Introducing ClaimsX Ultra: The next evolution in automated fraud detection and deterrence.

[Webinar Recap] Automating Underwriting: Turn Art into Science

Providing coverage for small businesses just went from problematic to difficult. New challenges demand new data elements which must be gathered from emerging and alternative sources and refined using artificial intelligence and analysis. But how does the underwriting process go from an artful approach to a scientific one? It comes down to data: not just finding new data types, but unleashing data’s true power to shape the underwriting of the future.

This blog is based on “Automating Underwriting: Turn Art into Science,” a webinar presented by Carpe Data’s Product Solutions Consultant, Chris Longano, CPCU, as well as Gail McGiffin, Principal/Partner at EY Advisory Services, for PC360’s webinar series on June 23rd, 2020. You can view the full session recording here.


New world, new challenges

The world has changed dramatically in 2020, and while the term “the new normal” is heard over and over, the truth is we don’t know what the new normal is going to be yet. The world is struggling with effectively managing a remote workforce, when possible, and this brave new world is providing more than enough challenges for employers—and underwriters. The unique insights provided by new data sources are required for accurate underwriting now and in the future.


Businesses are changing

There is a massive demographic shift going on right now in small and medium-sized businesses because of the global pandemic. Carriers should take a step back to look at the big picture—these make up a huge client segment, and how are you going to deal with them? How could these business models evolve due to the pandemic and its ripple effects? Carriers need more data, more insights and, most importantly, more RULES. “Those who have the best rules win,” said Gail McGiffin. 


Moving to a more scientific approach

The industry has yet to unleash the full power of the complete array of emerging data sources. The trick is how we look at data, and doing it differently than we have in the past. There are now “dimensions” of data that need to be galvanized to get the most from it. The aggregation of data provides perspective and a path to no-touch automation and improved underwriting capabilities.


Policy interaction and misclassification

A very important shift is happening in the commercial P&C world, one that is moving from a model that relies on human review to one that relies on human intelligence—but it’s not happening fast enough. Thanks to the new data sources, there is an increased level of interaction with the policy throughout its lifecycle. Businesses change, grow, and adapt, and a carrier that’s not keeping up can lose out on a great opportunity to fine-tune its book of business.


Ready to hear the entire conversation? Watch the full webinar free here:


Questions (and answers!) from the webinar

How should we think about the cost/benefit analysis of introducing new underwriting or loss control data?

This will certainly vary from carrier to carrier but, in general, we see the evaluation affecting hard costs like evaluating the premium lost/leaked from misclassification versus cost of first year and future years; soft costs like time saved within the underwriting and reunderwriting processes; and opportunity cost such as the ability to adversely select and competitively price for the desired businesses. 


Will there be standardization across the board in terms of RFQs, data exchanges, etc.?

It is certainly a question worth asking, and there are some ventures out there seeking to do this today. However, the distribution model often takes negotiating control away from the carrier who likely wants to have a direct agreement with the individual data providers. 


Can you provide an example of how Carpe Data helped a carrier increase their automation in underwriting?

One example is the use of classification profiles to model for and automatically select a classification, as well as decide which conditional questions to ask, if not already answered by the data provided, to more quickly process an application.