The 2025 engineering landscape is being defined by a single word: productivity. But as organizations rush to optimize delivery speed and team output, a critical insight is emerging from the latest data, sustainable productivity gains are nearly impossible without strategic investment in developer experience.
LinearB's 2025 Engineering Benchmarks Report analyzed data from over 3,000 organizations to uncover how elite engineering teams are balancing velocity with quality, speed with sustainability, and execution with alignment. The findings reveal a more nuanced picture of high-performing teams than many leaders expect, with surprising correlations between cycle time and quality, PR size and efficiency, and even project management hygiene and delivery risk.
For VPs of Engineering, Directors of Developer Experience, and CTOs planning their 2025 strategies, these benchmarks offer more than comparison points; they provide a roadmap for where automation, process improvements, and cultural shifts can have the most impact.
Benchmarks reveal where engineering teams unlock the most productivity
The 2025 benchmark report represents a significant evolution in how engineering leaders can evaluate team performance. Rather than relying on intuition or isolated metrics, the report provides external reference points that help leaders understand where their teams stand relative to industry standards, and more importantly, where to focus improvement efforts.
This year's report introduces several new benchmarks that offer more granular visibility into delivery flow. Two new metrics break down cycle time optimization into its constituent parts: approval time (the duration from first comment to PR approval) and merge time (the time from first approval to merge). These additions recognize that teams operate differently, and bottlenecks can emerge at distinct stages of the review process. For example, a team might have fast review cycles but slow merge processes due to integration failures or deployment pipeline constraints.
Another significant addition is PR maturity, which measures the ratio of changes added to a branch after a pull request is opened. This upstream signal indicates how ready code is for review when developers first submit it, a critical factor in reducing rework and review friction.
"What we're trying to do is help teams monitor their development progress and ensure that actual development work is being tied to defined tasks within the organization."
Perhaps most notably, the report has expanded beyond pure delivery speed to include project management hygiene indicators. These include metrics like issue hierarchy, which is the percentage of tasks linked to parent epics or stories, as well as branch-to-task traceability, estimation coverage for in-progress work, and clear ownership assignment. The primary goal of these metrics is to help teams monitor development progress and ensure actual coding work ties directly back to defined organizational tasks. While these metrics might sound mundane, they serve a critical purpose in connecting engineering output to business initiatives.
Understanding these metrics is vital. For instance, knowing when large proportions of work lack time estimates can be a helpful early warning sign that predictability risks are coming down the pipeline. These additions reflect a maturation in how engineering organizations think about productivity. Elite teams are fast, predictable, aligned, and able to demonstrate that their velocity translates into meaningful organizational value. The benchmarks provide planning inputs for 2025, helping leaders identify which metrics need attention and where automation or process changes could have the most impact.
Faster cycle times and smaller PRs improve quality and throughput
Developer productivity has emerged as the top priority for engineering organizations in 2025, but the benchmark data reveals that treating it as a vague goal is a recipe for failure. High-performing teams understand their current performance, identify weak metrics, and improve them deliberately through targeted interventions.
One of the most surprising findings from the report challenges a common assumption about the trade-off between speed and quality:
"The biggest one that was actually very surprising to me is related to cycle time, and that is specifically longer cycle times correlated with quality risks, specifically around change failure rate."
This correlation between faster cycle times and better quality outcomes, including lower change failure rates, reinforces that speed and stability can and should improve together. Organizations in the elite category for cycle time are roughly half as likely to fall into the "needs focus" category for change failure rate. This pattern extends to deploy time as well, where faster deployments also correlate with higher quality.
The explanation likely lies in the ecosystem surrounding developers. Teams with elite cycle times typically have robust automation, comprehensive testing, AI-assisted reviews, and streamlined processes. These same investments that enable speed also catch defects earlier and reduce production incidents.
Another major driver of flow efficiency is pull request size:
"PR size drives velocity. Smaller PRs reduce pickup time, they reduce merge time, and large PRs sort of do the opposite and actually even require more review modifications typically."
The data shows a cascading negative effect from oversized PRs. Large changes wait longer for initial review, take longer to merge after approval, and are more likely to require substantial modifications during the review process. Each of these delays compounds, creating friction that affects multiple team members and slows overall throughput.
Pull request maturity represents another efficiency lever. Better-prepared changes that require fewer commits after opening reduce rework, shorten review loops, and support higher merge frequency. When developers have clear expectations and leverage automated checks before requesting human review, they create less disruption for reviewers and maintain faster overall flow.
The report also surfaces a counterintuitive finding: teams with weak project management hygiene may appear faster in the short term. Organizations that don't consistently reference PRs against task tracking systems or maintain issue hierarchies often show faster cycle times. However, this speed comes at a cost, misaligned goals, hidden technical debt, and poor visibility into actual progress. The recommendation isn't to abandon hygiene practices, but to automate them so they don't become manual overhead that slows teams down.
