Revenue at Risk is a forward-looking exposure view: how much MRR could you plausibly lose soon, through which doors, and what would it cost you? Where Health Score describes your customers, Revenue at Risk prices the downside. Every number here is an estimate, and the dashboard says so.

You'll find it under Revenue Pulse → Revenue at Risk, in five tabs.

A customer is considered "at risk" when they trip one or more concrete, observable risk channels, each carrying a set probability of churn:

If a customer trips more than one channel, we use the highest probability among them rather than adding them together. From there:

The signals feed for risk, watching for jumps in expected loss, clusters of customers tripping the same channel at once, and clusters of upcoming card expirations worth a single proactive campaign.

The headline numbers: your loss range (expected loss and worst case shown together), how expected loss has moved over the last 30 days, at-risk MRR as a share of total MRR, and the customer count behind it.

One card per channel, showing customer count, MRR exposed, and share of total exposure — a clean breakdown since each customer is only counted once, under their single highest-probability channel. Use the mix to triage: delinquency dominating means focus on recovery; card-expiring dominating means run a card-update campaign; critical health dominating means the risk is relationship-shaped, not billing-shaped.

Expected loss and worst case charted over the last ~90 days, so you can tell whether today's number is a temporary spike (common at the start of each billing month) or a genuine upward slide.

Your call list — every at-risk customer ranked by expected loss (largest first), with their MRR, which channel(s) they tripped, their probability, and a link to their full Health Score detail. This is meant to be worked top to bottom.

Why doesn't this match Health Score's "MRR at Risk"? They answer different questions. Health Score's number covers all MRR in the Struggling + Critical tiers — a broad quality measure. Revenue at Risk only counts customers tripping a concrete, near-term risk channel, so it's typically a narrower, more time-sensitive number.

Is expected loss a prediction of next month's churn? It's a probability-weighted exposure estimate, closer to an insurance estimate than a forecast — its job is to stay consistent over time so changes in it are meaningful, and to help you triage correctly.

Why did exposure spike at the start of the month? Renewal charges land at month start and some fail temporarily before retries succeed. This usually resolves within days — a rise that persists past that window is worth a closer look.