Built for: Customer success, retention and commercial leaders.
What you can do
Contracts and repeat buyers
Score customers whose contract comes up for renewal, and customers who simply stop buying. For repeat buyers, overdue is measured against the usual gap for what they buy, so a clothing customer and a furniture customer are not held to the same clock.
What needs attention
Upcoming renewals at risk, champions who left, customers who went quiet or are slipping, service problems, and where the risk sits and how it moved.
Every score with its reasons
Each customer’s risk comes with the signals pushing it up or down, grouped by where they come from (experience, operations, financial and relationship), and a likely range that is wider where there is less history to go on.
Breakdowns
Risk by segment, by renewal or repurchase, and by acquisition cohort.
What-if readings
See how a customer’s risk would read if a signal changed, clearly labelled as a reading rather than a forecast of what an action will do.
Trained on your history
Several models are trained on your own past outcomes and graded on the most recent period, which none of them saw. The best one scores your customers, the full leaderboard stays visible, and an admin can pin a different model with a reason recorded in the audit log.
Check the predictions
How the model did against what actually happened last cycle, compared with a simple rule anyone could run. If it is not clearly better, its scores are withheld unless an admin chooses to publish them.
Signals from across Hub
Survey answers and Service Experience conversations feed the model live, and the accuracy view shows how much experience data adds on your own outcomes. Your data is used for your organisation only and is never pooled.
Settings
How your customers buy, what counts as leaving, how models are selected, alerts, access, and importing your history.
Churn Modeling predicts who is likely to leave. It shows associations, not causes, and it does not say what will make a customer stay.
Works with
- Account Health Score: one health score per B2B account, in your CRM.
- Customer Lifetime Value: predict each customer’s future value and act on it.
- Closed-Loop: turn unhappy responses into owned, tracked cases.
- Predictive Satisfaction Score: estimate satisfaction for every customer, not only respondents.