Anthropic launches reporting tools to help companies track AI spending

Tracking the ROI of AI

Anthropic recently shared financial data with investors that highlights the growing success of its AI coding tool, Claude Code. The company reported a 900% net revenue retention, meaning that customers who used the tool in Q3 2025 spent nine times as much in Q3 2026. This suggests that enterprise users are not just trying the software but are integrating it deeply into their daily operations.

To support this growth, Anthropic is launching a new suite of reporting tools for its business customers. These tools are designed to provide a granular look at what companies are actually getting for their token spend, which is the unit of measurement for AI processing. The goal is to help managers see the specific value generated by their investment in the technology.

The move comes at a time when many large companies are struggling to see major gains from AI deployment. While venture-backed firms are driving a lot of the current spending, many Fortune 500 companies are still trying to figure out how to make the technology work for them. Anthropic is positioning these tools to help bridge that gap by providing clear evidence of productivity.

How the reporting tools work

The new reporting suite uses internal analytics and employee surveys to give managers a clear picture of team performance. Users can view work completed over specific date ranges and see exactly what the team produced during that time. This allows for a direct comparison between the volume of work finished and the cost associated with the AI tokens used.

Beyond just showing costs, the reports will identify where specific workflows broke down. This helps teams pinpoint bottlenecks in their processes where the AI might be struggling or where human intervention is required. Anthropic also stated that the reports will offer specific recommendations on how to improve results based on this data.

Privacy is a key component of this rollout, as Anthropic stated it will not use the data from these reports for model training by default. This is a significant consideration for enterprise clients who are often wary of their internal workflows being used to improve a general product. By keeping the data separate, the company hopes to encourage more transparent reporting from its corporate users.

The rise of agentic coding

A recent report from Jellyfish, a company that creates tools for software developers, shows how quickly the market is adopting agentic coding tools. These are AI systems capable of performing multi-step tasks like writing and reviewing code. The survey found that tools built on Anthropic models, such as Claude Code, are among the most widely used in the industry.

According to the Jellyfish data, these tools are responsible for nearly one-third of all proposed code changes submitted for review. Despite the high volume of use, the cost of these tools currently accounts for about 2% of total engineering costs. However, spending on these tools has increased 10x since the beginning of the year, a trend that many chief financial officers did not anticipate.

Nicholas Arcolano, the chief scientist at Jellyfish, noted that while 2% may seem small, the 10x growth is a major shift. He explained that companies are reaching a point where they realize it is not enough to just spend money on AI. Instead, they now have to optimize the tools to ensure they are actually making the work more efficient.

Creating high switching costs

Anthropic is also introducing pooled group budgets to help teams manage their token spending more effectively. This allows for better oversight of how different departments share resources. By providing these management features, Anthropic makes it easier for large organizations to standardize their workflows around its specific ecosystem.

Industry observers suggest that these features create a high cost of switching to a competitor. Once a company has spent the time and effort to set up workflows, standardizing data, and integrating specific tools, moving to a different provider becomes difficult. Even if a competitor releases a newer model, the friction of moving all existing data and processes can be a major deterrent.

This strategy is bolstered by the fact that Anthropic is investing in forward-deployed engineering teams. These are consultants who embed themselves inside customer organizations to help get the AI up and running. By combining these hands-on consultants with automated reporting tools, Anthropic is making its products more 'sticky' for the enterprise market.