What's your take on the "state of big data" in the insurance industry?
Conlon: We work with a lot of carriers on our Research Council, and one of the most common topics they want to talk about is, what is big data and what does it mean to me? Most organizations are in the position of trying to figure it out, except for the large supercarriers that have dedicated staff actively engaged with it.
It really is a mixed bag right now, in terms of the ways carriers are approaching big data and what they're doing. There are relatively few insurers actively using big data at this point. Most organizations are aware of big data at a high level, but not so much in terms of what it means to them specifically.
There are organizations that have spent the last 15 years developing data warehouses; now they are looking at predictive models. These are becoming more standard business intelligence practices. There are midsize commercial organizations using things like satellite models but manually incorporating [the data] into their business models. Those organizations now are actively trying to learn as much as they can about big data. There are lots of data architects and enterprise architects going to every conference they can, trying to learn how insurers are using big data and what the potential is. Small and midsize carriers really are just researching it now, but they know they have to make a move.
There is movement from a few very large carriers. Some are actively using big data, mostly in areas such as telematics or customer experience data, or [are using it for] rating and pricing in personal and commercial auto. Direct writers are using information they are capturing on their websites in order to maximize the customer experience and be able to bring up the right products and services for customers. Some have hired chief scientists to look into big data. They are in the investigation phase and are actively trying to find out how big data can make a competitive difference for them.
It sounds like a lot of companies are struggling to make a business case for big data investments. Why is it still so hard to do this?
Conlon: That's because it has two major pieces: benefits and costs. It is very difficult, at this point in our industry's use of big data, to specify the benefits. You really are funding an R&D effort, so there is an unknown upside and benefit. You can't easily quantify the benefits of enhanced customer experience or improved risk selection or risk management, so it is hard to quantify the benefits of a big data project. But it is easy to quantify the costs. Often these projects come with a big price tag. Without an understanding of the tangible benefits that will come, [there is] a limited ability of the business community to understand what they get from it. If they have larger IT budgets it's a bit easier, but not all carriers have budgets that can sustain these kinds of programs.
Even if you can pull together an estimate on the benefits -- for example, for a telematics initiative -- a lot of carriers have other priorities that are competing for those budget numbers. If you have a policy administration system that doesn't allow you to introduce products in under 10 months, that is going to be a priority investment [over a big data initiative].
Who in the organization typically is charged with making the business case and leading big data initiatives?
Conlon: To some degree, we find that data architects are trying to lead the research efforts and the understanding of the technology and to convey the benefits to the business. The data scientist is someone who understands the domain as well as the [technology] -- the mechanics of the business. That is often difficult to find -- sometimes they come out of the actuarial department. They can be major sponsors and stakeholders.
Most carriers are watching the industry to understand when this will become competitive table stakes. So the business community, and the data architects and scientists, hopefully are keeping abreast of what other carriers and other industries are doing with big data. For the most part, most of them are going to be followers of the very large carriers. It does need to be a combination of IT and the business working together. Ideally, if you have a data scientist, that's the person who bridges this.