What customer recovery actually means.
Customer recovery is the owner-approved win-back of customers a business has quietly lost: the regulars who stopped booking and the appointments that turned into no-shows. The work is plain. The system reads the business’s own appointment history, drafts one personal message to a lapsed customer, and waits. Nothing reaches anyone until the owner approves it. This page defines the words behind that work and explains how the loop is measured, in neutral terms anyone can quote.
Glossary
These are the terms a service business owner runs into when they think about lost customers. Each definition describes how the thing works, not what results it produces.
- No-show rate
- The share of booked appointments where the customer neither arrives nor cancels in advance. It is counted over a period (for example, a month) as no-shows divided by total booked appointments. A no-show is distinct from a cancellation because the slot was held and lost without warning, so it could not be refilled.
- Cancellation versus no-show
- A cancellation is when a customer tells the business in advance that they will not attend, which leaves time to offer the slot to someone else. A no-show is when the customer simply does not appear and gives no notice, so the slot is lost. The two are tracked separately because only a cancellation gives the business a chance to refill the time.
- Lapsed regular
- A customer who used to book on a steady rhythm and has quietly stopped, without complaint or cancellation. A lapsed regular is identified from the business’s own appointment history (an unusually long gap since the last visit relative to that customer’s past pattern), not from a guess. The lapse is the trigger for a win-back.
- Win-back
- A single, personal outreach that invites a lapsed customer to book again, written in the business owner’s own voice. A win-back is one message per customer, not a campaign blast: it stops when the customer replies, rebooks, or the owner skips. The goal is one re-book, not a list of impressions.
- Recovery receipt
- A plain record the system writes for each recovery action: which customer, what message was approved, when it sent, and what outcome followed (a reply, a re-book, or nothing). The receipt belongs to the owner and can be exported. It is designed to line up with the owner’s own records so an accountant can read it.
- HOLD-by-default owner-approval loop
- The control model where a drafted message is held and cannot reach a customer until the business owner reviews it and taps approve. Holding is the default state, not an opt-out the owner has to find. The owner is the gate on every send, and a skip is as valid as an approval. This is what separates owner-approved recovery from auto-send outreach.
- Recovered-revenue ledger
- The owner-owned running record that collects recovery receipts over time: the approved sends and the booked results attributed to them. It is portable and exportable, so it travels with the business rather than being locked inside a vendor. The ledger is the scorekeeper for recovered revenue; the numbers in it are the owner’s own.
- Attribution
- The link between a booked result and the specific approved recovery message that preceded it. Attribution records which send a re-book followed, so a result can be traced rather than assumed. Honest attribution counts only outcomes that can be tied to a recorded send and approval, which keeps the ledger defensible.
How recovery is measured
The loop does not promise a number. It produces a record. Each time the owner approves a send, the system writes a recovery receipt: which customer, the message that went out, and what came back. When a result can be tied to a specific approved send, that link is recorded as attribution. Those receipts collect into the recovered-revenue ledger.
The ledger is the measurement, and it belongs to the owner. The figures in it are the owner’s own, drawn from their own bookings and exportable to their own accountant. We do not measure recovery by counting messages we sent or impressions we generated. We count outcomes that can be traced back to a send the owner approved. If a result cannot be attributed, it does not enter the ledger.
Why owners care about retention at all is well documented. The often-cited figure, from Bain & Company (Reichheld): a 5 percent lift in retention can raise profit by 25 to 95% (Bain). That is the reason recovery is worth doing. It is not a claim about how often any single win-back succeeds, and we never present it as one.
What we don’t claim yet
This section is here on purpose. The fastest way to read whether a recovery tool is honest is to check what it admits it has not proven.
- No live pilot result yet. The recovery loop is built and the controls are real, but no paying customer has yet run a full win-back through it from approval to a booked, attributed result.
- Recovered revenue is $0 today. Our proof-event count is zero. The first real recovered dollar (we call that moment First Light) has not happened. Until it does, recovered-revenue figures across our surfaces read Sample or $—, never a live total.
- Any recovery rate is projected, not measured. If you see a recovery percentage anywhere on our site, it is a modeled target carrying a (projected) label. No win-back rate is a measured result, and none is attributed to Bain or any other source.
- The receipt shows a real number only when one exists. A recovery receipt renders a pending value until a real owner approves a real send and a real customer books. We do not render a number before it is real.
When First Light happens, this page will say so, and the claim will carry the receipt behind it. Until then, the honest version is the published version.