Coverage
How much of what was planned was actually executed — always reported separately from the pass rate, so a high score on thin evidence cannot hide.
Definition
Coverage is the share of planned, applicable test cases that were actually executed: executed ÷ planned, counted in cases, not weighted by outcome. It is deliberately kept separate from the pass rate. A product that passes 10 of 10 executed cases has a 100% pass share — but if 90 cases were planned, its coverage is 10%, and MC Testing reports both numbers side by side so the thin evidence is visible.
A second, structural kind of coverage compares a product against its profile: how many of the profile's required modules were matched to features discovered in the product, and how many remain gaps. Under the platform's Section 22 rule, a gap is recorded for review — a module may sit behind authenticated navigation, be nested inside another module, be plan-gated, or simply not enabled in that installation — and a gap is never automatically a product defect.
How MC Testing applies it
Coverage drives the readiness gates directly: if P0/P1 coverage is below the proposed 50% threshold, the label is INSUFFICIENT EVIDENCE no matter how green the executed tests look. Quarantined flaky cases are excluded from coverage denominators (they still run and are reported), and blocked cases — for example those waiting for TEST-QA credentials — count as planned but not executed, which is exactly why credential-blocked journeys show up as an evidence problem rather than being quietly dropped.
On a product's profile page the console shows the full coverage snapshot: modules matched, gaps with their reasons, discovered-but-unmapped features, and executable versus blocked case counts. See also Readiness Score and Provisioning.