Change failure rate by industry

Fintech, healthcare, SaaS, e-commerce: the cohort cuts and the release-train confound.

Oliver Wakefield-Smith

Founder, Digital Signet. Built siemcost / monitoringcost / itbudgetcalculator.

Updated 21 Jun 2026

Direct answer

Industry CFR variance in the DORA cohort is smaller than the within-industry varianceDORA 2024 Report2024. Regulated industries (fintech, healthcare) do not uniformly report lower CFR; deploy-gate cadence (release-train vs continuous) is the larger confound.

The cohort medians

Horizontal SaaS clusters near the 10-12% median. Fintech runs slightly lower (8-10%) but with notable bimodality between continuous-deploy and release-train fintechs. Healthcare and regulated SaaS cluster in the 12-16% range, reflecting longer change-management cyclesDORA 20242024.

The release-train confound

Release-train teams deploy less frequently, batch more change per deploy, and therefore record a different CFR shape than continuous-deploy teams. The lower deploy frequency makes per-deploy failures more visible and longer to remediate, partly offsetting the lower count.

What to benchmark against

Match your industry AND your deploy-gate cadence. A continuous-deploy fintech should benchmark against the continuous-deploy SaaS distribution more than against release-train fintechs.