A payment score that lifted on-time payments 20–30%

I was a Product Analyst (later Senior Product Analyst) at Smart Energy Water for about two years, working on the payment and customer analytics modules of the product.

Payment analytics was missing from the user portal — the business had no systematic way to identify which users were paying on time versus defaulting, so there was no data-backed way to intervene with at-risk customers.

I owned the design of the payment score dashboard specifically — what it needed to do functionally and how it should behave — along with the underlying product analysis behind it.

Using the payment score, we targeted customers with a higher probability of defaulting and offered them an appropriate payment plan.

Making the score visible to customers drove behavior change directly — people cared about their own score and worked to improve it, which supported the push toward more on-time payments and the marketing of payment plans.

I also built customer journey analytics in Power BI, extending the product's analytics capability more broadly.

I scoped the payment score to retail clients only — I deliberately didn't build it for corporate clients. Corporate accounts were a much smaller population, and a reliable score needs enough training data per client; there wasn't enough volume on the corporate side to make that work, so I focused entirely where the model could actually be accurate.

The payment score work drove a 20–30% increase in on-time payments.