Conversion value QA: one misplaced decimal can distort the whole campaign
Validate amount, currency, quantity, discounts, and event identity with known orders. A purchase count can be correct while reported revenue is badly wrong.
Read the guide →Campaigns. Delivery.
Measurement.
The article library
Metric definitions, cost calculations, and reporting differences.
15 articles
Validate amount, currency, quantity, discounts, and event identity with known orders. A purchase count can be correct while reported revenue is badly wrong.
Read the guide →A separate booking or checkout domain can split a customer journey. Test the transition, identity continuity, and purchase event rather than only the final thank-you page.
Read the guide →Recent advertising cohorts have had less time to convert. Compare outcomes at the same age before cutting spend based on an apparently weak day.
Read the guide →A seven-day click window and a thirty-day window answer different reporting questions. Align the rules before using attributed conversions to compare channels.
Read the guide →Separate a traffic-quality alert from its possible causes. Segment the affected delivery, preserve evidence, and review acquisition sources before drawing conclusions.
Read the guide →Viewability is usually calculated from measurable impressions. Read measurement coverage alongside the rate so an improvement does not hide a shrinking sample.
Read the guide →A successful upload does not guarantee a useful optimization signal. Define the lead milestone, timestamp, identifiers, and correction process before connecting the CRM.
Read the guide →First-user, session, and event-scoped acquisition data answer different questions. Learn which comparison is useful before treating a mismatch as a tracking failure.
Read the guide →Average frequency can hide a small audience seeing most of the ads. Check the distribution, cap scope, and reporting window before changing delivery.
Read the guide →