What does a profit-aware proposal contain?
The products or collections to promote, chosen by margin after returns, discounts and dated landed cost rather than by revenue; the budget; the expected spend against the monthly cap; and the arithmetic behind the choice. A line that sells well and returns heavily does not become an ad proposal just because it is popular.
Because the commerce data is in the same database as the proposal, the proposal can say what a product actually earns per unit today and what a return costs, which is exactly the input a media buyer normally lacks.
- The products or collections to promote, chosen by margin after returns and dated landed cost.
- The budget, and what it would leave against the monthly cap.
- The arithmetic behind the choice, including what each unit actually earns today.
- The audience and creative the proposal would use, for review before anything runs.
- The measurement plan: which figures will be read back, over which window.
Why money safety comes first here
Advertising is the only surface where a mistake spends real money immediately. So the rule is stricter than elsewhere: no real spend can move without an explicit human approval, even when unattended sending is enabled for other classes of action. A proposal is a document until a person signs it.
The monthly cap is an absolute bound rather than a target. The kill switch stops everything at once and is designed to be used, not to be a line in a marketing page.
What is in the audit ledger?
Every proposal, approval, change and stop is written to a ledger with its time, its author and its figures. If a campaign spent more than expected, the record of what was approved is there, and if a campaign was stopped, so is the moment it was.
This is the same discipline the rest of the engine applies to sends, applied to money that leaves the business faster. An ad platform's own history is not enough, because it does not record who inside the brand agreed to what.
How are results read against the products?
Results come back from the ad platform and are read against what the store actually sold and kept. Spend against a collection is compared to that collection's units, net sales and margin in the same window, and the comparison states its assumption rather than claiming attribution the data cannot support.
Where the brand's own platform-reported figures and the store's own figures disagree, both are shown. An honest gap between platform-reported and store-measured is more useful than a single confident number.
What the control room does not do
It does not move budget on its own, bid, or launch a campaign without approval. It does not promise a return; it proposes, measures and reports. And it does not hide a bad month: a campaign that lost money is reported as such in Results.
The roadmap's outcome measurement, where a recommendation declares what it expects before it runs and the system publishes its own accuracy afterwards, is labelled as roadmap here as everywhere else.
Which role owns it on ftes.ai?
The Acquisition Officer FTE owns paid acquisition, works under a hard monthly cap and an approval gate for every spend, and is judged on contribution after ad cost rather than on revenue. Its proposals appear in Approvals with the same arithmetic the control room shows.
A brand starting here should set the cap first, approve one small campaign, and read the Results verdict before approving a second. That order is what keeps advertising a decision rather than a habit.
What does a careful first campaign look like?
A small budget, one collection chosen by margin rather than by popularity, a cap set before anything runs, and a measurement window agreed in advance. The proposal states what will be read back and when, so the review at the end is a comparison rather than an argument.
Then read Results honestly. If the campaign did not clear its own cost after returns, that is the finding, and the next proposal is built on it. The value of the control room is that this loop exists at all.