Meetings and interruptions are still the biggest obstacles for developers, even with AI

This post was originally published in Engineering Enablement, DX’s newsletter dedicated to sharing research and perspectives on developer productivity. Subscribe to be notified when we publish new issues.

With all the hype surrounding AI, and the many promises being made to increase developer productivity, one could be forgiven for thinking that AI represents the biggest opportunity for engineering teams to reduce friction and save time. We sampled over 130,000 data points related to developer experience, and the results were a (perhaps painful) reminder that AI is not a silver bullet for productivity.

Although it’s clear that AI is saving developers time, when we look at how much time is being lost to other tasks, those savings are eclipsed. We analyzed data on some of the most common sources of friction and time loss for developers across tens of thousands of developers. We then compared the time lost to these sources to the average time savings using AI.

When these time losses are annualized, the expense of these bottlenecks becomes strikingly clear. And when compared to AI time savings, it’s clear that though AI saves time, there are multiple opportunities for efficiency that significantly outweigh those benefits.

While developers on average report saving 3.35 hours a week with AI, they lose much of those savings to meeting-heavy days, as well as other significant time sinks such as daily interruptions and prerequisites such as code comprehension, information seeking, and Dev environment toil.

Despite these differences, it does bear mentioning that many of these bottlenecks are things that can be addressed, at least in part, by strategically integrating AI agents throughout the SDLC. Companies are seeing success in areas such as:

• Reducing overall review wait time by allowing agents to take a first pass at reviewing certain changes, such as our recent writeup of Faire’s Fairey agent solution
• Making code easier to comprehend, either through the automation of in-line comments, or even having agents and assistants explain the code
• Training AI agents on internal knowledge bases, easing access to information around code and architecture

Engineering leaders should first measure and surface the biggest bottlenecks to throughput in the organization, and then decide on the best ways to deal with those bottlenecks, whether through AI, or through other areas of process improvement. (Note: If you’re a DX customer, use the Workflow view in DX.)

While AI has not proven to be a full panacea for addressing productivity, by comprehensively measuring and analyzing developer experience, leaders can make informed decisions about the best ways to deal with productivity loss.