An integrity check that catches a mislabel before it becomes a wrong decision
An e-commerce software vendor asked for the key Google Ads and GA4 metrics for the last full month — with a twist: don't just report them, sanity-check them. AOS pinned every figure to its source, tagged each as [DATA], and ran the checks that catch the two classic reporting failures: a revenue-versus-spend mislabel, and a Databox datasource quietly connected twice. The run cost $0.07.
01 Ground
For a software vendor running performance marketing, the monthly metrics pull is a load-bearing input — it informs budget, channel mix, and the story told upward. But the most dangerous reporting errors aren't missing numbers; they're plausible wrong ones. A revenue column labelled as spend (or vice versa) reads perfectly until someone acts on it. A Databox datasource connected twice doubles a metric that then looks like real growth.
These errors are insidious because they survive review. A human skimming a dashboard sees a number in the right shape and the right ballpark and moves on. The mislabel and the double-count both produce figures that pass the eyeball test, which is exactly why they slip through to a decision.
The vendor didn't want a prettier report. They wanted a report that had been actively distrusted — one where the numbers had been challenged against their sources before being presented, so that an integrity defect would be caught at the data layer rather than discovered three decisions later.
- Last full month's Google Ads + GA4 key metrics, each [DATA]-tagged to its source Google Ads / GA4 2026-06-21 90% confident
- Integrity verdict: no revenue-vs-spend mislabel and no double-connected Databox datasource detected Databox / cross-source reconciliation 2026-06-21 88% confident
02 Decide
AOS treated the request as an audit, not a fetch. Every metric was pinned to its system of record and marked [DATA] so its provenance is explicit, and then the two named failure modes were checked directly: was anything labelled revenue actually spend, and was any single source feeding the aggregator twice and inflating a total?
The editorial discipline here is restraint. AOS did not dress thin numbers up as insight. It validated the month, declared the figures sound against their sources, and made the integrity check itself the headline — because a metrics report whose first job is to catch its own errors is worth more than one that confidently presents a number that happens to be wrong.
03 Deploy
AOS queried Google Ads and GA4 for the last full month, reconciled the overlapping figures, and ran the mislabel and duplicate-datasource checks against the Databox configuration — emitting each metric with a [DATA] tag tying it to source.
At $0.07 the check is effectively free, which is the point: integrity verification this cheap can sit in front of every monthly pull rather than being reserved for when someone already suspects something is wrong.