Developer experience sustains productivity and delivery predictability
The central theme of the 2025 benchmark report is the inseparable relationship between developer experience and developer productivity:
"You can't actually improve productivity without making DevEx a significant part of that strategy."
Developer experience encompasses team morale, engagement with the organization, and the quality of tools, processes, and environments developers work within. Developer productivity measures how effectively and efficiently developers complete meaningful tasks with minimal waste. While these concepts are distinct, the data makes clear that sustainable productivity gains require strategic DevEx investment.
This framing repositions developer experience from a "soft" concern to a strategic investment area, especially critical when engineering leaders face pressure to deliver aggressive roadmaps while maintaining team health. The challenge many leaders face is justifying DevEx spend to business stakeholders who prioritize feature velocity above all else.
The report's DevEx investment profile provides a practical framework for measuring this balance. Analyzing data from approximately 3,000 organizations, the benchmark found that on average, teams distribute their effort as follows:
- 55% on new value: new features, roadmap work, platform applications
- 20% on feature enhancements: performance improvements, reliability work, quality initiatives
- 15% on DevEx: refactoring, test automation, dev tooling, reducing operational overhead
- 10% on maintenance: keeping the lights on, service maintenance
These percentages serve as recommendations rather than rigid rules, and every organization should adjust based on their specific situation. However, the critical takeaway is avoiding chronic underinvestment in developer experience.
Underinvesting in DevEx creates long-term delivery risks. Teams begin experiencing more unexpected work creeping into sprints, reduced predictability in delivery timelines, and a declining ability to maintain consistent output of new value. Technical debt accumulates, tooling friction increases, and the cognitive load on developers grows, all of which compound over time.
Predictability, accountability, and workload clarity emerge as important experience outcomes supported by the project management hygiene benchmarks. Time estimation coverage and clear assignment of in-progress work aren't bureaucratic overhead, they're signals that teams operate professionally with reduced friction. When developers understand their commitments, have visibility into priorities, and aren't constantly context-switching between undefined tasks, they experience less stress and deliver more reliably.
Microsoft is accelerating AI adoption as automation removes engineering friction
AI-driven automation is rapidly moving from experimental to essential, both in industry practice and in the benchmark data. Microsoft's recent internal reorganization, combining its developer division and AI platform teams with a stated goal of compressing 30 years of change into three years, exemplifies how seriously large organizations are restructuring around AI capabilities.
The conversation around AI in engineering contexts has matured beyond the simplistic narrative of whether AI will replace developers. The more productive framing positions AI as a force multiplier that automates lower-value tasks and allows developers to focus on higher-value work. This shift is about augmentation, not replacement. It realigns developers around the tasks that require human judgment, creativity, and strategic thinking while offloading repetitive or mechanical work to automated systems.
Automation emerges repeatedly in the benchmark data as the primary path to improving performance. For cycle time optimization, the fastest teams leverage extensive automation in testing, integration, and deployment. For PR maturity, AI-assisted tools can catch issues before human reviewers engage, improving the quality of initial submissions.
AI-assisted review, testing, and code-quality checks represent particularly promising automation opportunities. When developers can run comprehensive automated checks before requesting human review, catching style issues, potential bugs, test coverage gaps, and performance concerns, they submit cleaner PRs that require less back-and-forth. This reduces wasted reviewer effort and repeated interruptions, allowing human reviewers to focus on architectural decisions, business logic validation, and knowledge transfer rather than mechanical feedback.
The project management hygiene metrics also benefit from automation. The advice to simply automate operational overhead applies directly to status updates, branch-to-task linking, and progress tracking. When these activities happen automatically rather than requiring manual developer input, teams can maintain strong hygiene practices without sacrificing velocity.
The path forward for engineering leaders
The key insight is that automation removes friction, reduces cognitive load, and creates space for developers to do their best work. Organizations that strategically deploy AI and automation across their development lifecycle will find they can simultaneously improve productivity metrics, enhance developer experience, and maintain the alignment and predictability that businesses require.
The 2025 Engineering Benchmarks Report makes one thing abundantly clear. The year ahead belongs to organizations that take a sophisticated, data-driven approach to productivity. Reading the report, understanding where your team stands against industry benchmarks, and then using automation and AI to improve targeted metrics represents the clearest path forward. The teams that will thrive are moving fast with quality, alignment, and sustainability built into every process.
To hear a deeper breakdown of the 2025 Engineering Benchmarks Report and how to apply these insights to your team, listen to the full episode on the Dev Interrupted podcast.
