'Tokenmaxxing' Hinders Developer Productivity, Study Finds

There's a classic management saying: What gets measured gets managed. And typically, you end up with more of whatever you're tracking.

Software engineers have debated productivity metrics for decades, starting with lines of code. But as the new generation of AI coding assistants produces more code than ever, what their managers should be measuring is becoming less clear.

Huge token budgets—essentially, the amount of AI processing power a developer is allowed to use—have become a status symbol among Silicon Valley developers. However, that's a strange way to think about productivity. Measuring an input to the process makes little sense when you presumably care more about the output. It might be logical if you're trying to encourage more AI adoption (or selling tokens), but not if your goal is to become more efficient.

Consider the evidence from a new category of companies in the "developer productivity insight" space. They're finding that developers using tools like Claude Code, Cursor, and Codex generate significantly more accepted code than before. But they also discover that engineers have to return to revise that accepted code far more frequently, undermining claims of increased productivity.

Alex Circei, the CEO and founder of Waydev, is building an intelligence layer to track these dynamics. His firm works with 50 different customers that employ over 10,000 software engineers. (Circei has contributed to TechCrunch in the past, but this reporter had never met him before.)

He says engineering managers are seeing code acceptance rates of 80% to 90%—meaning the share of AI-generated code that developers approve and keep. However, they're missing the subsequent churn that happens when engineers have to revise that code in the following weeks, which drives the real-world, lasting acceptance rate down to between 10% and 30%.

The rise of AI coding tools led Waydev, founded in 2017 to provide developer analytics, to completely overhaul its platform in the last six months to address the proliferation of rapid coding tools. Now, the company is releasing new tools that track the metadata generated by AI agents, offering analytics on the quality and cost of their code to provide engineering managers with deeper insight into both AI adoption and its effectiveness.

While analytics companies have an incentive to highlight the problems they find, evidence is mounting that large organizations are still figuring out how to use AI tools efficiently. Major companies are taking notice—Atlassian acquired DX, another engineering intelligence startup, for $1 billion last year to help its customers understand the return on investment from coding agents.

Data from across the industry tells a consistent story: more code is being written, but a disproportionate amount of it isn't sticking.

GitClear, another company in this space, published a report in January that found AI tools increased productivity. However, its data also showed "regular AI users averaged 9.4x higher code churn than their non-AI counterparts"—more than double the productivity gains the tools provided.

Faros AI, an engineering analytics platform, drew on two years of customer data for its March 2026 report. The finding: code churn—lines of code deleted versus lines added—had increased by 861% under high AI adoption.

Jellyfish, which bills itself as an intelligence platform for AI-integrated engineering, collected data on 7,548 engineers in the first quarter of 2026. The firm found that engineers with the largest token budgets produced the most pull requests (proposed changes to a shared codebase), but the productivity improvement didn't scale. They achieved two times the throughput at ten times the token cost. In other words, the tools are generating volume, not value.

These kinds of statistics ring true when you talk to developers, who are finding that code review burdens and technical debt are accumulating, even as they enjoy the new capabilities of these tools. One common finding is the difference between senior and junior engineers, with the latter accepting far more AI-generated code and consequently dealing with a larger amount of rewriting.

Still, even as developers work to understand exactly what their AI assistants are doing, they don't anticipate turning back anytime soon.

"This is a new era of software development, and you have to adapt. Companies are forced to adapt," Circei told TechCrunch. "It's not a passing trend."