Anthropic Research Tracks Claude Code Usage and Economic Imp

Anthropic Research Tracks Claude Code Usage and Economic Impact

@AnthropicAI· June 16, 2026 View original

▶ The 60-second brief

Summary

Anthropic's latest economic research introduces a framework to track the scaling and usage of Claude Code, analyzing who uses it, for what purposes, and how task value and domain expertise influence success. The study found that the average task value increased by 27% and success rates across various occupations were comparable to software engineering.

Anthropic has released new economic research detailing a framework for monitoring the growth and application of its Claude Code AI assistant. The study investigates user demographics, common use cases, the evolving economic value of tasks performed, and the role of domain expertise in achieving successful outcomes. The research, based on an analysis of 400,000 sessions, revealed that over half of Claude Code's usage involves writing or repairing code, with nearly 20% dedicated to operating software. A key finding was the significant increase in the monetary value of the average session, which grew by 27% between October and April, based on comparisons to freelance marketplace costs. Furthermore, the study indicated that while domain experts tend to have higher success rates, the difference between intermediate and expert users is minimal, suggesting that general proficiency in a subject area is sufficient for effective code generation. Success rates across various professional fields were found to be within 7 percentage points of software engineering, highlighting the tool's broad applicability.

Why it matters

This research provides valuable insights into the real-world impact and adoption of AI coding assistants, helping professionals understand their economic value, potential applications across industries, and the importance of domain expertise.

How to implement this in your domain

1Evaluate the potential for AI coding assistants like Claude Code to enhance productivity in your specific domain.
2Investigate how AI tools can automate or streamline coding and software operation tasks within your organization.
3Train your teams on leveraging AI coding assistants, emphasizing the importance of domain knowledge.
4Monitor the economic value generated by AI-assisted tasks to justify investment in such tools.
5Contribute to or utilize frameworks for tracking the impact of AI on the nature of work.

Who benefits

Software EngineeringConsultingEducationFinanceLegal

Key takeaways

Claude Code is primarily used for writing, repairing, and operating software.
The economic value of tasks performed with Claude Code has significantly increased.
Domain expertise improves success, but proficiency is often sufficient.
AI coding assistants show comparable success rates across diverse occupations.

Original post by @AnthropicAI

"Our latest economic research introduces a framework for tracking Claude Code as it scales. Who is using Claude Code, and what are they using it for? How is the value of tasks changing? And how much does domain expertise shape whether a session succeeds? We compared Claude Code su…"

View on X