What is the money finder?
It is the surface that answers who stopped buying. For each customer it computes their own purchase rhythm from their history, their value to the business, and whether they are past the point where they would normally have bought again. Customers who are late against their own pattern are lapsed; customers who buy twice a year are not lapsed in month three.
Alongside it sits a customer health view: new, returning, at risk and lapsed, with the value in each state, so an owner sees whether the base is growing or quietly eroding under a good revenue month.
- Each customer's own purchase interval, computed from their own order history.
- Their value to the business, so the list can be ranked by money rather than by count.
- How far past their usual interval they are, which is what makes a win-back timely.
- The customer health view: new, returning, at risk and lapsed, with the value in each state.
- The exclusions applied before review: open tickets, recent returns, unsubscribes and active flows.
Why does per-customer rhythm beat a global window?
A single ninety-day rule treats a weekly buyer and an annual buyer the same way, and it produces a list full of people who are not lapsed at all. Computing each customer's own interval from their own orders makes the list smaller, more accurate and much more valuable per message sent.
It also makes the timing right. A customer who is two weeks past their usual interval is a better target than one who is six months past it, and the list is ordered so the recoverable ones come first.
What vetted means
Before a list reaches review, the engine removes the people who should not be on it: customers with an open support ticket, customers who returned their last order, customers who have unsubscribed, and customers already in an active flow in the email tool where that is visible. The exclusions are listed so the reviewer knows what was removed and why.
That is the difference between a list and a blast. A message that arrives while a refund is unresolved does damage that a discount cannot repair, and the vetting exists to prevent exactly that.
The human review before the push
The list appears in Review & Send with its size, its total value, the exclusions applied, and the message that will be sent. A person edits, removes rows, changes the copy, or approves. Only then is the list pushed to the connected email marketing tool, which sends it.
The engine does not send marketing email itself. It prepares the audience and the message and hands both to the tool the brand already uses, with the approval recorded.
How is the list's recovery measured?
After the send, the Results ledger measures what happened: which recipients bought, what they spent, and how that compares to the period before. The verdict is per send, and it rolls up into the value ledger so an owner can see the cumulative recovery rather than a series of one-off claims.
A send that recovered little is reported as such. That honesty is what lets the next list be built differently rather than repeated hopefully.
Which roles own this on ftes.ai?
The Customer Intelligence FTE owns who lapsed and what they are worth; the Lifecycle Officer FTE owns the win-back and the flows around it, and is judged on recovered revenue from lapsed customers. The Voice of Customer FTE reads the reasons behind churn where tickets and returns explain it.
A brand with the store and an email tool connected can produce its first vetted list on day one. The useful discipline is to send one list, read the Results verdict, and let the second list be shaped by it.
What does a good first win-back look like?
Small and specific. One segment, usually the customers who are least far past their own interval and worth the most, with a message that names something real about the brand rather than leading with a discount. The engine ranks the list so that segment is at the top.
Then read the Results verdict before sending a second. A brand that sends one list, measures it and adjusts is building evidence; a brand that sends four in a month is training its customers to wait for the next message.