • Call out untested spend levels. Confidence drops when you propose spending at a level a campaign has rarely or never hit, because there's less history to learn from.
  • Round your numbers. 1.2 million reads as an estimate, while 1,213,477 reads like a promise.

Show why the model deserves your CFO's trust

Before a CFO uses MMM results to move real money, they'll want to know why they should believe them, and accuracy is only part of the answer.

A good fit isn't the same as correct credit

A model can predict your total revenue almost perfectly and still hand credit to the wrong channels, especially when spend on several channels rises and falls together. So show both sides: backtesting that compares the model's predictions with what actually happened, and an explanation of how the model separates one channel's impact from another's. If your CFO hears "the model is 95% accurate," they may assume that means 95% of the channel credit is right, which isn't what accuracy scores measure.

Where incrementality tests fit

Many CFOs have heard that incrementality tests are the gold standard. But a test measures one channel in one market over one window, so its results don't always hold for the rest of the business. Prescient can run your model with and without a test's results and show how accuracy changes. You make the final call, and a test you trust can be built into the model. Present it to your CFO as one input you checked, not the final word.

An example: why last year's holiday numbers might be inflated

Black Friday and Cyber Monday are often the biggest budget decisions finance signs off on, and they're when MMM results are easiest to get wrong. Brands raise ad spend right as shoppers are already ready to buy, so a model can mistake that seasonal surge for ad performance.

In Prescient's research on our model, researchers built a simulated marketing environment where the right budget for a BF/CM-style peak was known in advance. They then compared channel-level recommendations from our model and two widely used open-source MMMs:

ModelPeak-season budget recommendation vs. the known right answer
Open-source MMM AOverspent by about 81%
Open-source MMM BUnderspent by about 11.5%
Prescient's modelWithin about 1%

The takeaway for your CFO conversation: if last year's holiday ROAS looked unusually strong, ask how much of it came from demand that would have shown up anyway. Then plan this year's peak budget from the model's forecast instead of simply repeating last year's spend increase.

Bring scenarios, not a single ask

A CFO rarely approves a budget change just because a model recommended it. They want to see the options and the trade-offs, and three scenarios usually cover it.

ScenarioWhat it showsWhen it fits
Reallocate at the same spendRevenue you could add by shifting budget from weaker channels to stronger onesBudgets are flat, or the CFO wants efficiency first
Increase the budgetWhat added spend is forecast to return, with a rangeThe company is prioritizing growth and has cash to invest
Cut the budgetHow much revenue you'd likely lose, and where cuts would hurt leastCash is tight or margins are under pressure

The reallocation scenario is often the easiest win, since it doesn't ask for new money. Our Media Forecasting feature can build that plan for you, showing forecast revenue by channel or campaign at the same total spend.

Stage the rollout and define success

Before you ask for money, spell out how everyone will know it's working. A staged plan gives your CFO a way to say yes without betting the whole budget at once:

  • Start with part of the shift, not all of it, and confirm the spend actually goes out as planned.
  • Set check-in dates based on how long each channel takes to show its full impact, since some take weeks.
  • Agree on the minimum result that keeps the plan going and what you'll do if results fall short.
  • Compare actual results with the forecast and report back, even when the news isn't great.

That last point builds more credibility than any single presentation.

How to structure the deck

The most common mistake is building the presentation in the same order as your dashboard, with methodology first and the recommendation last. Flip it so you lead with the decision and use the MMM to back it up. This eight-slide structure works for most CFO meetings:

  1. The decision: your recommendation, the dollars involved, the expected return, and the biggest risk
  2. Goals and constraints: the growth target, margin, and budget you're working within
  3. How the numbers line up: revenue and spend definitions, and which storefronts the model covers
  4. What marketing contributed: MMM Revenue next to total revenue, by channel and storefront
  5. Profit by channel: Modeled ROAS against break-even ROAS, with the profit math
  6. Room to grow: saturation curves and confidence for the channels you want to scale
  7. Scenarios: reallocate, increase, and cut, each with a range
  8. The plan: staged rollout, check-in dates, and what success looks like

Put model details, backtesting results, and incrementality test comparisons in an appendix, where they're available without taking over the meeting.

Phrases to swap before you present

Small wording changes go a long way with finance, because they show you understand these are estimates tied to business decisions.

Instead ofTry
Meta drove $1 million in revenue.The model estimates Meta drove about $1 million in revenue, including halo effects.
This campaign has a 4x ROI.This campaign brings in about $4 in revenue per ad dollar, before product and shipping costs.
We need another $500,000 to scale.We forecast another $500,000 would add about $1.6 million in revenue and $220,000 in profit at our margin.
The platforms are over-reporting.Platforms report sales their ads touched. The MMM estimates the sales our ads actually added.

Where Prescient comes in

Prescient's marketing mix model updates daily and reports down to the campaign level, so you can see MMM Revenue split into base revenue and halo effects across your website, Amazon, and retail partners. Our Media Forecasting feature lets you build the scenarios your CFO wants to see, with confidence scores showing where a forecast stands on solid ground and where it's more of a stretch.

Walking into a CFO meeting with numbers that tie back to total revenue, a clear explanation of why they differ from platform reports, and scenarios with honest ranges changes the conversation from defending your budget to planning growth. Book a demo to see how Prescient can help you build that case.

FAQs

What ROAS should I show my CFO?

Show Modeled ROAS for each channel next to your break-even ROAS, which is 1 divided by your contribution margin. Platform-reported ROAS can live in the appendix, but it shouldn't drive budget decisions, because platforms often claim the same sales and miss revenue that shows up elsewhere. Adding the profit math gives your CFO a number they can compare with other ways to spend the company's money.

Why doesn't our MMM match what Meta reports?

Meta reports the sales its ads touched within its own attribution windows, and it can claim purchases Google also claims. A marketing mix model estimates the revenue each channel actually added across all your storefronts, including halo effects Meta can't see. Some channels will come out higher in the MMM and some lower. Meta's numbers still have uses, but the MMM answers the question your CFO cares about: what you'd gain or lose by changing spend.

How often should we report MMM results to finance?

Monthly or quarterly reporting works for most CFOs, ideally timed to budget planning. Prescient's model updates daily, which helps your team adjust campaigns quickly, but finance doesn't need daily swings. Use a consistent scorecard with the same revenue, spend, and margin definitions every time, and call out when the model has revised earlier estimates and why.

Can MMM forecasts go into the board plan?

They can inform it, as long as everyone knows which parts come from the model and which come from finance. A forecast of what happens when you reallocate budget or scale a channel is a strong input, while a full-year revenue plan also depends on pricing, inventory, and other factors finance owns. Share MMM forecasts as ranges, list the assumptions behind them, and agree with finance on how the two plans fit together before anything goes to the board.

What if the MMM says to cut a channel leadership likes?

Show the result anyway, since hiding an unflattering number undermines trust in everything else you present. Explain how confident the model is, check whether halo effects or retail revenue could change the picture, and propose reducing spend partway before deciding. That shows your CFO you're using the model to make decisions, not to defend a budget you'd already settled on.