changefailurerate.com
verified 21 Jun 2026

Change failure rate, measured properly.

Type your deploy numbers in. The needle below moves to your CFR, plots your team against the 2025 DORA percentile distribution DORA 2024 Report2024, and prices the annual engineering burn in dollars. Zero account. Zero demo. Math runs client-side.

OW
Oliver Wakefield-Smith
Founder, Digital Signet. Built siemcost / monitoringcost / itbudgetcalculator.
Updated 21 Jun 2026
Direct answer
Change failure rate is the share of deployments to production that result in a failure requiring rollback, hotfix, or downtime. In the 2025 DORA distribution, only the top ~16.7% of teams achieve sub-2% CFR DORA 2024 Report2024, the median sits near 10%, and a long tail extends past 30%. CFR is one of the four DORA keys DORA Four Keys2024, paired with deployment frequency, lead time for changes, and mean time to restore.
Window:
0%25%50%+
14.3%
37th percentile vs 2025 DORA distribution
14
Failed
111
Total deploys
across 30 day window
Annual engineering burn at this CFR
$186,960/yr
Customer-impact churn estimate: $447,720/yr at 35% customer-facing failure rate.

Where your team plots vs DORA cohort

CFR (y) by deploys per week (x). Elite teams sit bottom-right: high throughput, low failure.

0%10%20%30%40%010204060youdeploys per weekCFR %
elite cohort high medium low your team

Would a DORA vendor pay back at this CFR?

Assumes a 40% CFR reduction (the published Sleuth median, scaled). Quote-only vendors do not display a per-seat rate.

Free tier (up to 10 developers, basic DORA metrics)
Vendor / mo
$4,500
Burn saved / mo
$6,232
Payback
0.7 mo
Essentials $29/user/month (annual, includes 1,000 credits)
Vendor / mo
$5,220
Burn saved / mo
$6,232
Payback
0.8 mo
Quote-only (typical deals $30K-$100K+ ARR)
Quote-only vendor. We do not infer per-seat rates for this site; contact sales for a published number.
Faros CE: free self-hosted. Faros AI hosted: contact sales.
Quote-only vendor. We do not infer per-seat rates for this site; contact sales for a published number.
Quote-only (positioning to 500+ engineer orgs)
Quote-only vendor. We do not infer per-seat rates for this site; contact sales for a published number.
$20/developer/month (published)
Vendor / mo
$3,600
Burn saved / mo
$6,232
Payback
0.6 mo
Pluralsight Flow (Appfire)
pricing source
Quote-only (Appfire-managed)
Quote-only vendor. We do not infer per-seat rates for this site; contact sales for a published number.
Quote-only (per developer)
Quote-only vendor. We do not infer per-seat rates for this site; contact sales for a published number.

Why this calculator exists

Every SERP competitor ships a printed formula and a static table. Sleuth, LinearB, Faros, DX and Jellyfish all maintain CFR landing pages but none let a VP Eng paste their own deploy counts in and walk out with a defensible board number in 90 seconds. That gap is what this site fills.

The math is the standard DORA definition: failed deploys divided by total deploys over a rolling window DORA Four Keys2024 Google Cloud CFR Guide2024. The default exclusions (fix-only forward-fixes, externally caused incidents) follow DORA guidance and the operational consensus across the four vendors above; both are toggleable.

What the gauge tells you

The semicircular dial maps your CFR onto the 2025 DORA percentile bands. The mint sliver at the left is the sub-2% region only the top 16.7% of surveyed teams reach DORA 2024 Report2024. The sulfur band covers the broad middle. The long graphite tail past 15% is where most teams actually sit, despite the marketing.

Below the dial, the dollar-burn ticker translates your CFR into an annual engineering cost using your MTTR, dev count, and blended FTE rate. Default $180,000 FTE is a defensible US Series-C blended anchor; override it to your number. The customer-impact churn estimate uses a $7,500 per customer-facing failure mid-anchor (see /methodology for derivation).

What about the four DORA keys together?

CFR is half of the DORA stability pair (MTTR is the other half) DORA Four Keys2024. The throughput pair is deployment frequency and lead time for changes. Looking at CFR in isolation is the most common DORA mistake; the 2D scatter below the calculator shows why (high-DF teams sustain low CFR through small batches and fast rollback, not through deploying less).

For the framework explainers, see /dora. For deeper cuts (by team size, industry, deploy frequency), see /benchmarks/2025-dora-report.

Continue reading