To judge whether a campaign is hitting its target or underperforming, the engine needs a baseline. It checks four levels, in order of specificity, and stops at the first level with a usable signal:
The campaign's own Target CPA - most specific; if the campaign has a tCPA, that's the judge.
The campaign's own Target ROAS - the value-based alternative.
The account-wide Target CPA - set in Clicc's account settings.
The account's blended average - the last-resort fallback.
Each level is a gate: once one applies, the others are never consulted. A campaign with its own tCPA is never judged against the account average.
A campaign counts as hitting its target within a ±15% band (configurable) - so a $52 CPA against a $50 target is "hitting", not "missing by 4%". This stops normal variance generating churn.
RELATED ARTICLES
The eight lifecycle stages
From "no conversions yet" to "ready to scale up".
The tunable rules
All the defaults, and what each one controls.
Account targets & budgets
The two numbers the whole engine is judged against.