In our new video, Craig Evans (CEO), Chris Oatts (Head of Data & Product Strategy) and Jonathan Davies (Lead Data Scientist) sit down to talk about the thinking, the challenges and the late nights behind PoD®, the new risk score from Company Watch, which launches on 21st October.
It started with a question
About two years ago, Craig and Chris started asking what all our new data could do. We had been bringing in more UK and Irish data than ever, including our own proprietary sources. Machine learning had also moved on a long way since our last score was built.
The market had changed too. Customers no longer want only lagging, factual information. They want leading indicators: early signals that help them act before they take a loss. As Craig puts it, what Company Watch does is help businesses avoid financial loss. And everything we build is geared towards achieving that in the most efficient way possible.
Inside the build: how PoD® came to life
Building a new risk score is rarely simple, especially when the one you already have is trusted across the market. For PoD®, it meant rethinking how Company Watch uses its data, from the models behind the score to how the results reach customers.
In the video, our leadership and data science team go behind the scenes of that process. They explain why we moved away from a traditional 0–100 score, and how we kept a machine learning model transparent enough for compliance, credit and procurement teams to rely on. They also share the design decisions that let PoD® serve both quick, high-level checks and detailed portfolio analysis.
They also cover what customers told us during two beta programmes, how PoD® has performed over 12 months of tracking, and why we expect its accuracy to keep improving as new data comes in.
For organisations managing credit exposure, supplier risk or trade credit insurance, this conversation shows where business risk assessment is heading, and what will be available when PoD® is launched. Book a demo today to see how PoD® can sharpen your risk assessment.
Risk isn’t measured anymore. It’s predicted.
