Kometrics

SaaS teams that bill customers through one or more providers often end up with revenue dashboards that disagree with each other or misrepresent reality, for example showing a customer who switched billing providers as both a churn and a new signup instead of one continuous history. Standard dashboards report that MRR moved but do not explain whether new cohorts retain better than old ones, which days cash actually lands, or which revenue is already at risk of leaving. When a number looks wrong, teams are left arguing about the chart instead of being able to open it and see the customers and invoices behind it. Getting to a trustworthy, presentable revenue number today requires manually resolving proration, coupons, plan changes and multi-currency billing across systems.

Kometrics ingests the billing data a company already runs across providers and merges it into one ledger, resolving proration, coupons, plan changes and multi-currency at ingestion so the reported figures hold up under scrutiny. It presents standard metrics (MRR, ARR, growth and churn rates, subscribers, ARPA, LTV) with the period-by-period table behind every chart, and six explorer tools (Forecast, Cohorts, Maps, Goals, Revenue Calendar, Risk Radar) that answer specific questions like where MRR is trending, which days cash arrives, and which revenue carries a warning sign. Every customer record carries their own MRR history, subscriptions, movements and invoices, and past-due customers are flagged separately from churned ones. Kometrics AI reads the same ledger as the reports, is scoped to one workspace at the query layer, and answers questions with a linked report rather than a generic estimate; it is also exposed over the Model Context Protocol so the same tools work inside Claude, ChatGPT, Claude Code and Codex. A read-only REST API with signed webhooks and an iOS app with home/lock screen widgets round out access to the same data.

The product is framed explicitly as "Subscription & Revenue Analytics for the AI Era," with its differentiator being that the same revenue ledger can be queried conversationally through AI assistants like Claude and ChatGPT over MCP.