Ad platforms report the revenue they can attribute to themselves, which is a different number. These are the questions that close the gap, in the order a finance team would ask them.
Is it judged on profit or on revenue?
ROAS, return on ad spend, divides revenue by ad spend, so it treats a low-margin sale the same as a high-margin one. The number finance needs is profit on ad spend: gross profit from the sales, divided by what the ads cost. If the platforms are told the revenue of each sale, they will chase revenue, including revenue that loses money.
Do the platform numbers match the books?
Each platform counts the sales it can connect to its own ads, and two platforms can both claim the same sale. Add their reports together and the total can exceed what actually came in. Reconcile every month: platform-reported conversions against the enquiries or orders in your own system for the same period, and total revenue against total marketing spend, the marketing efficiency ratio, which no platform can inflate.
Would the sales have happened anyway?
Attribution decides which ad gets credit for a sale. It cannot say whether the sale needed the ad. Searches for your own brand name and ads shown to people who already visited your site are the usual places to look first, because they reach people already on their way to buy. The reliable way to know is a controlled test, set out in the incrementality guide.
What does the advertising cost in total?
Media spend is only part of it. Add management fees, tools, tracking, creative production and internal time, then judge the return against the whole. A fee charged as a percentage of ad spend deserves a second look: it can reward spending more rather than spending well.
Who owns the accounts and the data?
The ad accounts, the analytics property and the tag manager container should be owned by your business, with your agency or consultant given access. If the relationship ends, years of performance history and audience data stay with you rather than leaving with the supplier.
Is the tracking sound?
Every platform optimises toward what it is told is a conversion, so a tracking fault turns into wasted spend without any visible failure. A sudden fall in conversions while spend and clicks hold steady is a tracking break to investigate, not a collapse in performance. The conversion tracking checklist sets out what to check.
Is the budget set from the margin, or from last year?
A budget copied forward with a percentage added says nothing about whether the last pound earned its keep. Work out the most you can pay for a sale or an enquiry from your margin, then judge each increase on what the extra spend brought in. The average return across a whole budget hides the fact that the last pound usually returns less than the first.
Is the comparison like for like?
Compare a month with the same month a year earlier, or a period with its own matched window. A December judged against a November, or a week of a sale against a normal week, produces a change that is real arithmetic and meaningless for a decision.