How to justify marketing spend to your CFO: the numbers finance actually asks for

Because ROAS answers a marketing question. It tells you the revenue the platform credits to ads per euro spent. Finance asks a different one: how much money was left, and when did it arrive?

Three gaps explain most of the distrust:

• Revenue is not margin. ROAS counts revenue before cost of goods, shipping, payment fees, returns and discounts.
• Credit is not cause. The platform counts sales it was given credit for. Some would have happened without the ad.
• Reported is not reconciled. The panel and the finance system rarely agree to the euro. Often nobody has checked why.

Surveys on the topic exist. An adtech vendor surveyed senior marketers in the US and Canada and found that only a minority fully agree their CFO is aligned with them¹. A media write-up of the same survey led on how few feel sure they can justify value to the CFO². A Bain & Company survey of marketing and finance leaders, previewed at Cannes Lions, found the two sides share more goals than each assumes, while finance still distrusts how marketing is measured³. Neither survey was built around a mid-sized European ecommerce brand. Ask your CFO directly what they would need to see.

Pick a short list. Agree the definitions once, in writing, with finance. In the Bain survey, CFOs said they want the measurement framework agreed before campaigns launch³.

Two notes on our own convention. First, in our view "contribution margin" means revenue minus variable costs before marketing: cost of goods, shipping, payment fees, returns and discounts. What is left after marketing we call "contribution after marketing". Say which one you mean, and whether it is per order or a percentage of revenue. Second, in our convention MER divides revenue by all marketing cost, while "blended ROAS" divides it by media spend only. Some teams use the two names for the same ratio. Pick one name and write the definition next to it.

A vendor that sells financing to ecommerce brands makes a similar point: MER and ROAS should not be optimised in isolation, but read through to the P&L⁴. Keep the reasoning, and discount the sales motive.

For the cash side, cash conversion cycle shows how long money is tied up in stock and receivables. A campaign that scales demand also scales stock. The warehouse and the bank both have opinions.

Split every number into one of three bins.

1. Measured. Taken from orders, ledgers or a proper test. You can show the source.
2. Assumed. Needed to get from the data to the conclusion. For example, the share of platform-credited sales that are incremental, or the repeat rate you expect.
3. At risk. What would change the answer, and how bad it gets if you are wrong.

Put the three on the same page. Finance is rarely upset by an assumption. It is upset by one it finds on its own.

Example (illustrative): You spent 30,000 on acquisition and won 600 new customers. That is a CAC of 50. Each customer brings 20 of contribution margin per month, so payback is 2.5 months. The 20 is measured from past orders. The repeat purchases behind it are an assumption. The risk is that new customers behave differently from the ones you have today.

Show a range, not a point. A single precise number invites a precise challenge. A range with a stated reason invites a decision.

A test usually returns an interval, not a verdict. Say what it means for the decision: does the whole range sit above break-even, below it, or across it?

Example (illustrative): A test puts incremental ROAS between 1.2 and 2.4. Your break-even ROAS is 1.8. The range straddles the line. The honest message is "we cannot yet tell". You propose a cautious next step with a stop rule, not a confident yes.

"Cannot yet tell" is a result. It is cheaper than a confident wrong answer. If a test is out of reach, say what you used instead and what its bias is.

One page. Same layout every month. Changes explained in a sentence, not a chart. The layout below is our view, not a standard.

A BCG paper on CMOs and CFOs working together recommends a shared set of KPIs and a marketing investment committee with finance in the room⁵. In our view, for a mid-sized brand that can be one monthly meeting with the same page each time. If you want to read your agency's report the same way, see how to read an agency report.

Skip these. Each one comes back as a question in month four.

• A ROAS target as a promise. It is a platform figure, not a cash flow.
• A guaranteed payback. You can forecast one. You cannot guarantee it.
• Precision you do not have. Two decimal places on an estimate help nobody.
• Proof from a platform's own report. Treat it as one input, and say who produced it.
• Silence about assumptions. The first one finance discovers costs you the rest.

For ZFD: In our view, a report to finance belongs on contribution margin, payback and cash, with assumptions next to results. Anything weaker is a story about the ads.