By FINTECH Books Contributor, Schicho Markus
Insurance companies now experience an urgent need for action: During these past few years, the enormous digitisation process happening in other sectors has largely been ignored – only very little investments were made in new technologies, although they offer captivating possibilities in automated information processing.
One niche that is attractive for new technologies is the field of personal claims
management. Studies show that every tenth insurance claim (i.e. resulting from
car accidents, sports injuries, etc.) involves excessive claims and / or fraud.
This results in an annual loss of eight to twelve billion euro in Europe alone.
Although there are adequate insurance risk models, the core problem is the huge amount of data to be handled manually. On average, a single record contains more than 30 documents, with correspondences, medical protocols, etc. When the insurance officer has to evaluate the claim, all this information has to be read, conspicuous data has to be identified, and then further steps for investigations have to taken.
For a medium-sized company, this results in about 3 million papers to be read over a year. Furthermore, there is no guideline on recovery time or health impact of certain injuries – thus preventing monitoring the progress, and in the end the legality of the claim. This is very much a Big Data to Smart Data challenge.
Technically, smart text analytics technologies are essential to support the insurance officers in their tedious effort to track the documents created, extract the health relevant information and assess a recovery prognosis for every injury automatically. This reduces the workload on every claim massively. For every incident, insurance companies have to form reserves ranging from 100k to millions of euro. An increased success rate of conspicuous records helps the reserve to be resolved earlier by a general compensation. Eliminating illegitimate claims leads to phenomenal savings. This, in the end, would benefit us all.