per week
days
hours
Today (2025 report): where you stand against everyone else
The 2025 report no longer ranks teams into levels: it publishes how many teams gave each answer. “Top X%” means X% of teams are at your level or ahead.
Deployment frequencytop 44.6 %
Lead timetop 56.4 %
Recovery after a failuretop 56.5 %
Change fail ratetop 62.2 %
If it were another year: your level according to each report
| Report | Deployment frequency | Lead time | Recovery after a failure | Change fail rate | Closest to |
|---|---|---|---|---|---|
| 2019p. 18 | High | High | High / Medium | Elite / High / Medium | High4 of 4 |
| 2021p. 9 | between bands | High | High | Elite | High2 of 4 |
| ⚠ The table has gaps: some answers don't fall into any level. | |||||
| 2022p. 11 | between bands | High | High | High | High3 of 4 |
| ⚠ No Elite that year: “no cluster showed enough of the characteristics of last year's elite cluster”. | |||||
| 2023p. 12 | High | High | High | High ≈ | High4 of 4 |
| ⚠ Published erratum: the Low row repeats Medium's frequency and lead time.⚠ The change fail rate is each group's typical value, not a threshold: it's compared with the closest one.⚠ MTTR is renamed “failed deployment recovery time”. | |||||
| 2024p. 13 | High | High | High / Medium | Medium ≈ | High3 of 4 |
| ⚠ The change fail rate is each group's typical value, not a threshold: it's compared with the closest one.⚠ The Medium group failed less (10%) than the High one (20%).⚠ Rework appears as a 5th metric, but the levels are still computed with four. | |||||
“≈” = the change fail rate is compared with the group's typical value, not a threshold. “A / B” = that year's table doesn't separate them. “Between bands” = your answer falls in a gap of the table. The last column says which row of the table most metrics match; DORA grouped by cluster analysis, so this is an approximation, not the “official” level.
The seven 2025 archetypes
They replace the levels, but CANNOT be assigned from these five metrics: they were computed from eight factors including burnout, friction and product performance. They help you recognise yourself, not score yourself.
Foundational challenges10 %Survival mode: low indicators, high burnout and friction, unstable systems.
The legacy bottleneck11 %Delivers regularly, but instability rules: lots of reactive work and a weak product.
Constrained by process17 %Stable systems, but inefficient process eats the effort: high burnout and little impact.
High impact, low cadence7 %Strong product and little friction, but low throughput and instability.
Stable and methodical15 %High quality and value at a deliberate pace: low throughput, high stability.
Pragmatic performers20 %Good speed and stability; burnout and friction around average.
Harmonious high-achiever20 %Everything positive: stable, low friction and no burnout.