Anthropic rolls out ROI dashboards as Claude Code spending jumps 10x

Key Points

1. Anthropic launched reporting tools letting enterprises track what their token spending on Claude Code actually produces.
2. The company told investors it posted a Q2 operating profit and 900% net revenue retention.
3. Most Fortune 500 firms report little gain from AI, making proof of return the industry's central sales problem.

The latest:

Anthropic unveiled a suite of reporting tools Wednesday that let enterprise managers see what a team produced over a given period, how much it cost, and where its workflows broke down. The release follows financials shared with investors this week showing an operating profit in the second quarter and 900% net revenue retention, meaning spending from the same customers grew nine times between Q3 2025 and Q3 2026.

Details:

• What the tools do: The reports draw on internal analytics and employee surveys, letting managers view a team’s output across a chosen date range alongside its cost and the points where workflows failed. They also generate recommendations on improving results. Anthropic says it will not use this data for model training by default.
• Budget controls: Anthropic is also creating pooled group budgets so teams can better manage token spending, pairing visibility with a spending cap mechanism. Both address the same customer complaint about costs growing faster than planned.
• The financials: Anthropic recorded an operating profit in the second quarter, according to figures shared with investors. Net revenue retention of 900% is the headline number: it measures growth in spending from existing customers rather than new accounts, and points to deepening usage inside companies already on the platform.
• The adoption gap: Usage among venture-backed firms is driving token spend, but most of the Fortune 500 are not seeing major gains from deploying AI tools. That gap is the commercial problem the new reports are built to address, by showing customers not just what they spent but how to spend it better.
• Third-party data: Jellyfish, which makes tools for software developers, released a report based on agentic coding tool use by 276,000 of its customers. Claude Code and other tools built on Anthropic models account for nearly a third of all proposed code changes submitted for review.
• The cost curve: Spending on those tools has grown 10 times since the start of the year, according to Jellyfish, with their cost now equal to roughly 2% of total engineering cost. Nicholas Arcolano, Jellyfish’s chief scientist, told TechCrunch that the growth rate is what most CFOs did not plan for.
• The competitive picture: OpenAI’s Astra model, released this summer, looked like a potential threat but has not slowed Anthropic’s adoption numbers so far. Nikolas Albarran, Jellyfish’s AI product lead, said Codex took a long time to register in their data at all and has risen with Astra.
• Why customers stay: Albarran attributed Anthropic’s position to enterprise execution and incumbency rather than model superiority, saying companies standardized workflows around tools that worked and lack a big enough reason to switch. Even broadly similar tools, he added, carry quirks that make switching costs real.
• The consultant angle: Major AI labs are investing in forward-deployed engineering teams that embed consultants inside customer organizations. Albarran framed Anthropic’s new reporting tools as a lighter version of that hands-on help, delivered as software rather than staff.

Between the lines:

The 900% retention figure and the 10x spending growth describe the same phenomenon from two sides: customers already inside are spending far more, not switching. That makes Anthropic’s problem less about winning accounts than defending the spending inside them, which is what a dashboard proving return on investment is designed to do. Albarran’s point about non-zero switching costs suggests incumbency, not benchmarks, is the moat.

What’s next

Watch whether Fortune 500 deployments start showing measurable gains, whether OpenAI’s Astra registers in third-party usage data in coming quarters, and whether Anthropic discloses adoption figures for the new reporting suite.