A recent Accenture survey in the United States highlighted a puzzling statistic — 70 per cent of an underwriter’s time at work is involved in non-underwriting activities.
Meanwhile, 40 per cent of their time was spent on administration and 30 per cent on negotiation and sales support.
I suggest the experience in Australia is similar.
Given that underwriting is so fundamental to insurance, surely underwriters should be able to devote more of their time to actual underwriting?
The three top reasons cited by the survey for the lack of time spent underwriting are redundant inputs/manual processes (71per cent), outdated/inflexible systems (40 per cent) and a lack of information/analytics at the point of need (36 per cent).
In many cases, underwriters are still reliant on several manual inputs, making the gathering and analysis of data more difficult.
Intelligent decision-making in a competitive environment becomes more time consuming and may lead to questionable results.
Getting on top of the data
Whether working directly for an insurer, or with an underwriting agency, it is very important that the underwriting role is supported by the appropriate organisation structure, processes and decision-making support systems.
All decision support systems require accurate and timely data to be on hand. A lot of insurance software architecture involves multiple sources of data in the form of unwieldy extracts loaded into specialist applications. Unreconciled data means people have difficulty understanding the analysis and delivering underwriting results.
It’s important to have one completely reliable source of truth, or a golden record so that everything can be reconciled. Without this, valuable data is overlooked, which hinders good underwriting.
Technology can be evil
Sometimes the technology can be so complex that it just takes up too much time to get a result or, for underwriting agencies especially, is simply not powerful enough to effectively support the underwriting team.
I often hear professionals asking whether underwriting is an art or a science. My view is that it’s probably a bit of both. If the process becomes too structured — based on algorithms without interpretation — we can get weird results. Robodebt is an example of such evil technology.
If your processes and technology mean that underwriting takes too much time and you are getting strange or incomplete results, it’s time to review how your processes and technology work together.
With the underwriting cycle becoming much tighter and markets in Australia suffering from floods, fires and other events, there is currently considerable portfolio remediation.
This can mean having to re-underwrite an entire portfolio in order to deliver an acceptable return on equity and understand what’s required in terms of risk selection, pricing and deductibles.
For every bad risk, good risks need to cover the claims cost and loss experience.
To successfully remediate, underwriters must have access to solid data and forms of analysis. You don’t want to remediate your portfolios every year. It is painful enough the first time, particularly, if the results mean there are significant changes to the products and pricing required.
Explaining why the same product costs significantly more (or less) can be very difficult and impacts your customer value proposition.
Align your systems and data
The answer lies in systems, data and processes that align.
Too much manual intervention can be dangerous because it takes up valuable time and you get inconsistent results.
On the other hand, too much technology and reliance on algorithms can mean losing the value derived from the ‘art’ of underwriting.
Look for a balance that gives underwriters the opportunity to underwrite while maintaining solid data oversight and portfolio analysis to give the best chance of strong, ongoing profitability.
*2021 P&C Underwriting Survey, Accenture - The Institutes Risk and Insurance Knowledge Group
Are you sure you want to delete your comment?
This cannot be undone.