# Facebook ads benchmark review

Complete one copy per source row and account cohort. Write **Unknown** for missing evidence and **not applicable** where a field does not apply. Never treat blank or Unknown as zero. Record a public value for context even when comparison is held. Calculate a difference only when the same metric has a supported definition, unit, denominator, objective, scope, attribution basis, and suitable time basis on both sides. Supply the business threshold independently; never derive it from the source value.

## Reader business and offer

Business type, buyer, offer, and intended action:

## Source URL and publication date

Exact URL and publication or update date:

## Source retrieval and reporting period

Date retrieved; period covered by underlying campaigns or data:

## Source objective and industry

Platform, campaign objective, and the source's exact industry label:

## Source geography and currency

Country or region; currency and any conversion:

## Source sample and aggregation

Sample size, inclusion rules, median/mean/other, and row-level sample if known:

## Metric name and definition

Source's exact metric label, unit, and formula if supplied:

## Numerator and denominator

Exact counted numerator and denominator on both sides; Unknown where absent:

## Click or conversion definition

Link click, all click, landing-page view, lead event, CRM event, or Unknown on both sides:

## Attribution and date basis

Attribution setting, timezone, whether dated by impression/click/conversion/CRM event, and lag on both sides:

## Source value and missing details

Quoted value with unit; list each detail the publisher does not specify:

## Account scope and metric value

Account or cohort identifier, platform, objective, industry/offer, geography, window, currency, metric value, and extraction date:

## Account maturity and CRM linkage

Time allowed for outcomes to arrive; stable join key, matching rate, exclusions, and status or Unknown:

## Qualification rule and outcome counts

Written rule; raw and qualified counts; verified zero versus missing/immature:

## Comparability decision and reason

Aligned, descriptive context only, mismatched, or pending; name each matching and missing/different field:

## Absolute and relative difference

Only if aligned: account minus reference, with unit; (account minus reference) / reference × 100%, with direction. For two own cohorts, name the earlier denominator. If the aligned reference is zero, compute the absolute gap when both values are known, and hold only the relative gap because it would divide by zero. If a required value or alignment detail is Unknown, hold the affected calculation. For rates, absolute difference is percentage points and relative difference is percent:

## Raw and qualified lead cost

Same-cohort spend / raw leads and spend / qualified leads. Positive spend with verified zero outcomes has no finite cost; 0 / 0 is undefined. Unknown counts produce Unknown cost:

## Business threshold and provenance

Owner-supplied maximum cost or minimum outcome, amount and unit, source/date, and business logic. If absent, write Unknown; do not take a target from the public benchmark:

## Next check and responsible owner

Specific data, decision, and who will check by when:

## Unknowns and limitations

List unresolved definitions, small samples, reporting lag, inconsistent scopes, and what this record cannot conclude: