- Industry
- Banking
- Department
- Customer
- Team
- Customer Lifecycle
- Kind
- Investigator
- Approver
- Head of Retail Banking
- Code
- CUS-14
It flags customers likely to move their main account, savings or mortgage elsewhere within the horizon, weighing salary moving out, balance transfers to other banks, mortgage fixed rates ending, rate-comparison complaints and closed direct debits. It explains each risk and proposes a save play the conduct rules allow. **The problem it solves:** Why each customer may move their banking: the salary, balance, mortgage, complaint and payment signals, when they began, and what changed before them.
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 risk data warehouse: salary and balance flows, outgoing transfers to other banks, mortgage maturities and complaints for the customers..
- 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 Head of Retail Banking with its reasoning before anything goes out.
- 06Acts
Once approved, it acts in the CRM: it records the attrition flags and queues the save plays for the relationship team. and records what was done.
What it works with
Your knowledge, your systems, your people.
Knows
The attrition model: switching signals, mortgage maturity and rate-review rules, the requirement that retention offers are fair to existing customers, the account-switching service rules, and contact limits.
Reads from
the risk data warehouse: salary and balance flows, outgoing transfers to other banks, mortgage maturities and complaints for the customers.
Acts in
the CRM: it records the attrition flags and queues the save plays for the relationship team.
Hands back
A list with one entry per customer, each with customer_id, risk_percent, products_at_risk, signals (a list), likely_cause, save_play (or null) and confidence.
Asks a person when
A flagged relationship is at or above 25000 USD and switching within 30 days, a save play offers a rate not on the product shelf, or the customer shows financial difficulty.
Settings
Your limits, set by your managers.
| Setting | Default | What it controls |
|---|---|---|
| Risk threshold (%) | 60 | At or above this a customer is flagged. |
| Horizon (days) | 90 | The window predicted. |
| Mortgage fixed-rate end to watch (days) | 120 | Customers whose fixed rate ends within this are watched. |
| Balances and lending for a save play (USD) | 25000 | Below this no personal save play. |
| Exclude new customers (days) | 60 | Customers newer than this are not scored. |
Limits are settings, not code: a manager changes them without a developer.
Rules it always follows
- Flag at or above 60 percent within 90 days.
- Contact customers whose fixed mortgage rate ends within 120 days with their product-transfer options.
- A save play goes only to relationships at or above 25000 USD, and a retention rate is offered only if it is available to comparable existing customers.
- Never use protected characteristics; explain each flag by its signals.
- Exclude customers newer than 60 days.
See the Churn 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.