There is a particular kind of marketing report that arrives on the fifth of every month. Forty slides. Impressions up. Reach up. Engagement rate up. A word cloud, for some reason. And nowhere in it, an answer to the only question the person paying actually has: did this make us money?
That report is not incompetence. It is a choice. Impressions always go up when you spend more. A report built on impressions can never deliver bad news — which is precisely why it gets built.
The six numbers
A monthly report should fit on one page and answer six things. If your agency cannot produce these, the problem is not the report. It is the measurement underneath it.
| Number | What it answers |
|---|---|
| Spend, by channel | Where the money went — media separated from fees, never blended into one line. |
| Qualified leads or orders | Volume of outcomes that a salesperson or a warehouse would recognise as real. |
| Cost per qualified outcome | Spend divided by the line above. The single most useful number in the report. |
| Revenue attributed, and how | The figure and the method. "Last click, 30-day window" is an answer. "Platform reported" is not. |
| Contribution after ad spend | Revenue minus cost of goods minus media. The number your accountant would care about. |
| What we cut, and what we are testing next | Proof that someone made a decision this month. |
Notice what is missing: impressions, reach, engagement rate, follower growth, "brand sentiment". Those are diagnostics. They belong in an appendix, useful when a number moves and you need to know why. They are not the report.
Why ROAS on its own is a trap
Return on ad spend is the most quoted number in digital marketing and one of the most misleading, for a simple reason: it ignores your margin.
Take a 4.0x ROAS. It sounds excellent. Now apply it to a business with a 25% gross margin. For every $1,000 of media, you generate $4,000 of revenue and $1,000 of gross profit. You have exactly broken even on media — before you have paid your agency, your staff, or your rent.
The same 4.0x on a 70% margin business produces $2,800 of gross profit on $1,000 of media. One of these companies should be spending everything it can find. The other is quietly losing money while its dashboard glows green.
Ask what your break-even ROAS is. If nobody in the room knows, nobody in the room knows whether the campaign is working.
The formula is not complicated: break-even ROAS = 1 ÷ gross margin. At 25% margin, you need 4.0x just to stand still. At 60%, you need 1.67x. Every target above that line is profit; everything below it is subsidy.
The platform numbers will not add up. That is normal.
Meta will claim credit for a sale. Google will claim credit for the same sale. Add the two dashboards together and you will find you sold 140% of what your accounting system recorded. This is not fraud, and it is not a bug — each platform measures its own contribution, in its own attribution window, with its own view-through rules.
There are three honest responses:
- Pick one source of truth — usually your own analytics or your order system — and report platform numbers as secondary.
- Ask buyers directly. A single "how did you hear about us?" field on the order confirmation, answered by even 30% of customers, is often more directionally honest than any attribution model.
- Test. Turn a channel off in a region for four weeks and watch total revenue, not platform revenue. Expensive, uncomfortable, and the only method that measures what advertising actually caused.
The two questions
You do not need to audit a tracking setup to find out whether a report is honest. Ask these in the next review:
1. "What did you turn off this month?"
Every account contains waste. Search terms that never convert, an audience that eats budget, a creative that stopped working three weeks ago. An agency actively managing an account can always name something it killed. "Everything is performing well" is not a reassuring answer — it means nobody looked.
2. "What would make you tell us to stop spending?"
This one is uncomfortable by design. An agency paid a percentage of your media spend has no answer to it, because the honest answer costs them money. An agency on a fixed fee can tell you the number at which the channel stops being worth it — and will tell you when you reach it.
What a good review actually sounds like
Twenty minutes. Six numbers. Three decisions: what we are scaling, what we are cutting, what we are testing next month. A named person who owns each decision. Then the appendix, for anyone who wants to see why.
If the meeting is longer than that and you still cannot repeat the cost per customer from memory afterwards, the report is not working for you. It is working for whoever wrote it.