Glossary
Deduplication
Recognising that two records describe the same event, so one purchase reported twice is counted once. The routine failure mode of sending events from the browser and the server.
Also called: dedupe, event deduplication
Send the same purchase from the browser and from your server — a common and sensible setup, because each path catches what the other misses — and the platform receives two events describing one sale. Without deduplication it counts both, and every number downstream is inflated.
The fix is an identifier both copies carry. Each sender stamps the same event ID, and the receiver treats two events with matching IDs inside a window as one event. That is the whole mechanism, and almost every failure is a variation on the ID not matching.
Where it goes wrong.
- The two senders generate their own IDs. They must share one, derived from the order, not created independently at each end.
- The IDs match but the event names do not. Most platforms match on both. A
Purchaseand apurchaseare two events. - The second copy arrives after the window closes. Server-side events that queue or retry can land outside it and get counted fresh.
- A retry sends a new ID. Now the retry is a second sale.
The one that cannot be fixed by configuration: deduplication only works inside a single receiving system. Two different vendors each receiving the same conversion will both count it, and no event ID makes them agree — they never see each other. Totals that exceed real revenue across tools are usually this, not a tracking bug.
Do not confuse with
Close enough to get mixed up, different enough that the mix-up costs something.
- Server-side tagging Sending marketing and analytics events from your own server instead of from the visitor's browser, so a request the browser might block is made somewhere it cannot be.
- Modeled conversions Conversions an ad platform estimates statistically rather than observes directly, used to fill gaps where consent, tracking prevention or privacy limits block measurement.