One of the defining characteristics of the digital lending market in Nigeria is the low level of individual access to smartphones and digital devices. A single smartphone may be shared among several family members or even households, making borrower identification difficult and increasing the risk of repeat loan applications from the same device.
Market Specifics: Limited Access to Digital Devices
CashExpress (operating under the CashX brand in Nigeria) applies advanced fraud prevention rules based on JuicyScore’s device fingerprinting technology to detect such repeated cases and reduce risks associated with multiple users per device. Specifically, the following parameters are used:
- Unique technical parameters of the device – help detect attempts to apply for multiple loans from a single smartphone.
- Behavioral indicators of borrower activity, such as the number of credit applications in the last 1, 7, and 30 days – a key signal of potential fraud or over-indebtedness.
- NPL 90 Plus – reflects the probability of loan default beyond 90 days, based on client behavior and device profile.
Using these parameters, CashExpress filters out up to 8% of potentially fraudulent borrowers. Data analysis during periods of technical disruption (when JuicyScore data was unavailable) confirmed that disabling these fraud-prevention mechanisms led to a 3–5 percentage point decrease in the default rate.
As noted by Temitope Adetunji, CEO of CashExpress Nigeria, the challenges of digital lending in emerging markets require a fundamentally new approach to credit risk assessment:
“When one smartphone is used by ten people, traditional scoring is powerless. JuicyScore became our eyes and ears in this digital chaos. The product helps us see behavioral and technical signals that simply can’t be obtained by other means. As a result, we’re able to assess risk more accurately — even with minimal traditional data.”
Improving Scoring for Clients Without Credit History
In a market where a significant portion of the population lacks a bureau credit score, alternative credit scoring models play a critical role. CashExpress operates 5–6 scoring models, each tailored to different customer segments. These models show particularly high accuracy for credit-invisible clients, with up to 50% of variables based on JuicyScore’s behavioral and technical data.
By incorporating alternative data and device analytics, the Gini coefficient of scoring models improves by +3 to +10 points, depending on the client type. Key metrics include:
- RAM productivity / storage productivity – measure how efficiently the device operates, allowing segmentation of users by digital maturity and likelihood of genuine behavior.
- Cursor movement speed / cursor distance covered – show how naturally and actively the user interacts with the interface, helping to detect bots or low-engagement patterns.
- Browser version ageing – indicates how outdated the browser version is, often correlated with digital literacy.
- IDX4 connection markers / IDX6 internet infrastructure quality / IDX7 device applications quality – composite indices reflecting connection stability, infrastructure quality, and app characteristics.
- IP ageing in months / IP ZIP code – show IP address usage history and client location.
- Loan limit utilization / observed quarters – how the client utilizes credit limits and how long they’ve been tracked in the system.
- Scroll movement speed / single click / time on page – additional behavioral signals indicating client attention and intent.
According to Temitope Adetunji, the company’s success in navigating the complexities of fraud detection and credit invisibility wouldn’t have been possible without JuicyScore:
“In Nigeria, you simply can’t rely on classic data sources alone. JuicyScore helps us see risks where others see nothing. The product allows us to distinguish critical signals — especially when the client has no credit history. As a result, we’re seeing real reductions in defaults and feel more confident when working with new-to-credit borrowers.”
Results: Reduced Defaults and Improved Scoring Accuracy
JuicyScore integration enabled CashExpress to:
- Decrease the default rate by 3–5 percentage points, proving the practical value of device intelligence in fraud prevention;
- Improve scoring accuracy through behavioral and technical data;
- Expand lending to clients without traditional credit histories.
The company continues to scale the use of JuicyScore’s alternative credit scoring tools, evolving its risk assessment approach to better fit the realities of digital lending in Nigeria and other emerging markets.