BI IQ

Remember the old Etch A Sketch toy that allowed you to “draw” things on a screen by turning dials either vertically or horizontally? Making a choice about which dial to turn is a decision many insurance carriers are facing today when they contemplate which way to proceed with business intelligence (BI) technology. Should they go vertical with an industry-specific tool? Or horizontal with a tool that crosses industry boundaries?

Today's BI software market is split between these two types of products, explains Matt Josefowicz, group manager of the insurance practice at Celent. “It depends what the task is and what people are looking for,” he says. “There are some insurance-specific data warehouses that have some reporting tools with them, but a lot of BI work–reporting, data analysis–is done with cross-industry tools.”

The cross-industry products tend to have a set of verticalized technology the vendors can give to their insurer clients, Josefowicz indicates, even though it's essentially a cross-industry platform. “The insurance-specific tools tend to be less in the reporting end and more in the repository end because those [repository tasks] imply a need for a specific data model,” he says.

In a recently published report for Gartner, Kimberly Harris-Ferrante, research vice president, argues the industry needs more vertical BI platforms and points to the fact many cross-industry BI vendors are investing in their products to make them more vertical. However, she concedes, finding a vendor for industry-specific projects may be a challenge due to market confusion. “Insurers must have a clear understanding of the demands of business users and select solutions that provide useful information for their job functions,” writes Harris-Ferrante. “Insurers should try to minimize the number of BI vendors used within the organization by selecting solutions that have strong horizontal capabilities as well as insurance-specific capabilities.”

As for where to deploy BI projects, distribution is a big target for insurers, notes Josefowicz, particularly producer profitability. Insurers also are generating some metrics and transparency to the claims process for P&C. “They are looking at the total book of business and customer data,” he says. “Any type of management reporting used to understand the operation and deploy resources to make the operation more efficient” is a focus.

INSURANCE SPECIFIC

Insurance House is a personal and commercial lines carrier that has developed multiple uses for business intelligence. “We use [BI] to aid our product development decisions, watch our claims activities, and adjust how our products satisfy the demands of our agents and insureds. It's more of a tool to help us put the right product in the marketplace, watch how it performs, and adjust [the product], if necessary,” says Robert Golden, director of strategic business services for Insurance House.

One of the reasons Insurance House selected the INSsight product from Skywire Software was because the product is insurance specific, explains Golden. The company didn't want to expend the extra time and work involved with tools that needed to be crafted toward the insurance industry when vendors such as Skywire already had provided a portion of the due diligence. By considering only products that dealt with property/casualty insurance, Golden believes Insurance House hastened the project's time line.

Without the benefit of INSsight, Insurance House would have needed to compile multiple concerns into a common view of standardization, which Golden indicates would have become Insurance House specific as opposed to insurance-industry specific. “Given our growth strategy, which is in part by acquisition, we want to position ourselves so we have a standard that will allow us to integrate data as we move forward,” he says.

“Where the insurance-specific applications add a lot of value is in collecting and normalizing the data and putting it into an insurance-specific data model,” Josefowicz says. “In terms of the reporting tools, the cross-industry tools do a good job once the data has been consolidated and normalized.” He believes the first element in any kind of data mastery project is identifying and normalizing the source data and pulling that all together into a place where it can be accessed by the reporting tool. Doing the reporting on top of that data is comparatively easier, he adds.

One area where vendors have been aligning themselves with insurance is in visualization, according to Art Barrios, managing director of the insurance practice at PricewaterhouseCoopers. Tools that support visualization take data and represent it graphically. For an insurer, this would involve mapping relative to where the carrier has a concentration of policies or a specific type of coverage, such as flood or earthquake. “[Carriers] use that for reinsurance purposes to identify where they might need to lay off risk or swap policies with another company that may be underexposed in that area but overexposed in another area,” he says.

Still another activity vendors focus on involves business analytics or predictive modeling–tools that are strong in the statistical space because insurers have to predict catastrophic losses. “[Insurers] use tools to come up with statistical or economic models they could feed into their pricing algorithms,” says Barrios.

Of the various available options, many cross-industry tools that have adopted the ACORD data standards tend to work better in the insurance space than others, Barrios contends. “Certainly there are uses within the insurance industry that specific tools align with better than others.”

