Most fractional CFOs hand the marketing team a budget and report on it afterward. I take named access to the ad platforms and measure what the spend earned the business after the cost to deliver the product. That figure is ad margin, I track it in dollars and as a percentage, and it is the number the budget gets set against.
Platform ROAS is the revenue a platform attributes to itself, divided by the spend on that platform.
It is a gross number, and for most brands it is the only one they ever see.
Ad margin asks the same question financially. It takes the revenue the advertising produced, subtracts COGS, freight and everything else it costs to deliver the product, then subtracts the advertising itself.
What is left is what the spend earned the business.
The two can point in opposite directions in the same month. A 4.0 platform ROAS says the campaign worked. Ad margin on those same orders can be thin or negative once product cost, freight, returns and discounts are loaded in.
This is most of what your agency and your accountant are actually disagreeing about. The agency reports the first number. The accountant reports the consequence of the second. Neither on its own is enough to set a budget.
| Per order | $ |
| Average order value | 60.00 |
| Ad spend at 4.0 platform ROAS | (15.00) |
| COGS at 38% | (22.80) |
| Freight, fulfillment, packaging | (11.00) |
| Payment processing at 3% | (1.80) |
| Returns and discount allowance at 9% | (5.40) |
| Ad margin per order | 4.00 |
| Ad margin as a percent of revenue | 6.7% |
| Fixed and operating costs at 12% of revenue | (7.20) |
| Left for EBIT | (3.20) |
Illustrative figures, not a client's. The platform reports a 4.0 and the business loses $3.20 an order. Note also that the $15 assumes the platform's attribution is correct. Where two platforms both claim the same order, real spend per order is higher and ad margin is thinner than this.
Expressing the overhead floor either way is the same arithmetic. You can say the ad margin dollars have to cover $7.20, or that the ad margin percentage has to clear 12%.
I use whichever one the business already thinks in.
This is the most common thing I get called about, and it is usually one of four causes. The test tells you which.
The spend landed in one period and the revenue landed in the next. Nothing is wrong with the advertising.
COGS or OpEx moved for reasons that have nothing to do with advertising, and absorbed the gain.
The same order count brought in less money, often because discounting did the work the creative was credited for.
Spend that buys repeat buyers is likely buying revenue you would have had anyway. An acceptable ROAS next to flat growth.
There is a fifth case worth naming separately, because the response is different.
Sometimes the advertising is fine and the shortfall was never in the advertising. Compare the size of the gap to what a realistic CPA improvement would produce. If the gap is bigger, the cause is volume, pricing or fixed cost, and more spend will not fix any of the three.
A brand can optimize its advertising for a year and stay unprofitable, not because the optimization failed, but because it was aimed at the wrong thing.
The full six-step method, in order, with what each step requires before it will compute, is written out at setting an ad budget from the P&L.
| Ad spend as a share of net sales | 47% to 30% |
| EBIT margin | -0.7% to 16.6% |
| FY2025 ad spend | -24% |
| FY2025 sales, against that spend cut | +10.4% |
| Google spend, year over year | -64.5% |
| Active channels | 3 to 6 |
| Markets | 1 to 4 |
One brand, figures from the client's own monthly P&Ls. Absolute dollars are redacted at the client's request and all ratios are real. The full engagement is at /case-study.
Spend came down and sales went up, which is what the method is for.
Cutting the total and reallocating it are two separate decisions. Deciding them together is how a brand ends up cutting a channel that was working because the overall number needed to come down.
Because platform ROAS excludes every cost of delivering the product.
Load in COGS, freight, fulfillment, payment processing, returns and discounts and the same order can leave very little behind. The worked example above shows a 4.0 ROAS producing an ad margin of 6.7% against overhead running at 12%.
Nothing in the agency's number is wrong. It is just not a profitability number.
Yes, named access to every ad platform, plus the storefront, the accounting system and the bank.
A monthly export from the agency is not enough. The figures that matter are at campaign and channel level and have to line up against the same periods in the P&L.
This is the part most fractional CFO engagements skip.
No, and not because they are unwilling.
The agency does not hold your COGS, freight, returns, discounts or overhead, so it cannot compute what the spend earned. It also has a structural interest in the spend level.
The agency should run the campaigns. Someone whose data includes the cost of delivering the product should set the budget.
Then the first job is the books, not the ads.
The monthly P&L has to be normalized so each month carries its own rent, interest, depreciation and inventory, and the data has to tie out before anyone trusts a conclusion drawn from it. Twenty-four months is the right depth, because it covers two of every season.
With less history the spend response is still readable, just with wider error bars, and I say so rather than presenting a number I do not have.
Yes.
A brand taking share or entering a market may choose to run at a thinner ad margin on purpose, and that is scaling wisely as long as the business can fund the gap while it does. A brand protecting cash may choose the other side.
What makes it a decision rather than an accident is knowing what the next point of growth costs in margin before you buy it.
Send your last twelve months of P&L and whatever SKU or channel detail you keep. I work through it and bring you where I think the margin actually is, before the first call.
Mutual NDA before you send anything, no charge, entirely optional. Plenty of good calls happen without it.
Start a conversation