Context
A lender noticed that repayment obligations bunch up in the last ten days of the month. Collections teams work in cycles, so a spike matters to operations as well as to cash.
Problem
Could the peak be reduced without disturbing repayment dates that customers choose for good reasons, such as salary day?
The seven lenses, applied
| Lens | What this example showed |
|---|---|
| Surface | Repayments pile up late in the month, and collections activity is several times higher in those days. |
| Structure | Products set repayment dates in different ways: by the customer’s salary date, by product structure, or by business cash flow. |
| Relationships | Agents tend to originate loans late in the month. A 30-day structure then pulls the due date into the same window. |
| Information | Loan-level origination date, product rule, repayment date and collections activity by day. |
| Dynamics | The pattern repeats every month, and collections capacity runs in cycles too. |
| Emergence | The peak comes from salary dates, product design and sales behaviour together, not from any single rule. |
| Intervention | A controlled pilot on new flexible loans only, with existing schedules untouched. |
System and insight
Product rulesCustomer salary datesSales and origination timingRepayment scheduleCollections capacityCredit and impairment cycle
Part of the peak is fixed by customers’ pay dates and should be protected. Part comes from product structure and from when loans are originated. Only that second part is a choice.
Trade-offs
- Moving repayment dates helps the calendar but could hurt customers whose income arrives on fixed days, so salary-linked products stay as anchors.
- A pilot on new loans only is slower than restructuring the book, but changes nothing for existing customers and produces evidence first.
- Whether concentration drives collections strain or losses is the first thing the pilot is built to test.
Intervention
- Separated the fixed anchor (salary-linked repayments) from flexible exposure (products with structure-set dates).
- Recommended a controlled pilot on new flexible loans: where the standard first-payment date lands in the hotspot, compare approved alternative offsets the customer can support and pick the one that concentrates least.
- Left existing repayment schedules unchanged, and proposed starting with underwriting plus a spreadsheet before any automation.
My role
I prepared the analysis and the recommendation for the management committee.
Result
- A recommendation to approve a controlled pilot on new loans, presented in September 2026.
Learning
Anchor what is fixed. Balance what is flexible.