Publisher revenue KPI dictionary
A concise publisher-side KPI dictionary for reading offerwall performance, postback health, and payout quality without mixing up traffic and revenue signals.
Publisher revenue KPI dictionary in brief
This KPI dictionary is the fastest way to align publisher, ops, and advertiser conversations before someone changes placements, caps, or payout assumptions.
Use it when your team needs a shared meaning for revenue metrics like EPC, acceptance rate, fill rate, rejection rate, and latency. If the definitions are drifting, every optimization decision becomes harder to trust.
Who this is for
- Publishers reviewing placement and geo performance.
- Ops teams reconciling postbacks and payout visibility.
- Advertisers who need to interpret publisher-side quality signals before scaling.
Definition
A publisher revenue KPI is a metric that helps explain one of four things:
- how much value a traffic segment creates,
- how cleanly conversions are tracked,
- how reliably payouts are confirmed,
- and when a source should be scaled, held, or paused.
Use the same definitions across dashboards, weekly reviews, and support tickets.
Decision table
| KPI | What it answers | Good for | Watch-out |
|---|---|---|---|
| EPC | Are clicks turning into revenue efficiently? | placement and source ranking | Can mislead if postbacks are delayed |
| eCPM | Is page or session inventory monetizing efficiently? | comparing monetization surfaces | Not a substitute for payout quality |
| Postback acceptance rate | Are conversions being accepted cleanly? | tracking health and QA | High acceptance alone does not prove good traffic |
| Rejection rate | How much traffic or conversion volume is being filtered out? | fraud and policy review | Needs reason-code context |
| Fill rate | Are users seeing enough monetizable offers? | offer availability review | More fill is not always higher-quality fill |
| Conversion latency | How long does it take for conversions to confirm? | reconciliation and pacing | Delay can look like revenue loss |
| Approval rate | Of accepted conversions, how many remain approved? | advertiser-side quality alignment | Must be matched to the same time window |
How it works
EPC
Definition: revenue per 1,000 clicks.
Use EPC to compare placements, geos, or sources only when the tracking window is consistent. Rising EPC is helpful only if rejection and latency stay stable.
eCPM
Definition: revenue per 1,000 impressions or sessions, depending on the reporting model.
Use eCPM when you want to compare monetization surfaces, not just click efficiency. A placement can have better EPC and worse eCPM if too few users engage with it.
Postback acceptance rate
Definition: accepted conversion events divided by all received conversion events.
This is one of the fastest ways to spot tracking damage. If acceptance falls after a launch change, debug IDs and parameter mapping before touching offer strategy.
Rejection rate
Definition: rejected conversion events divided by all received conversion events.
Track rejection with reason codes. A rise in invalid_payload means a different problem than a rise in quality_reject or late_window.
Fill rate
Definition: the percentage of eligible traffic that receives a monetizable offer or wall experience.
Low fill can cap revenue. High fill can still be poor if the extra inventory lowers approval quality.
Conversion latency
Definition: the time between the click or qualifying action and the confirmed postback.
Use latency when finance, ops, or campaign teams are seeing “missing” revenue that later appears. It is a pacing and trust metric, not just an engineering metric.
Approval rate
Definition: approved conversions divided by accepted conversions, usually after advertiser validation.
Approval rate helps explain whether accepted traffic is still commercially valid after deeper review.
Example
A publisher sees EPC rise but payout visibility stays flat. The right next move is not “scale the winning placement.” Check postback acceptance and conversion latency first. If latency spiked or acceptance fell, the EPC move may be temporary or misleading.
A second publisher sees steady EPC and higher approval rate after tightening geo filters. That is usually a stronger scaling signal because both monetization and quality improved together.
Common mistakes
-
Comparing metrics from different windows Fix: keep 7-day and 30-day views explicit.
-
Using EPC without tracking-health context Fix: review EPC beside acceptance and rejection.
-
Treating rejection as one number Fix: split by reason code and source.
-
Using eCPM and EPC interchangeably Fix: use eCPM for surface efficiency and EPC for click efficiency.
Checklist
- KPI definitions are documented in one shared source.
- EPC and eCPM are not mixed in the same decision without explanation.
- Acceptance and rejection are reviewed with reason codes.
- Latency is visible before payout disputes escalate.
- Approval rate uses the same source and time window as conversion reporting.
- Weekly reviews use the same KPI names as support and ops logs.
FAQ
Which KPI should a publisher watch first?
Start with EPC, acceptance rate, and rejection reasons together.
Can one good KPI outweigh one bad KPI?
Usually no. Rising EPC with broken postbacks is not a clean win.
Is approval rate a publisher KPI or an advertiser KPI?
It is both. Advertisers own the deeper quality rule, but publishers need the number to understand sustainable revenue.
When should latency block scaling?
When delays are large enough that teams cannot trust current revenue or payout visibility.
Conversion link
Use this dictionary before changing a live revenue setup. Then move to the parent hub at Publisher revenue systems, the diagnostic path in Publisher EPC optimization checklist, and the setup path in Affiliate offerwall. If the KPI picture is clear and you are ready to act, continue to Publishers.
Evidence notes
- Strategy source:
docs/content-silo-plan.mdand the synced Obsidian strategy note dated2026-06-18. - This page is intentionally concise because it is a glossary/dictionary artifact under the locked launch rule.