For a few years, the DORA metrics had been the gold commonplace when it got here to measuring software program growth efficiency, however in accordance with JetBrains State of Developer Ecosystem 2025 report, builders right this moment need to be measured on extra than simply technical metrics.
DORA tracks 4 metrics: frequency of deployments, period of time between acceptance and deployment, frequency of deployment failures, and the way lengthy it takes to recuperate from failures.
In keeping with JetBrains’ report, 51% of respondents say that technical elements are crucial to efficiency and 62% say non-technical elements are. “Inner collaboration, communication, and readability at the moment are simply as vital as quicker CI pipelines or higher IDEs,” JetBrains wrote.
Sixty-six % of respondents consider that present metrics don’t replicate their true contributions. Eighty-nine % say that non-technical elements—job design, clear communication, peer and supervisor help, and actionable suggestions—affect their productiveness, whereas 84% say technical elements—efficiency and reliability of growth instruments—affect their productiveness.
“The info suggests it’s time to rethink how we measure success and to construct work environments that reward not simply outcomes, however the best way they’re achieved,” JetBrains wrote.
The report additionally checked out how builders are utilizing AI instruments, with 85% saying they recurrently use AI instruments of their growth workflow. For these utilizing it, nearly 9 out of 10 builders say they save a minimum of an hour each week and one in 5 saves greater than eight hours.
Nonetheless, there may be nonetheless a mixture of emotions round AI’s position in society, with 51% having constructive emotions, 28% having adverse emotions, and 18% feeling unsure.
In keeping with the report, the event actions that builders are most probably to let AI deal with embody writing boilerplate code, trying to find growth info on-line, translating code into different languages, writing code feedback or documentation, and summarizing current code adjustments.
The largest issues builders have about utilizing AI instruments embody inconsistent high quality, the restricted understanding AI instruments have of advanced code and logic, privateness and safety dangers, the potential adverse affect on their very own coding expertise, and lack of context consciousness by the AI instruments.
The report additionally checked out language recognition, highlighting the truth that TypeScript has seen a major rise in real-world utilization over the previous 5 years, as have Rust, Go, and Kotlin.
In keeping with JetBrains, the highest 5 languages that builders wish to undertake subsequent embody Go (11% of builders), Rust (10%), Python (7%), Kotlin (6%), and TypeScript (6%).
JetBrains performed this survey between April and June of this yr, and the ultimate report contains knowledge from 24,534 builders.