Spot underperforming ad spend in one afternoon
Three passes over data you already hold, each one narrowing the list, ending with a single channel worth testing rather than a ranked table nobody acts on.
Three platforms, three answers, one budget to defend. Your GA4 export already holds the answer, and the first finding is free and never leaves your browser. Get your first finding free
By Joris van Huët, Founder & CEOUpdated 3 min read
Run the numbers for your store: the free break-even ROAS calculator.
Three passes, each narrowing the list, ending with one channel worth testing rather than a table nobody acts on.
The afternoon
| Pass | What it removes from the list | Time |
|---|---|---|
| 1 | Channels whose spend is too small to matter either way | 15 minutes |
| 2 | Channels where reported and causal figures broadly agree | 45 minutes |
| 3 | Channels whose interval is too wide to conclude anything | 15 minutes |
A ranked table of every channel is what most audits produce and it is why most audits change nothing.
Pass one: drop the immaterial
Sort channels by spend. Anything below a threshold you set, and below the level at which any method can separate its effect from ordinary variation, comes off the list. Not because it is fine, but because no analysis will resolve it and the decision there is a judgement call rather than a measurement one. The arithmetic is in the measurability floor.
Pass two: compare reported against causal
Export the acquisition window from Google Analytics and run a causal read. Put the causal estimate beside the platform figure for each remaining channel.
Where the two broadly agree, the channel is probably doing acquisition rather than harvesting, and it comes off the list. Where they diverge sharply, it stays. The pattern is explained in why platform and causal ROAS disagree.
The read is €99, once, refundable within 30 days, no questions asked, and returns per channel what it caused next to what last-click gave it, a data-health score rather than a confidence interval, and a next step.
Pass three: drop the unresolved
Any remaining channel whose interval spans break-even stays on a watch list rather than a cut list. An unresolved estimate is a request for more data, not evidence of weakness, and cutting on it is the mistake described in the cost of cutting the wrong channel.
The output
One or two channels where the gap is large, the interval is narrow, and the spend is material. Those are the candidates, and the next step is a geo holdout on whichever carries the most budget rather than an immediate cut.
What to write down
The shortlist, the numbers behind it, the date, and the threshold you would use to decide. That record is what makes the eventual decision defensible and the next audit comparable.
Related answers
Frequently asked questions
How do I audit ad channels in an afternoon?
Three passes: drop channels too small for any method to resolve, drop those where reported and causal figures broadly agree, and drop those whose interval spans break-even. What remains is a shortlist of one or two.Why not produce a ranked table of every channel?
Because ranked tables of everything are what most audits produce and why most change nothing. A shortlist of one or two channels with a specific next step is actionable; a complete ranking invites discussion instead.What should happen to an unresolved channel?
It goes on a watch list, not a cut list. An interval spanning break-even is a request for more data rather than evidence of weakness, and cutting on it is the most common way a working channel gets removed.
Go deeper: Causal attribution, explained.
Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.
Have an audience of ecommerce operators? Earn 20% recurring commission by recommending the €99 read.
Keep reading
Terms in this article
- Ad SpendAd Spend is the total amount invested in advertising campaigns. It is measured against Return on Ad Spend (ROAS) to evaluate campaign effectiveness.
- AnalyticsAnalytics is the systematic computational analysis of data. It reveals customer behavior and measures campaign performance.
- AttributionAttribution identifies user actions that contribute to a desired outcome and assigns value to each. It reveals which marketing touchpoints drive conversions.
- Attribution ReportAttribution Report shows which touchpoints or channels receive credit for a conversion. It identifies which campaigns drive desired actions.
- Confidence IntervalConfidence Interval is a statistical range of values that likely contains the true value of a metric. In marketing analytics, it quantifies uncertainty around estimates, indicating the precision of an outcome or causal effect.
- Google AnalyticsGoogle Analytics is a web analytics service that tracks and reports website traffic.
- Holdout TestA holdout test is an experiment that keeps a random part of an audience from seeing a campaign and compares its results with those of the part that could see it. The held-out part is the control group, so the gap in outcomes estimates what the campaign added.
- IncrementalityIncrementality is the part of an outcome, such as orders or revenue, that a marketing activity caused: the extra results compared with what would have happened without it. It is a causal effect, not the number of conversions credited to a channel.