“Big data is here, but some insurers are still struggling with small data... They have Stone Age systems in place, where the structure of the data is poor and access is difficult. It's not the best starting point. Those that manage to get the core data storage correct will find it easier to build in more data.”
This is a quote from Michael Graham at Sequel. “Stone Age” may be a little harsh... Those systems more correctly belong to the (big) Iron Age. Graham's point is well made. How are we to cope with an onslaught of unstructured data if our existing data sets already suffer from poor structuring? There needs to be a culture of standardization ahead of any success with big data. Insurance is a business that's hungry for indicators. The industry is founded on the isolation of data items that inform about risk. It grows by finding new indicators. It is also forced to adapt when existing indicators become obsolete or disallowed. So, for example, in Europe gender is becoming a disallowed indicator in car insurance. But a slew of indicators connected to driving ability are becoming central to the same product line.
So, insurance is, among other things, a data definition business. Regularizing its data stores is not an externally driven concern. It's at the heart of what we do. Always has been, always will be. Insurance already has a culture of standardization – we just need to include technology in this perspective. Insurance Insight
Comments