Visit rhythm and reliability: spot customers who stop coming in time

Piyzi answers two separate questions about every customer automatically: are they still coming? (visit rhythm) and do they cause problems when they come? (reliability). You don't enter anything; statuses update by themselves with every visit, booking and payment.

What counts as a visit?

A completed or unmarked past appointment, class attendance, group event attendance, a turnstile entry and a facility entry count as visits. Several entries on the same day count as one visit. Cancelled and no-show appointments are not visits.

How the visit rhythm is worked out

• The rhythm is the median number of days between the customer's consecutive visits over the last 12 months (the latest 8 gaps). The median, not the average, is used so one long break does not distort it.

• A personal rhythm needs at least 3 visits. For customers with fewer visits, Piyzi looks in turn at the spacing of their future bookings, the pace of their active package and your business's overall rhythm. The card says which one was used.

• Your holidays and the customer's subscription freeze days do not count as delay.

• The thresholds are the same for everyone; there is no per-business setting.

Visit statuses

• No visits: has not visited yet.

• New: only a few visits so far; a personal rhythm needs at least 3 visits.

• Active: visiting at their own rhythm.

• Scheduled: has a future appointment, a class or event booking or a facility reservation — they are expected, so they are not counted as late.

• Late: hasn't visited for longer than 1.5 times their rhythm (at least 7 days overdue).

• At risk: hasn't visited for longer than 2.5 times their rhythm.

• Lost: hasn't visited for longer than 4 times their rhythm or 180 days.

• Returned: came back after being At risk or Lost; shown for 30 days after the return.

Reliability score

Reliability is a 0–100 score; a HIGH score means a reliable customer. It looks at bookings the customer didn't show up for and cancellations they made at the last minute (inside your cancellation window) over the last 180 days, together with past-due unpaid debt and how late the oldest debt is. It is smoothed with a few trouble-free bookings, so on thin data a single no-show does not make a customer "Unreliable" at once. Reviews do not affect the score.

• Reliable (75–100): shows up for bookings and pays on time.

• Caution (50–74): occasional no-shows, late cancellations or late payments.

• Unreliable (25–49): frequent no-shows, late cancellations or past-due debt.

• Very unreliable (0–24): a clear no-show or debt problem; consider asking for prepayment.

• Not enough data: a score needs at least 3 bookings in the last 180 days; fewer are enough if the customer has past-due or late-paid debt. No label is given on thin data.

The Reliability card on the customer detail shows the reasons that affect the score most (e.g. "Didn't show up for 2 of 8 bookings") and how much each one lowers it.

Where you see them

• Customer list: the Visit and Reliability columns, the Visit status and Reliability choices in the Filters panel, and rhythm and reliability sorting.

• Customer detail → Overview: the Visit rhythm card (last visit, rhythm, expected visit, threshold dates, sources of recent visits) and the Reliability card.

• Operations dashboard: the "Off rhythm" tab.

• Reports: the customer metrics on Overview (on rhythm / off rhythm / lost) and the filters, columns and sorting in the Customer Report.

• SMS campaigns: the "Visit Status (own rhythm)" field in segment rules (e.g. a win-back campaign for At risk and Lost customers).

• Automations: the "When a Customer Falls Off Rhythm", "When a Customer Is at Risk" and "When a Customer Returns" triggers; the Win-back scenarios in Ready-made Automations.

• Online booking rules: with the Reliability Score rule you can ask low-scoring customers for approval or a deposit.