Automating the underwriting process holds great potential, but to realise that potential you need the right product or products. There are plenty of software suppliers and plenty of decision engines out there but there is a big range of capabilities too.

Some have theoretical capability. We worked with an insurer looking to overhaul its back office and introduce on-line submission of applications with automated underwriting. Its admin software supplier said it had a decision engine too and so the insurer decided to adopt it – only the engine turned out to be a basic ‘shell’ with limited capability and nothing like the power of a well developed tool. The back-office system works fine but the automated underwriting has yet to be implemented.

You need an engine that has been designed especially for the complex life insurance underwriting environment. But even some of those products don’t come with a rule-set or ‘knowledge base’ installed, which means that there is the big task of programming the engine with the intelligence required. You could argue that you would end up with a set of rules bespoke for your business, but the effort required far outweighs that of customising an already comprehensive knowledge base; why sweat away when the hard work could have been done for you already?

Avoid self-build

And don’t be tempted to build your own. That route might sound appealingly cost-effective but… The better engines have been invested in and developed over a number of years, and boast a range of powerful features. A self-build project is highly unlikely to be able to replicate anything like them, even if you use a proprietary decision engine (and there are plenty available) as a base. How about timing and speed to market? Can your IT team schedule the build you want?

And how long is it going to take? Buying in proven technology means that the new system can be up and running within a few months, with full capability from outset; the alternative is an incremental, hesitant implementation over a period of years. And don’t forget project risk and the potential for negative impact on the business: proven tech wins hands-down, and furthermore the supplier has an experienced implementation team to help smooth out any wrinkles that arise.

Ownership by the business

Finally, business software needs to be owned and managed by the business users. The best automated underwriting tools are not just managed (monitored, tweaked and maintained) by the business, but they are easy to work with: no specialist programming knowledge or skills required. Could the same be said for a ‘home-grown’ system? And for sure you don’t want to be asking the IT guys every time you want to add a rule or adjust an existing one. (They would probably say they could schedule that in in three months’ time.) At a conference we listened to a major US insurer describe how they went the self-build route. The firm was pretty pleased with what it achieved… but what it didn’t know was how much better an outcome they could have had by buying a good product from an external supplier.

Your shopping list

So what essentials should be on your engine shopping list?

  1. A modern system architecture making for easy integration, reliable development and updating, and API interactivity with other systems and tools; given the connected ‘eco-system’ structure of business today, that last point is important.
  2. The product should be highly configurable so that it can be customised to your individual needs.
  3. Multi-channel capability is required.
  4. Multiple data feeds at various stages in the underwriting process need to be accommodated given the increasing use of direct data inputs, especially from external agencies.
  5. A comprehensive rule-set (beware the empty engine referred to above); and the rules should be able to use a piece of information more than once – repeating a question detracts from the customer journey.
  6. The ability to vary rules, perhaps according to channel or individual distributor; it shouldn’t be that one rule-set fits all.
  7. A strong capability for handling risk factor interactions – ‘combination rules’ that deal with, for example, the interdependency between diabetes and hypertension.
  8. Ease of use by non-IT specialists, especially the editor tool that underwriters will use to create and re-configure rules.
  9. Strong, clear version controls so that it is easy to see what underwriting rules were in force at any given time, and what changes were made and when.
  10. Powerful analytics for business insight.

Harnessing the power of data

Point number 10 is absolutely crucial. Data produced by the engine and throughout the new business process, especially when allied with in-force data as the portfolio develops, is key to understanding the efficiency of the whole on-boarding set-up, the make-up of the new business stream in terms of customers and risks, and the performance and value of individual channels and distributor partners. The results of data analysis are an essential part of the successful strategic management of the business.

So a good data analytics module is a key component of an engine, and some feature a ‘workbench’ that provides an overview of the new business pipeline, tracks progress of individual cases, monitors engine/rules performance, produces comprehensive reports and maybe directs cases to the appropriate underwriter when human involvement is required.

And given that the e-application form has a close and direct relationship with the engine, you might want to consider sourcing the customer interface software from the engine supplier, if it offers one. It is not essential that all the software driving the new business process comes from the same manufacturer, but sometimes sourcing more than one piece of kit from the same place can make life a bit simpler.

Not all are equal…

So, not all automated underwriting systems are created equal. Shop around and choose carefully. Ensure that you acquire the strong capability that will continue to benefit your company as it and its business grows.

Leave a Reply