A telecom sells phones on instalments to customers with little credit history. Here's how AI credit decisioning made that profitable.
The problem, simply
Telecoms like Ooredoo Qatar sell phones on a buy-now-pay-later (BNPL) basis — the customer takes the device today and pays in instalments. The catch: many buyers are thin-file (little or no formal credit history), so it's hard to tell who will keep paying. Ooredoo was losing about 12% of these device-financing plans to default — high enough to eat the margin on every phone sold.
What CXingularity did
- Real-time decisioning — scored each applicant in seconds at the point of sale, using alternative and behavioural data rather than a traditional credit score alone.
- Approve more, safely — auto-approved low-risk buyers instantly while flagging higher-risk applications for a closer look, so good customers weren't turned away.
- Keep watching — monitored repayment behaviour after the sale, surfacing early signs of trouble so the team could act before an account went bad.
The result
A roughly 74% reduction in defaults — turning device financing from a loss-leader into a profitable, scalable product, without slowing down the checkout.
The same engine that decides a business loan in minutes decides a consumer device plan in seconds — the difference is just the data it reads and the speed it runs at.
- BNPL
- Buy Now, Pay Later — take the product now and pay it off in instalments.
- Device financing
- Selling a phone or gadget on instalments instead of one upfront payment.
- Thin-file
- A customer with little or no formal credit history, so a normal credit score says very little.
- Default rate
- The share of those instalment plans that stop being repaid — lower is better.
- Credit decisioning
- The automated yes/no (and on what terms) at the moment someone applies.