Matching and confidence

Our service is based on data: we have some of the brightest data scientists and data engineers around and they are focussed on curating and matching data to allow our clients the best insight into the risks they manage.

Data is drawn from public sources – but not all our sources use the Company Registration Number as a unique identifier. In the absence of unique identifiers, we have invested significant time building in-house matching algorithms to ensure we can link data sets.

Data that doesn’t pass the test

There are some data sets that we don’t use – for example Payment Data. Our research has shown that, because of data quality issues (it relies on multiple third parties keeping their ledgers up to date) and the fact that it is an incomplete set (i.e. absence of information cannot allow us to draw a conclusion) the value in risk scoring is limited compared to other more robust sources.