Fraud Detection

Whilst the majority of Celebrus clients use our data and real-time personalisation software to achieve marketing objectives around loyalty, revenue and costs, some have found invaluable and ingenious uses for us as another weapon in their fraud detection armoury. 

The individual-level nature of Celebrus’ multi-channel behavioural data gives a unique view of each visitor including their environment (IP address, device, browser etc.) and behaviour (from navigation and clicks right down to mouse moves and keystrokes) over time means that Celebrus can give a complete picture of that individual’s behaviour across your online channels. 

A few examples of this highly granular data helps identify potential fraud include:-

• If a retail website visitor tries inputting several credit card numbers before completing the order, a well-crafted customer friendly call from the call centre to confirm the order before it’s shipped could reduce exposure to online fraud.  We can do this because we see everything that someone types into a form regardless of whether they hit submit or not.
• If a car insurance company receives multiple visits to their online pricing calculator with several of the key variables entered changing each time, e.g. where the car is kept, driver’s age etc., this could highlight that the individual might be being a little flexible with the truth in order to secure a good deal.  We can do this because we can see the behaviour by individuals across time so piece together their online history on your website.
• If an insurance company receives multiple claims on different policies from the same IP address within a certain period of time this could be an indicator of insurance fraud which requires further investigation.  We can do this because we collect the IP address from each visitor.

Other fraud detection ideas crop up consistently in different types of Celebrus clients - once you have the data your analysts and fraud team will love playing with it to identify potential fraud.