The Case for Developer Experience
Enterprise developers are being asked to do more than write and ship code. They are expected to navigate a growing array of tools, environments, and technologies—including generative AI assistants—while keeping up with security demands and broader business objectives. Yet the metric that matters most to the individual engineer isn't story points or deployment speed; it is the overall developer experience (DevEx).
DevEx is more than a buzzword; it is a working formula that accounts for how simple it is to make changes, how smoothly an idea moves to production, and how positively the work environment affects the individual. In practice, that means a developer's productivity, their ability to see impact, and their overall satisfaction. For leaders, the aim is to build a collaborative environment that fosters all three.
To understand the current state of that formula, we partnered with Wakefield Research to survey 500 U.S.-based developers working at enterprise companies. The findings paint a clear picture: organizations are evaluating engineers on the wrong things, spending too little time on the work that matters, and underestimating how central collaboration is to both code quality and job satisfaction.
Too Many Wrong Metrics, Too Many Waits
A clear gap exists between the metrics developers are judged by and the metrics they believe represent strong performance. Developers overwhelmingly think how they handle bugs and issues matters more than the raw number of incidents they resolve, and they rank collaboration and communication alongside code quality as a top performance determinant. Yet only 33% of survey respondents say their employer actually factors communication skills into performance reviews.
When asked what would have a positive impact on their workday, developers point to learning new skills (43%), getting feedback from end users (39%), automated tests (38%), and designing solutions to novel problems (36%). But their actual daily schedules are dominated by other tasks. Alongside writing code, developers report spending large chunks of time fixing security vulnerabilities—a sign of security's growing importance to engineering priorities.
The most striking finding is that developers report spending as much time waiting for builds and tests to run as they do writing new code. Despite a decade of investment in DevOps tooling, wait times remain a persistent drag. These delays reduce the opportunity for upskilling and for tackling new problems, and the survey indicates wait times have a major effect on overall satisfaction. Similarly, while developers value end-user feedback, they face friction getting it—product managers and marketing often stand between engineering teams and the people using their software.
The bottom line: Developers want to upskill, build novel solutions, and receive direct end-user feedback. Wait times on builds and tests, poor channels for feedback, and legacy performance metrics are the biggest obstacles standing in the way.
Collaboration Is the Core of the Job
Software development is rarely a solo pursuit. Survey respondents say they work with an average of 21 other developers on a project, and 52% work with other teams daily or weekly. Developers cite regular touchpoints as the most important factor for effective collaboration—but they define collaboration far more broadly than meetings and messages.
Three specific elements shape the quality of team interaction:
- Specified blocks of time with no team communication, which give developers the rare, valuable chance for uninterrupted work.
- Access to fully configured developer environments, which boosts consistency and eliminates the "but it worked on my machine" problem.
- Formal mentor-mentee relationships, which accelerate upskilling and help build the interpersonal skills necessary for a collaborative setting.
The inverse is also true. When these factors are absent or handled poorly, the impact on a developer's day is demonstrably worse—ineffective meetings exist to distract, not to help. Notably, when engineers describe effective collaboration, they connect it directly to better software: improved test coverage and faster, cleaner, more secure code. Simply put, when developers work well together, they believe they build safer products.
Even with this consensus, collaboration rarely makes it into performance reviews. Developers have adopted agile and DevOps practices predicated on teamwork, yet they are still individually evaluated on metrics that don't include it. Engineering leaders can respond by:
- Making collaboration a goal in performance objectives. Tie it to expected behaviors via lunch-and-learns, joint projects, and other shareable work.
- Defining what collaboration means inside the organization. Make the distinction between being informed and being consulted, and use responsibility matrices to codify roles—a practice in place in GitHub teams.
- Carving out time for people to connect informally. Remote and hybrid setups require scheduled, deliberate space for relationship-building.
- Identifying and elevating principal and distinguished engineers. Academic research shows change agents have an outsized effect—when model collaborators are promoted, others follow suit.
The bottom line: Effective collaboration produces better code and should be treated as a performance measure. Regular touchpoints, unimpeded focus time, consistent environments, and mentorship are the levers that make it work.
Developers Report Broad Gains from AI Coding Tools
Adoption of AI coding tools is now effectively universal among developers at large U.S. companies. According to a GitHub-commissioned survey of 500 U.S.-based developers at firms with 1,000+ employees, 92% report using an AI coding tool either at work or in personal time, and 70% say they see significant benefits from doing so.
The near-total adoption rate carries an implicit warning for engineering leaders: if developers are already using AI tools, the question is not whether to allow them, but whether those tools are enterprise-grade and governed by clear standards. Without approved options and usage policies, developers may turn to unvetted applications.
Beyond Code Volume: New Performance Metrics
Survey respondents say AI coding tools help them meet existing performance standards through improved code quality, faster output, and fewer production-level incidents. They also believe those metrics should carry more weight in performance evaluations than raw code volume.
That view is timely: roughly one-third of developers report their managers currently assess performance by the volume of code produced, and the same share expect that to remain true once AI tools are in use. Given that AI can inflate code output, leaders may need to reconsider whether code quantity remains a meaningful productivity signal or simply a proxy that AI has made obsolete.
Collaboration and Cognitive Offload
More than 4 in 5 developers surveyed (81%) say AI coding tools will increase collaboration within their teams and organizations. Security reviews, planning, and pair programming stand out as the most significant collaborative tasks developers expect to work on with AI assistance. The emphasis on security reviews suggests those processes will remain a critical checkpoint as AI-generated code becomes more common.
Developers also report that automating routine parts of their workflow frees time for solution design—shifting effort away from boilerplate code and toward new features and products.
AI's benefits extend to skill development and burnout prevention. More than half of respondents (57%) say AI coding tools help them improve their coding language skills, ranking as the top benefit they see. The same tools reduce cognitive load: 41% believe AI helps prevent burnout. Prior GitHub research reinforces this, finding that 87% of developers said Copilot helped them preserve mental effort on repetitive tasks.
The upskilling angle is notable because developers consistently rank learning new skills as the top contributor to a positive workday, yet 30% say learning and development can negatively impact their day—likely because it adds to an already full plate. AI tools offer a way to absorb new skills while working rather than treating learning as an extra task.
Efficiency Within Existing Workflows
Developers do not see AI as fundamentally restructuring the software development lifecycle. Rather, they report it layering into established processes—builds, CI/CD pipelines, security checks—and making them more efficient. That efficiency gain is what gives developers time back for solution-focused work.
The bottom line
Almost all developers (92%) are using AI coding at work—and they say these tools not only improve day-to-day tasks but enable upskilling opportunities, too. Developers see material benefits to using AI tools including improved performance and coding skills, as well as increased team collaboration.
What Engineering Leaders Should Do Next
Developer satisfaction, productivity, and organizational impact are all positioned for a boost from AI coding tools, and the usage numbers make clear the shift is not temporary. With 92% of developers already using AI, 70% reporting significant benefits, and 81% expecting greater team collaboration, the tools are positioned as a net positive for both velocity and developer experience.
Engineering and business leaders weighing how to improve the developer experience should consider three priorities:
- Help developers enter a flow state with tools, processes, and practices that enable productive, impactful, and creative work.
- Break down organizational silos and provide developers with efficient communication channels to empower collaboration.
- Build upskilling into developer workflows through AI investments that support experimentation and innovation.
Methodology
This report draws on a survey conducted online by Wakefield Research on behalf of GitHub from March 14, 2023 through March 29, 2023 among 500 non-student, U.S.-based developers who are not managers and work at companies with 1,000-plus employees. For a complete survey methodology, please contact [email protected].
