Glossary · Term
The translation layer between buyer rejection codes and internal reason codes used for reporting and source-level action.
Definition
Every buyer emits its own rejection taxonomy — some as codes, some as free text, some as HTTP status. Rejection mapping normalizes these into a small internal set of reasons (duplicate, filter, cap, quality, unknown) so that source-level reporting is comparable across buyers and so that publisher-side action can be taken automatically.
Why it matters
Without a rejection map, a spike in rejections looks the same as a spike in cap-outs and cannot be diagnosed without manual work. With a rejection map, source-level rejection rates drive automated publisher throttling, quality alerts, and payout renegotiation triggers.
How it fails in practice
Buyers add new rejection codes without notice, and they fall into the `unknown` bucket. Rejection reports look clean while a specific failure mode compounds. The map has to be reviewed on a schedule; nobody does this reliably without a documented process and an accountable owner.
Related terms
A 30-minute working session to map how this term shows up in your platforms, workflows, and reporting — and where it is currently exposed.