Top 5 Trends for 2013 — Analytics

At a time of incredible change in the insurance industry, IT is right in the middle of that change.

"It is extremely critical across the economy, but especially in information businesses such as insurance," says Matthew Josefowicz, managing director and partner for Novarica. "And the challenge is we're in a world of rapidly increasing complexity and rapidly decreasing user patience, and it's up to IT to bridge this gap."

It starts with data, Josefowicz says. "It used to be that insurers could compete on knowing more than their competitors. They'd actually find out information their competitors didn't know. This is becoming not possible. Everyone can buy the same information. What companies have to do is out-analyze that data, out-smart the competitors in that data, but it's not possible to know more than the competitors."

This is where advanced analytics in all forms have come into play for multi-line insurer, Alfa Insurance Group. "It's not a question of whether you think this is happening or not," said Al Schellhorn, SVP, chief underwriting and development officer for Alfa Insurance, during his presentation at Property Casualty Insurers Association of America's Information Technology Conference in October. "Your competitors are using data, analytics, predictive modeling, generalized linear modeling; they're using all of these against you, either in ratemaking or-as the best-in-class are using it-in operationalizing in other areas of the company."

Novarica's "US Insurer Budgets and Projects for 2013" report backs up Schellhorn's assertion. Among property/casualty insurers, business intelligence and data is at the top of three business capability priorities for 2013. For large property/casualty insurers (more than $1 billion in premium) it is the No. 1 priority. For property/casualty insurers with under $1 billion in premium, it's No. 2. These insurers currently, or plan to, apply data and analytical capabilities to a number of areas, according to the Novarica report.

For large insurers, the most common use will be to drive frequency and severity prediction, loss reserving, and detecting unknown patterns of fraud. Other insurers will also use it for frequency and severity prediction and loss reserving, but third on the list is premium audit.

Alfa, which is made up of eight property/casualty companies, one life company, four non-insurance companies and four agency operations, currently has more than $25 billion in force and Alfa's property/casualty companies service more than 1 million policies.

The challenges it faced almost four years ago are familiar to many insurers: silos of data, 120 sources of information, 17 lines of business, data that wasn't very easy to get to, and a number of definitions. "The reasons there are different metrics [and definitions] is because people want to measure things differently and they probably have legitimate reasons for doing so," Schellhorn says. Executives knew they had to make the investment. Schellhorn told the president and CEO what they were going to face with this type of project: millions of dollars, and a staff solely dedicated to this project for three years.

So the company started in 2007/2008 with building a scalable, sustainable enterprise data warehouse. "You don't have to have a warehouse to get into this space. I believe it works well to have your history for your modeling work," Schellhorn says.

In 2011, Alfa began its three-year project by loading all of that history and operationalizing with the customer data mart and master data mart. Then they moved on to customer-lifetime value. From there, Schellhorn says, the plan is to get into predictive sales leads-finding out who the best customers are and targeting them.

Alfa is now applying analytics across a number of areas in the organization. It is able to take the data, put it in a predictive model, and at customer renewal, price more accurately. "This has been years in the making, something our actuaries have been talking about for years and is proprietary and uses advanced analytics," Schellhorn says. "We're also able to integrate different levels of competitive analysis on ratemaking, benchmark our sales performance, and predict more accurately who is going to give us losses in the future."

Enterpriswide or Targeted or Both?

Precision, personalization and collaboration will propel the coming wave of analytic capabilities, according to Kurt Schlegel, analytics research VP for Gartner. In a recent presentation entitled "The Future of Business Analytics," Schlegel said data architects should look to augment rather than replace existing operations with an eye toward agility and acceptance of external and varied data sources. Here, Schlegel sees opportunity for more acceptance in data discovery tools, especially where there is little training required and as big data frameworks become more visually appealing.

 

There is also an opportunity for industry-specific, trusted data aggregators to fill in enterprise data gaps with as-a-service options. Schlegel says existing providers of benchmarking and industry analysis are putting together niche offerings that provide the content that many enterprises may not be able to for lack of in-house talent, budget or existing data resources.

Don't underestimate the leadership and expertise aspect, Schellhorn says. "Data is foundational, data quality matters, but having a jammed-up data warehouse that no one uses doesn't make any changes in your organization. If you're looking for this kind of talent, you have a choice: You can spend a lot of money for somebody who's been there and done it before, but probably can't figure it out now with today's rules or find a kid who's been training [in data analytics] right now and then teach him or her insurance. We went the latter route. I want someone who looks objectively at the data and then presents that to management." At that point, better decisions can be made.

Just like any project an insurer takes on, it all boils down to enabling business capabilities through technology, Novarica's Josefowicz says. "That means working hand-in-hand with business leaders, with organizational planning, with corporate strategy, making them aware of what's possible, the costs and the restrictions of the things they may be considering."

For reprint and licensing requests for this article, click here.
Analytics Data and information management Policy adminstration
MORE FROM DIGITAL INSURANCE