Anthropic has announced the launch of a new suite of tools for Claude Enterprise customers to provide businesses with a clear understanding of the effectiveness and return on investment (ROI) of AI. New features include Smart reports, In-app surveys, and Pooled group budgets.
Anthropic has announced the launch of a new suite of tools for Claude Enterprise customers to provide businesses with a clear understanding of the effectiveness and return on investment (ROI) of AI. New features include Smart reports, In-app surveys, and Pooled group budgets.
The release comes amid strong financial results from the company. According to investor data, Anthropic turned an operating profit in the second quarter and had a net revenue retention rate of 900% — meaning the same customers spent nine times more in the third quarter of 2026 than in the same period in 2025. A key driver of this success was its AI coding tool, Claude Code.
As Anthropic reports on its blog, to group workflows, administrators can select a team, time frame, and products, after which Claude analyzes a sample of transcripts and groups tasks into work streams, showing the costs for each, the deliverables created, and points of failure, such as missing connectors or tool call errors.
In addition, reports generate a list of actions with recommendations on which repetitive processes should be turned into skills or added to the code repository, and administrators have the opportunity to ask custom questions about the use of models and save them as templates.
Delegation without loss of privacy has also been implemented: account owners can allow team leaders who are not administrators to run reports for their groups, although Anthropic specifically emphasizes that these reports are designed solely for planning investments and evaluating AI implementation, and not for assessing individual employee performance or making decisions about dismissal.
These tools complement quantitative data with qualitative feedback from employees, as short survey cards appear directly in the interface after Claude's responses.
Standard questions are aimed at measuring value: they help to find out what was most usefully done with AI, how the time spent on a task changed, and what exactly would break or slow down if access to Claude disappeared tomorrow.
At the same time, the list of questions can be customized to the needs of the company, and the collected results can be exported to tables for further analysis.
Anthropic also emphasizes that by default this data is not used to train new models.
Allows you to set a single monthly token limit for the entire team instead of rigid individual limits. This prevents a situation where active users exhaust their limit at the beginning of the month, while others' budget remains unused.
The system automatically suggests an amount based on recent activity and sends notifications to administrators when 50%, 75%, 95%, and 100% of the budget is reached.
All three new tools are now available to Claude Enterprise users. Availability may vary depending on security configuration (e.g., for organizations with HIPAA requirements or CMEK-based encryption keys).
