Underwriting engines have been around for over 20 years. Initially they were used in head offices, processing application form information keyed in by clerical staff, and avoiding the need for (expensive) underwriters spending time on the more straightforward cases. From there data entry migrated to the point of sale, but still often with processing and decision-making deferred until the information could be uploaded in a head-office environment.
But since those early days, use of underwriting engines has grown, and in certain markets, such as the UK and Australia in which specialist medical information is used sparingly, underwriting automation has become ubiquitous. What is more, high straight-through processing (STP) rates – that is, where a final decision is made without any human involvement – are high. Indeed some are very high: depending on classes of policy target market and underwriting policy, STP rates around the 90% mark are being achieved.
Recently interest in underwriting automation has spiralled, and even in the USA where traditional underwriting, with medical exams and lab tests, has long been king. American insurers have looked at ways of minimising costs and tapping into new markets via propositions that make life insurance easy and quick to buy. So-called ‘simplified underwriting’ uses algorithms applied to third-party data (‘predictive analytics’) plus information from traditional databases (previous rated decisions, motor vehicle records, prescription histories) to supplement the usual application form. If the app is clear and the database information and analytics add up, then the case can be accepted. After some initial caution, insurers are now prepared to accept sums assured of several million dollars on younger applicants in this way. And of course an underwriting engine is a key part of the risk evaluation and acceptance process.
Three things have ushered in an exciting era for automation. One is that experience in the US. The second is the growth in interest in use of external data. And the third is the use of modern technology – machine learning, natural language processing, etc – to extract key information from medical reports and records without the need for laborious work by humans. (And arguably there is a fourth too – COVID-19 which has emphasised the benefits of buying and dealing on line.)
So underwriting engines are no longer confined to processing application form information: they can now take charge of the whole underwriting process. Information from application forms, third-party databases, medical exams, lab tests and (via the latest data extraction techniques) medical records can all be fed into the engine at appropriate stages for automated decision-making.
Clearly there are big benefits to be gained from deployment of a capable underwriting engine:
- Efficiency – less reliance on human underwriters
- Speed of service to customers and distributors – case processing in minutes rather than in weeks
- Quality of service – convenience and transparency, a better customer journey
- Consistent application of underwriting philosophy.
But to focus on just these is to miss the point: there is a bigger picture. Automating the underwriting process has the power radically to improve the business – transform it even. It enables new customer propositions and underpins a multi-channel strategy. It revolutionises the on-boarding process for all stakeholders and truly moves the insurer into the digital age.
Data: ‘digital diamonds’
Just as important as all these, automated underwriting generates data, those ‘digital diamonds’ that, if used smartly, enable vastly improved management – not just of risk but of the business itself:
- Monitoring of engine performance and efficiency
- Understanding the make-up of the new business flow – classes of policy, sums assured, buyer profiles
- Getting a decent handle on non-disclosure – what, who, which distributors?
- Monitoring channel and individual distributor performance: who is really bringing in the valuable business?
- Getting a better fix on persistency and weed out under-performers
- Understanding the risk make-up of the portfolio: is it as expected or does there need to be a change in underwriting, targeting and marketing?
The results of this sort of data enable improved business planning, strategies and tactics.
Underwriting automation makes sense at every level and benefits all stakeholders in the business. But insurers need to think big, to think strategy and transformation. The worst mistake would be to create an e-process based on a traditional paper one.
The right engine
Deploying the right engine is vital. Given the highly developed and effective products out there, don’t even think of building your own. And don’t just go it alone. Partner with a trusted supplier with a proven track record for true excellence. Be sure you are deploying the best tools for the best process and the best overall outcomes.
So, what is your company’s digital strategy and how does automated underwriting fit into it?