We see it all the time: Insurance companies are striving to get ahead in the field of business intelligence. They’ll invest, often heavily, in systems such as Oracle Business Intelligence, Microsoft Power BI, and so on… and yet they struggle to find the value from these investments. Why is that? Many times, it’s because they have not used the value of business intelligence in insurance.
In this article, we’re going to reveal, from our numerous business intelligence and advanced analytics engagements in the insurance industry, where the business intelligence value is hiding… and how you can extract it. It’s based on real stories and real insurers, ranging from regional specialists to dominant multi-line giants. As you read, you’ll see what we’ve seen, and hopefully will be able to apply the lessons learned to your own insurance organization in the pursuit of optimal operational business intelligence.
Business intelligence in insurance: Turning unstructured data into gold with insurance BI
Many of the issues arising in insurance business intelligence can be traced to “completing the final mile.” We’ll often encounter insurance enterprises that are practically swimming in data, yet failing to leverage its power. The problem is that these insurers aren’t reconciling the data available for business intelligence with their own business needs.
It sounds simple, but the problem is common, and its roots are deep. The company’s business units and its IT people are always wrangling data and generating reports, but they’re doing it in a vacuum. They’re working with insufficient direction as to what the executives want to learn from the front lines.
This is where we often step in and, based on our experience, are able to quickly provide actual “intelligent” insurance business intelligence that proves an eye-opener for claims executives. Here’s an example: We’ve encountered insurers that want greater visibility into their subrogation operations. They may be collecting billions from other parties each year, but is that truly satisfactory? Industry benchmarks can certainly help (and we provide them when they’re missing), but they also must be treated with caution. Simply knowing that you’re in the top quartile, for example, can hold you back from attaining peak performance. Ditto for averages: They can squelch the standard deviation for metrics like reclaim rates.
Based on our experience, we can help insurers focus on the business intelligence that truly makes an impact.
Business intelligence in insurance: How to button down business-unit BI requirements
Modern insurance business intelligence software is impressive. It can pull data from a nearly limitless supply of sources. But which ones do you choose and what systems or claims data sets do you connect? The power itself can lead to problems. We see this all the time.
And that’s assuming that the data is even readily available. Too often, it’s not. We’ve worked with prominent insurers whose data has resided in, say, 20 different sources which we’d need to query (and clean up!), daily, simply to generate the initial insurance business intelligence dashboards. (We were able to eventually overcome this daily grind for our clients, as we remediated the issues that created the situation in the first place.)
Here’s one of the sources of the confusion: The executives want business intelligence in insurance to generate real-time reports or dashboards that give them actionable insight. But the people tasked with making those reports don’t know enough about the executive/big-picture-view of the business to configure the insurance BI application with the right things to do. They’ll generate lots of reports, all right, but they won’t be terribly useful. An IT person could—and often does—assemble and manage business intelligence for business units. But they’d need to understand which parts of the data to use and aggregate so that the final KPI (key performance indicator) in the resulting dashboard represents reality. Many of the lower-level business people lack this perspective as well. So their “business language” will talk with the “IT language,” and the executives will continue in the dark.
The answer is proper reconciliation of the data: marrying the right numerators and denominators to produce a set of “drill-down-able” KPIs that insurance industry leaders crave.
Business intelligence in insurance can stumble on “false precision” and measuring too much
Our business intelligence work with insurance enterprises often reveals some interesting differences between the data that exists and the data that matters. What will happen is, over time, new systems, with new capabilities, will get added to an insurer’s operations, and the sprawl of new data goes unchecked. We worked with insurers who have had about 200 different policy definitions. But it turned out that fewer than 20 of them represented 80 percent of the volume. We parked the remaining 180 in an “Other” category in order to “properly feed” the business intelligence application. The IT people wailed: “You can’t ignore 90 percent!” Sure we can—if that 90 percent isn’t worth anything. Put it one way: We help fix the “garbage-in” problem that clogs too many insurance business intelligence implementations. Put it another way: It’s like getting 500 channels with your cable TV package… and then scanning through them all to unpack the half-dozen that you’ll actually end up watching.
Insurance Business intelligence: Plugging it all in and summing it all up
“Data enrichment” is the name of the game in insurance business intelligence. Once the information is properly cleaned, set up, built, and tagged, we can wrangle it in ways that “the data guys” simply lack the business insight to do. That’s not a shortcoming of theirs; they just lack the operational experience.
But once the insurance business intelligence engine is properly primed, sit back and be amazed! We recently showed the head of P&C (property and casualty) claims at a major insurer how easy it was to drill down through data—from the enterprise level to the desk level—with the intuitive ease you’d find in, say, a CNN election-night map. “Look! That’s our subro organization! That’s our demand process! That’s our respond-to-request process! Here are recovered claims dollars! Here, for the first time, are the unrecovered claims dollars!” Those were the kinds of reactions we got from this leader and his team. What he was seeing was true business intelligence in insurance, properly implemented and optimized. You should see it, too.
Is your insurance business-intelligence initiative not quite as intelligent as it should be? Contact The Lab today. In just one 30-minute call, we can give you a live screen-sharing demo that will show you just what BI can do. Even better, our business-intelligence engagements self-fund—that’s right, they pay for themselves—in just six months or less. And that promise is money-back guaranteed. Contact us to book your demo call today.