Guest Lapse Prediction Agent
Flags loyalty members and corporate accounts at risk of booking elsewhere.
- Industry
- Hotel
- Department
- Customer
- Team
- Customer Lifecycle
- Kind
- Investigator
- Approver
- Guest Relations Manager
- Code
- CUS-14
It flags loyalty members and corporate accounts likely to stop staying with the group, from the gap since their last stay against their usual pattern, falling booking pace, recent complaints and poor reviews, tier downgrades and bookings moving to online travel agencies. It explains the top reasons for each with the evidence and logs the at-risk flag and a save step in the CRM. **The problem it solves:** Why each loyalty member or corporate account may stop staying with the group: the change in their stay pattern, booking pace and channel, and the events that could explain it.
How it works
6 steps, and a person signs off before step 6.
- 01Looks it up
Finds the record the request is about in the guest messaging and CRM platform: the guests' and accounts' stay history, loyalty tier changes, booking channel mix, complaints, review scores and marketing consent..
- 02Reads the request
Works out what is being asked and what a complete answer needs.
- 03Checks the facts
Checks figures, dates and rules against the source records, not from memory.
- 04Works it through
Weighs competing explanations against the evidence and shows which one holds up.
- 05Asks a person
Hands the work to Guest Relations Manager with its reasoning before anything goes out.
- 06Acts
Once approved, it acts in the CRM: it flags each at-risk guest or account and creates a save task for its owner. and records what was done.
What it works with
Your knowledge, your systems, your people.
Knows
The lapse model: the signals and their meaning, each segment's normal stay frequency, the save plays by segment and what each play may offer.
Reads from
the guest messaging and CRM platform: the guests' and accounts' stay history, loyalty tier changes, booking channel mix, complaints, review scores and marketing consent.
Acts in
the CRM: it flags each at-risk guest or account and creates a save task for its owner.
Hands back
A list with one entry per guest or account, each with id, kind, likelihood_percent, flagged (true or false), reasons (a list of up to three, each with reason, evidence and date), how_to_confirm and save_play (text or null).
Asks a person when
A save play includes an offer, points or a rate concession, or a corporate account worth more than 3000 USD is flagged.
Settings
Your limits, set by your managers.
| Setting | Default | What it controls |
|---|---|---|
| Flag at likelihood (%) | 60 | Guests or accounts at or above this lapse likelihood are flagged. |
| Look ahead (days) | 180 | Lapse is predicted over this horizon; hotel stays are less frequent than most purchases. |
| Overdue when the gap exceeds the usual by (times) | 1.5 | A member whose days since last stay exceed their usual gap by this multiple is overdue. |
| Annual spend that earns a save play (USD) | 3000 | Only guests or accounts spending at least this a year get a save play proposed. |
| Skip guests with fewer stays than | 2 | Guests with fewer stays have no pattern to measure and are not scored. |
Limits are settings, not code: a manager changes them without a developer.
Rules it always follows
- Score only guests and accounts with at least 2 stays, over 180 days unless the request sets another horizon.
- Treat a member as overdue when the days since their last stay exceed their usual gap by 1.5 times, and weigh it with booking pace, OTA share and complaints.
- Flag at or above 60 percent likelihood, and give the top three reasons with the evidence and date for each.
- Prefer a dated cause, such as a service failure, tier downgrade or competitor opening, over a general trend, and say what would confirm it.
- Propose a save play only for guests or accounts spending 3000 USD or more a year; any offer in it needs the Guest Relations Manager.
- Never contact a member without marketing consent; a save step for one goes through their account owner or the front desk on their next stay.
See the Guest Lapse Prediction Agent agent on your own work.
We’ll run it on a sample of your work and show you what your approver would see.