How the humble database can help prevent insurance fraud
Fraud is a challenge businesses struggle to find remedy for. The Office for National Statistics (ONS) recently stated that last year saw nearly six million fraud and cyber crimes take place in 2015, in the UK. What is more, these crimes impact everyone in some way. Take insurance fraud for instance; personal premiums may rise as insurance providers seek to recover the money lost. Ultimately, this is a problem which needs tackling from the ground up.
In 2014, insurers found 150,000 fraudulent claims which cost over £1.3 billion. This has led to companies, who have a vested interest in defeating fraud, to invest £200 million to identify and prevent fraud. That would be £200 million well spent if it indeed reduced the risk but it’s hard to tell if it is.
Investing in digital technologies is key to securing the future of the insurance industry. Digitisation has rapidly transformed the landscape for consumers in how they can look for, and acquire, policies. Customers are no longer going to their high street brokers. The impact of less footfall was illustrated by Swinton who plans to close 130 of its high street branches within the next 12 months. High street brokers are being replaced by online price comparison and brokerage websites that can give you the best deals, for your particular circumstances, in real-time.
This sheer number of information requests and transactions online, coupled with the needs to satisfy customer demands in real-time, has significantly increased the risk of online fraud. Unfortunately, the means to detect these risks are still relatively archaic. KPMG found that most companies rely on old-fashioned tips rather than technology to overcome fraud. The question is, why is technology, which can help monitor behavioural patterns, overlooked when trying to combat this?
Something which has intrigued all data driven businesses is the thought of data analytics on an industrial scale. Data analytics is touted as a viable means for brokers to detect fraud by examining the swathes of data insurers have.
The de facto post-incident response in insurance is erroneous, relying on post-breach investigation and hours of legwork by “field agents”. Detecting it whilst it is happening, or better still, predicting incidents based on behavioural analytics, is a little more enticing and it is possible. A real-time look at validation data is the holy grail of detecting fraud, not just on new policies but renewals as well. To do this, a robust database infrastructure is needed to handle the queries of not only prospective clients but brokers too.
Backend technologies which make up the underlying infrastructure are able to provide this sort of capability. By building specialist applications and using them to aggregate data of individuals trying to take out policies, especially analysing past and current customer behaviour, would make it possible to detect and root out those who are trying to defraud the organisation, before they strike.
To do this with efficiency, it is important to have the database, which serves and receives the data, functioning in good order. One which is robust and scalable, capable of handling billions of queries coming from prospective customers on a price comparison website and the brokers who need to confirm or deny application success. Management of the database is also critical; if it isn’t properly monitored and implemented with a view of handling large quantities of data, downtime and revenue loss, due to unidentified fraud, can almost be guaranteed.
Technology might not be the only solution, it might take a combination of tactics to help reduce the risk from the 150,000 fraudulent claims which were uncovered within the last year, with probably several thousands more still hidden. The astronomical growth of data, and the ability to analyse it, will definitely play a part in the future success of the insurance industry.
The smart bet for insurers to move forward is by having a strong database foundation, one that can handle transactions at scale and continue fighting hard against the new world of cyber fraudsters.
- » Calculating the Kube roots: Why 2019’s KubeCon represented a milestone for the industry
- » Enterprises not seeing total fulfilment with cloud strategies – but hybrid remains the way to go
- » How to get your data scientist career up and running: A guide
- » How cloud computing is changing the laboratory ecosystem
- » How to improve supply chains with machine learning: 10 proven ways