Cadence detection
Every regular has a visit cadence — fortnightly fade, monthly cut, six-weekly set. The system learns it. When the cadence breaks, the regular shows up on the retention queue — before they're gone.
Per-regularLearns from booking historyCatches drift, not just churn
Pattern per regular
The model is per-regular, not a one-size-fits-all rule. A monthly cut and a fortnight fade have different cadences; both get tracked accurately from booking history.
Risk alerts before they're gone
A regular's cadence stretches. Then breaks. You see the alert with the number of days past their usual cadence, what they usually book, and when they last came in. A clean handoff to the retention play.
See the cadence model on real data
20-minute demo, with sample regulars showing a clean fortnight, monthly, and six-weekly cadence.