USER-MANAGED SYSTEM

Prior to joining American Modern Insurance Group last summer as an assistant vice president, Mike Koscielny spent nearly five years with AAA Michigan, and both carriers are using the Fair Isaac product Blaze Advisor, a cross-industry tool with a strong focus in insurance and financial services. When going through the assessment process at AAA, the carrier studied all sorts of tools. Five years ago, the BI technologies still were raw in many respects, points out Koscielny. “They were code driven and aligned more with the technology people rather than the business users,” he says. “AAA's intent was to have the business users own this for various reasons–flexibility, regulatory tools, and the rest. When it came down to making decisions, the business drove that.” The only available business-user-managed system at that time that allowed business users to manage the flow of information was the Fair Isaac tool, recalls Koscielny. As for today, “having the technology in the hands of the users has been the focus [at American Modern], as well,” he says.

BI IN ACTION: MARKETING

Insurance House has territory managers in the field, and Golden explains the company relies on them for their assessments of opportunities the carrier might seize along with good competitive information on why the carrier's products might not be performing well. “The problem is if we took on every one of those opportunities, we'd be chock full of projects,” he says. “We try to use our intelligence to 'true up' those opportunities if they are hunches and try to prove what, in fact, is a good product move for us. Once we make that move–which gets my folks engaged, and we do some application development–it gives our product development and marketing folks an opportunity to wrap a campaign around it and monitor [the product] as it goes into production.”

GAB Robins is a third-party provider that handles claims operations for property/casualty carriers. The company uses a horizontal BI solution from Cognos, which despite being cross-industry, is popular among insurance companies, according to Harris-Ferrante.

One application the TPA provider uses Cognos for is to check what it calls service-quality metrics. “We use that to monitor the level of service we provide our clients,” says Curt Culver, associate vice president, resource management, for GAB Robins. “We identify potential problems we have in our organization. We can look at the process from a corporate level and draw a breakdown to the individual adjuster.” Culver notes this allows the TPA to address any potential issues it might have and keep the level of service where it wants it to be.

The second area in which GAB Robins uses its BI tool is to market its services to clients, enabling the customers to identify potential problems within their organizations. Such early warnings eventually would lead to a reduction in losses.

The TPA is using the Cognos BI product as the development platform, and this allows multidimensional reporting. “The Cognos platform is not [insurance specific], but the applications we develop are,” says Culver. “One thing we need to consider is the variety of clients we service on a day-to-day basis. We have self-insured from all industries as well as a variety of insurance companies that have similar requirements but frequently have their own requirements. It's important whatever application we offer has to have a level of flexibility that provides the capability to service each of our clients on an individual basis.”

CLAIMS AND UNDERWRITING

Prior to hooking up with Skywire, Insurance House didn't have any sophisticated monitoring tools in its claims operation to watch out for things such as fraud or evaluate the average close time on open claims. “A lot of things that drive our metrics from a claims perspective needed to be monitored more actively than we could prior to partnering with Skywire,” Golden says. “We are trying to keep the claims expense minimized by using the tool to monitor the adjusters' activities.”

At AAA, the carrier used the technology as part of the fraud detection process at the point of capture of loss, states Koscielny. “We were able to automate the process to alert adjusters to some losses that on the surface appeared to be questionable,” he says. AAA also was working to use the Blaze Advisor tool in the customer service area. “As a call came in and the caller was identified, the CSR would have a pretty good idea whether this was a high-profile person–not only with [AAA] membership, but auto, home, and life products–and this was someone to give extreme care.”

When carriers apply some business rules to their data, it allows the underwriters to make better decisions on exceptions, although Koscielny points out there are fewer exceptions today than there were prior to the use of rules technology. “One of the pieces we missed in that process was identifying the quality of the agent submitting the business and recognizing certain agents, based on their performance with us, may be more worthy of exceptions [for their clients] than others,” he says. “Not all agents are created equal.”

Making data available in a way that doesn't require the underwriter to spend hours looking through it means the underwriter can measure the agent's success rate relative to a half-dozen different criteria that are stored in the data warehouse and can be passed through rules. “That would be another step in the process to determine whether this is the right decision,” he says.

REPORTING CAPABILITIES

Bankers Financial Corporation recently decided it could enhance its reporting capabilities with a cross-industry BI tool from Lawson, according to Brad Martz, executive vice president and CFO for Bankers. “It's going to start out as an accounting tool,” he says. “We're going to automate managerial reports with our financial reports. We will be bringing in nonfinancial information or financial information related to production or claims activity that wouldn't necessarily be tracked efficiently in the general ledger.”

