The Verdict-First Framework for Answering "How Are the Ads Doing?"

"How are the ads doing?" is the single most common question in performance marketing, and it is almost never answered directly. Instead it gets answered with a report: a wall of charts, a table of click-through rates, a row of percentages next to last month's numbers. The founder or the client is left to do the one thing they hired someone else to do for them, which is figure out what it all means.

This is backwards. When someone asks how the ads are doing, they are asking for a verdict: good, bad, or fine-but-watch-this. Data is supposed to support that verdict, not replace it. A report that opens with a spend chart and ends, three scrolls later, with a sentence buried in paragraph six that says performance was "broadly stable" has answered the question last, and weakly, when it should have answered it first, and clearly.

At Riibon, every reporting surface is built around one rule: the first sentence a person reads has to be a real verdict, not a data point and not a hedge. "Ads are healthy, no action needed this month." "Ads had a rough month, and it's one thing: cost per lead on the retargeting campaign doubled." "Ads are fine, but we're watching rising CPMs on Meta, if it continues into next month it'll eat the margin." Each of those is a complete answer on its own. Someone could stop reading after that sentence and know exactly where they stand.

The test for whether something counts as a real verdict is whether it's falsifiable, meaning someone could later check whether it was true. "Performance was mixed" is not falsifiable. It cannot be wrong, because it doesn't actually claim anything. "Ads are healthy, no action needed" can be checked against the account a month later. That difference, between a claim that can be checked and a phrase that can't, is the whole framework in one sentence.

The verdict-first structure sounds simple and is rarely done, for a specific reason: writing a real verdict up front requires the analyst to have already finished the analysis before they start writing the report. They have to know the answer before they type the first word. Most reports get built the other way around: the analyst opens a template, pulls in the metrics, and uses the act of writing the report as the place where they figure out what happened. The report becomes a working document, not a finished answer, and the reader gets handed the unfinished version.

There's also a psychological reason vague hedges are so common. "Performance was mixed this month, with some campaigns up and others down" protects the person writing it. It can't be proven wrong, because it doesn't commit to anything checkable. A real verdict, by contrast, is a bet. If an analyst writes "ads are healthy" and next month tells a different story, that's visible and has to be reconciled. Hedging is what happens when the report is optimized for the analyst's comfort instead of the reader's decision. A verdict-first report only works if the person writing it is willing to be pinned down, and that willingness is the actual differentiator, not any particular formatting choice.

The verdict is not the whole report, it's the headline the rest of the report has to earn. Everything after it should exist to justify that specific claim, not to give a generic tour of every metric available. If the verdict is "ads had a rough month, driven by retargeting CPA doubling," the supporting section should walk through what changed in that campaign specifically: what the CPA was, what it became, and what's known about why. It should not also include an unrelated paragraph about a channel that was flat, just because that channel has a metric that could be reported. Every additional fact included after the verdict should be there because it's needed to trust the verdict, not because it was available.

This also changes how much detail belongs in a given report. A founder who has never run a campaign needs less supporting detail to trust "ads are healthy" than an agency manager auditing whether their vendor's month-end report holds up. The right amount of evidence is however much is needed to make the verdict checkable and credible to that specific reader, not a fixed template applied regardless of who's reading it or what's actually at stake that month.

If you manage an ad account through an agency and never touch Meta or Google Ads directly, this framework gives you something concrete to check instead of a vague feeling that reports could be "clearer." Open the last three monthly reports your agency sent. For each one, look at the first paragraph and ask: does it state, in one sentence, whether the ads did well or poorly, and why? Or does it open with spend totals, a chart, or a phrase like "solid month overall" that doesn't actually commit to anything?

If the verdict is missing, delayed, or hedged, that's a specific and reasonable thing to raise with your agency, not a soft complaint about communication style. You can say exactly this: "I want every report to open with one sentence that tells me if the ads are healthy or not, and why, before any charts." That's a request an agency can act on directly, unlike "please communicate more clearly," which nobody knows how to fix. And if the agency can't produce that sentence when asked, that tells you something too: it usually means the analysis wasn't finished before the report was written, which is worth knowing regardless of how the report itself reads.