By Angus MacCaull

2020. The year suggests perfect vision—the beginning of an era when everything is easier, cheaper, and more abundant. But what can we really see about the future? And how will emerging trends affect insurance?

Developments in a variety of sectors have the potential to transform aspects of the insurance value chain. AI. Insurtech. IoT. It can be hard to keep up with the buzzwords and acronyms. In my view, the megatrend running through them all is an explosion of data. We now have the ability to measure and analyze more than ever before. For insurance, we think big data will enable us to approach our relationship with risk in a new way.

Here are three principles for working with big data. These principles will become increasingly relevant to insurance companies, brokers, and startups looking to keep or gain the trust of communities that need financial protection.

 

#1 Garbage In, Garbage Out

This is a popular saying in computer science. If your data collection or entry is flawed, your data set produces flawed summaries. When you set out to scale up a data set, it’s critically important to be careful—that is, to legitimately care about—the seemingly mundane work of data collection and entry. If you’re only focused on a flashy presentation deck at the end of the process, you may find that the summaries in your slides are meaningless.

 

#2 The Map Is Not The Territory

Just like you can’t eat a recipe, you can’t actually travel anywhere on a map. Data always points back or forward to something else. The concept at stake here is called reification, which means treating an abstraction as the thing itself. Reification can be very helpful, but we have to remember that business numbers don’t just exist on spreadsheets or application dashboards—they always represent some kind of material behaviour.

 

#3 A Model Makes It Work

A carefully collected data set treated as an abstraction still needs a model for interpretation. Business models usually only apply to specific industries or geographic regions. The best ones sometimes only apply to individual companies. You need to know the social, economic, and regulatory environment. What does the data mean to the people you work with and serve? How do your coworkers and clients behave?

 

A group of people discuss some data on their devices

 

New kinds of data at new volumes will shake up the insurance industry in 2020 and beyond. The good news is that our industry has always included a lot of data. Pricing for policies has long come from tracking different risk indicators at scale—sometimes over centuries as actuaries look at fire records and weather patterns.

Many of us in insurance already have some familiarity with the three principles above. We apply them in different ways intuitively. Our challenge in the future will be to activate our implicit knowledge of big data and use it more consciously.