Big Data and Data Warehouses
In the early, to mid-2010s it was all about Big Data. If you didn’t have an own Data Warehouse or at least a project, you were not going in the right direction. The effective outcome of these projects, especially in the more traditional sectors such as insurance, was often not that high. So the “trend” got “improved.” Data Warehouses were replaced by Data Lakes, in which all data we could get our hands on in the company could be “dumped” for further analyses. The results of these analyses were often quite a disappointment. It exposed some major flaws in how we capture and store data. Very little data was available and accessible in structured formats.
Never miss InTech insider information, top jobs and mandates. Sign-up here.
If I look back at the past 15 years in which I have worked within the claim process sector, I see progress, but also a lot that reminds me of the days when I started out (especially the lack of structured data management).
Collecting data
At Spearhead, we take collecting claims data in a structured way seriously. For the first time, the circumstances and actual damages or losses are described in such a way that they can be used in smart analytical models. And this can all be done within the GDPR framework. Doing the groundwork for digitalization, capturing data in your processes in a structured manner, and qualifying them well is the basis for most digitalization initiatives. It is less flashy than a Big Data project, but it is the foundation of your company’s performance in the future. The value lies in the data that is true, not because you need to sell it like some of the well-known internet and social media companies do, turning data into a business model of its own. But we should consider this option because it drives the improvement of your business processes, your services and allows you to get to know your customer a lot better.
It also helps you to take advantage of new technology innovations a lot more efficiently. We see this with the Internet of Things, in our case vehicle telematics. It’s a great source of new and objective data. But in order to make it actionable, in order to be able to turn the data into value, you need to be able to connect it with data from your existing business. In the end, this is less a matter of technology and more one of proper data management and data design.
Machine Learning / Artificial Intelligence
Now we are in the age of Machine Learning (ML) – Artificial Intelligence (AI). The buzzwords of today. And as with the Data Warehouses and Data Lakes, the technologies of ML/AI have huge potential. But we can only use them to their full potential when we collect the right kind of data, in a structured format.
By Cees Van Dijk.
