DX

ICYMI: join us tomorrow for a live walkthrough of what's new in DX. Our product team will demo several of our latest capabilities, including a first look at the new AI lifecycle impact report, which helps engineering teams measure AI's impact across the software development lifecycle. Save your spot: https://lnkd.in/gbzsyHhw

Friday is for sharing with peers! One of the tools in my bag that helps me be a better CTO (to my board, my peers, and my team) is DX . I can share transparency and benchmark where we are against peers without anecdote. The power of this is something I cannot work without now. “How are we doing on AI adoption?” “Whats blocking our teams from success?” “How do we compare to our peers, in our market, at our size?” I can answer with a click. The team at DX didn’t ask me to write this… they had no idea… but when something is so powerful… I felt it was only right to share with my friends here. Shout out to Ian O'Brien who first introduced me to this black magic (thanks man) and the team at DX who help me make it work, who I will tag in the comments. Data is power!

We recently released our Q2 State of AI Impact in Engineering Report, a quarterly analysis of 500+ engineering organizations. In this analysis, we see that AI spend has increased 28x since last year. Code output has increased as a result, but innovation has remained flat and new bottlenecks have emerged. Download the full report: https://lnkd.in/g6YhuYzS

Proud to announce the launch of our Q2 2026 State of AI Impact in Engineering report -- hands down our best AI impact report yet. With adoption at over 95%, and more than 50% of production-merged code now generated by AI, it was time to move beyond comparison of user cohorts into deeper, outcome-based research. We found dozens of interesting trends, including: - AI spend is accelerating faster than outcomes. Median quarterly organizational AI spend climbed from ~$1.5K to ~$44K over four quarters. Leaders who cannot connect this investment to downstream outcomes (feature velocity, innovation ratio, quality) may face increasingly difficult budget conversations - Saved time isn’t converting into innovation. AI users are now saving an estimated 4 to 6 hours per week. However, the innovation ratio, defined as the percentage of time spent on building new features versus maintenance and overhead, has remained flat over the same period of study. - AI is making codebases easier to understand, but it’s also making the code it generates harder to trust. From Q1 2026, Code Maintainability improved by 3.8%, whereas Change Confidence decreased by 6.1%. 👉 Read the full report here: https://lnkd.in/exWdZBA5 We'd love to hear your feedback in the comments!

This Thursday, join our live discussion of the findings from the new AI Impact Report, hosted by Brian Houck and Justin Reock. They’ll discuss: - Why isn't rising AI spend translating into better engineering outcomes? - What does AI ROI actually look like beyond adoption and usage metrics? - Why are developers becoming less confident in code they rate as becoming easier to understand? Save your spot: https://lnkd.in/gbfSYeJb

New on the Engineering Enablement podcast: Sejal Amin and Pedro Gutierrez share how Priceline went from a project-based organization to a product operating model. They cover: ❒ Which qualitative signals revealed organizational bottlenecks that system-based  metrics missed ❒ Why engineering managers at Priceline now own quarterly developer experience goals ❒ How having measurement frameworks already in place gives them visibility into where bottlenecks shift as AI accelerates code output Listen to the full conversation: https://lnkd.in/gPzGxEym

New data from 500+ engineering organizations: AI spend jumped from ~$1.5K in Q3 2025 to ~$44K by Q2 2026, yet innovation ratio has essentially remained flat over the same four quarters. AI-driven time savings are not yet visibly converting into new value creation. This suggests spend is outrunning the return it’s meant to justify. Join Brian Houck and Justin Reock for a live readout of the Q2 2026 AI Impact Report releasing next week: https://lnkd.in/gbfSYeJb

UKG’s Thomas Newton recently shared how they built a system to guide AI adoption and assess whether it’s translating into meaningful engineering outcomes. They built a metrics dashboard that combined AI usage patterns, TrueThroughput, and spend awareness to help managers have stronger coaching conversations with teams. Some of their findings and lessons: - Senior engineers saw outsized throughput gains after adopting AI - Managers and directors increased hands-on technical contribution - Metrics work best when they enable better conversations, not as surveillance Read more about UKG’s approach: https://lnkd.in/gpqQQQgQ

Where is AI making software engineering faster, and where is it creating new pressure on our workflows? Some outstanding researchers across our industry are beginning to provide answers. In this week's Engineering Enablement newsletter, I summarize five recent papers that have fundamentally changed how I'm thinking about AI-assisted software engineering, and why I believe the next generation of AI tooling will be less about generating code and more about helping humans understand, validate, and deliver it. Across these studies, I kept seeing the same pattern emerge: - AI coding assistants substantially increase coding throughput. - Those gains diminish as work moves through the delivery pipeline. - Developers increasingly want AI to help with verification, review, debugging, and understanding (not just code generation). - The next challenge isn't writing more code. It's preserving shared understanding while shipping it safely. The bottleneck hasn't disappeared. It has moved. I'd love to hear whether these findings match what you're seeing in your own organization. And if there are other papers you think belong in this conversation, I'd love the recommendations. The newsletter can be found here: https://lnkd.in/eCXmeXAU #AI #SoftwareEngineering #DeveloperExperience #EngineeringLeadership #DeveloperProductivity

"If every software organization were twice as productive as they were a year ago, why aren't they making twice as much money?" Justin Reock recently hosted a readout of the upcoming State of AI Impact in Engineering (Q2 report), sharing findings that indicate throughput is accelerating faster than organizations can convert that capacity into business outcomes. A few highlights: - 50%+ of merged production code is now AI-generated, up from 30% three months ago - Engineers report saving 6 hours/week, but that saved time isn't translating into more new features shipped - PR size has nearly doubled in the last year. Every extraneous line is a potential bug, vulnerability, or review bottleneck - Deep work scores are declining as engineers context-switch across multiple agent CLIs Stay tuned for the release of the full Q2 report later this month.