The tools are in every editor. The invoices are real. But when a leader asks the obvious next question — who on my team is using this well, and who is generating cleanup work — the honest answer is nobody knows. Self-report is a mood. Seat counts are a headcount. Neither tells you what's happening in the work.
That blind spot is expensive, because the distance between a median AI user and a great one isn't a few percent. It's an order of magnitude.
Cursor studied 18 months of real usage across the industry. The gap between the most AI-active developers and the median wasn't incremental — the top 1% instruct a network of agents to do the grunt work, then review, steer, and stitch. That's a learnable skill, and right now you can't see who has it.
more AI-generated lines of code per day than the median active user (the top 1% of AI-active developers)
more pull requests merged per week than the median active PR author
Source: Cursor — Developer Habits Report; Forbes coverage (May 2026). Figures compare the top 1% of AI-active developers to the median active user.
The business decision underneath it: if the gap is 46×, then the return on your AI tools isn't set by the seats you bought — it's set by how many of your engineers work like the top 1%. Moving the median even a little is worth more than any discount on the license.
Experienced developers felt AI made them 20% faster. Measured on real tasks, it made them 19% slower.
Only 36% of employees say their AI training was enough. The other 64% are improvising on a tool they use every day.
A 25% jump in AI adoption came with higher individual productivity, and a 7.2% drop in delivery stability. Teams felt faster while delivery got shakier.
DORA (Google), 2024 · per +25% AI adoptionRead the research →Every chart above is self-report or an aggregate. None can name who on your team uses AI well.
Claude Code ships new features weekly. The models change under you. The patterns the best engineers lean on today didn't exist last quarter. Keeping a whole team current on all of it is a full-time job — so we make it ours.
Your telemetry tells us how your team works. The frontier tells us how the best teams work right now. Promptster folds both into the recommendations every engineer and manager sees — so the guidance is never generic, and never stale.
Why 90% of engineers use AI but almost no org can say who uses it well — and what a real measurement looks like.
The case for treating AI-coding spend and leverage as something you measure per developer, not per invoice.
A concrete look at how invisible, always-on config quietly taxes every session — and how much it costs.
Detection is the wrong question. The one that matters is how well the tools your team already uses are paying off.
Bring a VP Eng or platform lead, and we'll show the manager view and an engineer's private view, and you decide in 15 minutes whether to connect a repo and see your own.