An impression answers only a small question
It documents that a delivery event was counted under the measurement definition being used. It does not automatically tell us whether a human was behind the contact, whether the ad was visible or whether the contact contributed to brand or sales impact.
Reach needs deduplication and quality
Impressions cannot be translated one-to-one into people. Reach requires a suitable identity or modeling logic. The underlying contacts should also be validated for relevance and quality; otherwise high delivery volume can create only the appearance of high reach.
Attention is a different measurement dimension
Viewability creates a technical minimum condition but does not prove conscious perception. Attention measurement attempts to capture or model additional signals. It should also disclose what was directly observed and what was inferred.
Effectiveness requires a causal question
If the question is whether media changed behavior or brand perception, delivery KPIs are insufficient. Depending on the objective, advertisers may need brand-lift studies, control groups, geo tests, experiments or other incremental designs. Attribution alone is not causality.
What good media management combines
Planning needs reach and frequency, operations needs delivery and quality data, forensics needs signal validation, and effectiveness research needs an appropriate causal method. The disciplines complement each other; no single metric replaces the rest.
This article describes principles and questions. Specific fraud, bot or quality rates cannot be transferred directly to individual campaigns without understanding methodology, scope and data basis.