Bankers has produced its product profitability reports manually outside the general ledger in Excel. “We feel with BI we can bring in the reinsurance information, the allocation information, and some of the product-level information necessary to do the profitability by product, agent, and distribution channel all within an automated fashion within Lawson,” says Martz.

GOOD DATA

No BI project will succeed without quality data being in place, and Golden reports Insurance House had some challenges in that regard. “One of the things we encountered was disparate data once we tried to bring all the data together,” he says. That was the initial point of the carrier's relationship with Skywire. “We procured the actual model so we could cleanse, transform, and standardize the data, and at least we could have one truth, if you will, and then push the data into the marts that became and are becoming our intelligence edge,” adds Golden.

Insurance House worked with the technical folks at Skywire to cleanse the carrier's data and make sure it was useable before the carrier got into the phase of using the data in intelligence. “The biggest concern I had was to have data that was highly available and on the Web–but wrong,” Golden says. “We've taken great pain to make sure that's not the case and that we are working with good, solid information at the point of decision.”

While the insurance industry is data rich, insurers are faced with barriers that involve data quality as well as legacy issues that contribute to poor data quality. “To drive best practices BI in insurance–and financial services in general–you really have to address some broad fundamental issues around the quality of the data and how companies are managing that specifically so you then can feed it into a BI solution,” says Barrios. He mentions the old data lament “garbage in, garbage out” and points out, “You can't implement a great BI tool if you have really bad data.”

Martz explains Bankers must make sure the interface with the company's legacy systems is accurate. “Once it's interfaced, it acts like a querying tool–a powerful report writer that will attach to our claims system, our product system, and our general ledger system all at the same time,” he says. “It will consolidate the tools used throughout the company. There is not a standard uniform process of pulling down data from our systems. That makes it difficult for IT to support. The licensing costs for these different applications are ridiculous. It's going to be nice to mandate a standard across the enterprise on the reporting tool.”

LENGTHY PROJECTS

BI and data projects are time intensive, but carriers have to view this in the context of the value they receive from such projects. “A lot of companies, especially in the insurance space, do not fully appreciate the strategic importance of data to their company,” asserts Barrios. “In the 21st century, most financial services companies are in the relationship business. They have to build a relationship with individuals or organizations.”

Carriers also are in the information management business. Everything they do to develop that business relationship almost immediately becomes a piece of information–policy information, servicing, investment of funds. “If they don't manage data competency as critical to achieving their strategy, they just are not getting it,” says Barrios. “The ability to manage data is that important to achieving their strategy.”

Barrios warns carriers not to bite off everything at once. “Do it in a tactical and incremental way,” he suggests. Most companies tend to start with customer data in terms of standardizing it, integrating it, cleansing it, and keeping it clean. That is an area they easily can assign a specific business value, which tends to be around revenue generation–for example, the up-sell/cross-sell opportunities–as well as overall customer service in terms of being able to achieve the single view of the customer. “Ultimately, start small and enable a specific business process,” he says. “Make sure those lessons and investments are reusable for another project. If you keep doing it that way, eventually you look back at all the projects and you will see you have transformed yourself significantly. Bigger projects tend to lose focus and not be clear on what specific business value they are delivering.”

Barrios has seen companies buy a BI tool and then try to figure out what to do with it. “Focus on what your strategy is first,” he says. “Then figure out the processes you need to apply that data to the business process. Finally, you can take a look at the tools that can support this. If you pick the tool first without a clear strategy, the tool defines what the process is going to be.”

NIGHT AND DAY

Golden believes Insurance House has tweaked its long-range business strategy significantly as a result of good data currently available to support trends and make strategic moves. “The area I inherited has been able to contribute at the strategy table in a much more significant way,” he says. “We bring a ton of credibility now, whereas before what we had were questions coming at the information reports we provided. It's literally a night-and-day experience from three years ago, when we first began the data model.”

Koscielny believes insurers have to think out of the box when it comes to business intelligence. “The focus I've had in my last two jobs has been raising the level of productivity at the point of decision-making and driving these decisions to human beings and giving them enough information to make a decision within a short period of time,” he says. “From my perspective, we haven't tapped the full potential of the rules technology that is available to us partly because we've just begun as an industry to gather all this data and put it someplace where we can do these things.” TD