Mise en Plase
I read a terrific article by Caribou Honig, Chairman and Co-Founder of Insuretech Connect, where he referred to APIs as a “business strategy masquerading as a technology strategy.”[1] We at Fenris couldn’t agree more.
Up until now, insurance was not a data-driven industry. That sounds ridiculous, right? After all, insurance companies have vast data stores including quotes, binds, premiums, personal data, household data, business data, claims, and more, for every line of business, covering a long period of time. It seems there’s so much data that predicting claims, understanding propensity to buy or sue, cross-selling, or assessing the quality of a lead would be normal business. It isn’t.
There are a lot of reasons why, but one is that many insurance companies are the result of acquisitions, each having its own policy administration and claims systems. On one hand, the cost for migrating all lines of business and every acquired entity onto a single platform is prohibitive. On the other hand, getting the data all in one central place to analyze is essential. What to do?
Insurance got part of the way there with data warehouses and conventional BI tools, but warehouses can’t take them to the next step: machine learning, in which systems are able to automatically learn from data and improve from experience without being explicitly programmed. It is these kinds of programs, models, and results that will move insurance forward and answer questions like, “is this a good lead,” or “is this person ready to buy,” or “does this person need another type of insurance too.”
The key to it all is data. The art of “liberating” one’s own data is only part of it. It is the mixture of that proprietary data with public and private data found in thousands of repositories all over the country that will make the difference in a carrier’s bottom line.
Back to APIs: finding data, curating it, combining it, learning from it, and deriving business-impacting results can only really happen in an industry where data is readily exchanged, enhanced, and modeled. Those activities require Automated Programming Interfaces. They’re not new, but they’re new to insurance. Caribou calls for the “API-ification” of the industry. I would go further and suggest that public databases held by the states need APIs, public data from universities need APIs, and data vendors need APIs; it is in combination that the magic happens. It’s hard enough to do even with APIs. In their absence, finding, acquiring, training, and testing data is a tough job. Good thing you have Fenris to do it.
“Mise en place” is a French culinary phrase which means “everything in its place.” The concept is to gather all your ingredients at your cooking station: organizing and arranging meat, sauces, spices, broths, and cut vegetables that the chef will combine to create a wonderful meal. So too, does Fenris organize and arrange the data components that make a wonderful model. We use every means available to stock our kitchen and to deliver results, including APIs. Like any chef, the more components Fenris has at its fingertips, the fresher the data, and the excellence of our recipes will delight the insurance palate.
In my next paper, I will share the way our APIs can positively impact leads and quotes.
- Dana Isaacoff is a 30-year veteran of information technology and the CIO at Fenris LLC. She has helped dozens of companies grow by streamlining processes and tapping into stored data.
[1] https://insurtech-startups.insuranceciooutlook.com/cxoinsights/the-road-to-success-in-the-insurance-industry-nid-314.html
