“All right, but apart from the sanitation, medicine, education, wine, public order, irrigation, roads, the fresh water system and public health, what have the Romans ever done for us?”
Reg, Life of Brian (1979)
Claim analytics: one of the current crop of buzz phrases enlivening the insurance world. Does it fulfil its promise? Is it worth the effort? What have data analytics ever done for us? Well, quite a bit actually, but they still haven’t given us anywhere near as much as they will. We have only just begun to mine the Claim Analytics seam.
I’m using the term ‘analytics’ here very broadly, referring to any sort of analysis we do on the data we capture and generate, including simple reports, trend analyzes and flashy, multi-coloured dashboards. Using technology to report on the progress of the insurance operation has been developing since it first appeared. Management reporting gave birth to Management Intelligence, (MI), which morphed into Business Intelligence, (BI), which in turn evolved into analytics.
What happened in the past?
The ‘reporting’ aspect tends to throw light on the past, showing us information and patterns in large quantities of data over time that reveal trends and associations that may not have been discernible as they happened. These reports, favoured by higher levels of management and executives, show us financial performance, volumes and percentages that illuminate the progress of the business over time frames such as months, quarters and years. These have been the staple of reporting produced by IT since IT was first used.
What’s happening now?
With the advent of online real-time IT support for business processes, the door was opened for recording and analysing data as it was produced. This gives an insurer the opportunity to monitor claims as transactions happen to them throughout the working day. With the development of workflow and case management, the fine detail of how many claims are being dealt with, who is doing what to any claim in the system and what stage it is in the overall process is available. This gives very valuable information to those responsible for the smooth running of the operation as they can detect bottlenecks and slow responses; they can see numbers, such as claim values or expenses, being processed and potentially crossing lines as they happen. The Key Performance Indicator, (KPI), became a target for real-time monitoring and the operational dashboard was born.
There is a lot more scope for traditional reporting and real-time operational dashboard style presentation of information to grow and develop. Business users now have more capability to drill down into data to discover more and more layers underpinning aggregated information.
What’s going to happen?
The exciting developments only just developing that I referred to at the beginning are those that seek to predict what is going to happen. One of our customers recently told me that 19% of their bodily injury claims absorb over 75% of their staff resources. These are the claims that could, if identified early enough in the process, be managed closely, shortening their duration, getting the claimant back to health and work and significantly reducing not only the operational cost but the indemnity spend. This can make significant differences in:
- Allocating resources
- Identifying potentially high value claims
- Recognising potentially fraudulent claims
At FINEOS we see reporting as a clear progression from understanding what has happened through to predicting future events:
- Reports – What has happened?
- Analysis – Why did it happen?
- Monitoring – What’s happening now?
- Prediction – What is likely to happen?
FINEOS Predictive Analytics
At the heart of predictive analytics are the algorithms that produce the scores from the raw data and the logic that assembles, categorises, assesses and interprets it.
Surrounding this are the manifestations of the intelligence gleaned from those processes:
- Real-time execution of scores to give current readings
- Scores refreshed in real-time as new data arrives
- Displays of information in a format easy to read and intuitive to understand
- Use of the analysis to triage, allocate and route each claim
- Further generation of tasks in the workflow system based upon the continual scoring of the claim as it progresses.
- Ability to rapidly edit and fine-tune the scoring
For example:
- What workflow(s) should be triggered?
- Which alerts and warning should be given?
- What escalation or reassignments should occur when scores reach certain levels?
The truth is that analytics have already done a tremendous amount for insurers, but we feel that the potential is only beginning to be unlocked. At FINEOS we are working with a global team of academic experts, insurers and government agencies to create predictive analytics that enable claims teams to determine in advance just which claims are likely to unravel and which will benefit from early intervention. And, of course, which ones you can leave alone and run through automatic or straight through processing. The beneficiaries will be your customers, who get a better service and faster return to health, and the insurer which reduces staff stress, gets better use of resources, lower operational cost and reduced indemnity spend. The seam is only just being opened up.
I’ll look more closely at this in the next blog.
