Filed Under:Technology, Analytics & Data

Turn It Around: Changing Data into Business Intelligence

There’s no shortage of data in the insurance industry, but most carriers realize there is a shortage of operational intelligence. BI projects are designed to take some of the dizziness out of that information overload.

Insurance carriers have accumulated massive amounts of data over the years, which can present a daunting task when it comes time to turn that data into business intelligence (BI). That is why gaining some control of that data is the first step recommended by industry analysts and insurers to achieve a successful business intelligence plan.

For Karen Pauli, research director in the insurance practice for TowerGroup, the first step calls for getting your data in order, but there are other steps that need to follow closely.

“You have to have some business objectives,” says Pauli. “What are you trying to accomplish with this business intelligence initiative?”

Business intelligence projects can come into play across most of the functional areas of an organization, according to Steve Callahan, practice director for Robert E. Nolan Co. Today, BI tends to apply to dashboards and executive scorecards, but Callahan believes the industry is shifting to more predictive analytics, the next generation of BI, which he feels gives organizations the opportunity for competitive differentiation.

The success of a BI project often depends on the maturity of the BI platform, according to Qasim Hussain, senior architect for the technology consulting firm X by 2.

Insurers need to remember it’s better to start small rather than big, adds Hussain.

“There definitely is a learning curve when it comes to new tools and a comfort change needs to happen with business and IT,” he says. “In the past [business users] may have done a lot of one-off reports, but when you start talking about a modern BI platform it gives you the flexibility to do a lot of ad hoc reporting.”

Hussain also believes it is better to start BI from a departmental perspective and try to get those processes adapted to the new technology and the modern paradigms before you start to take on a huge initiative.

“If insurance carriers take on fairly large-scale initiatives there are a lot of learning curves,” he says. “Those implementations take longer and frustrate the business [side] because they were promised something that will be useful to them within a year or so and the project often ends up taking years to build.”

DIGGING IT OUT

Hussain believes insurers need to take measured steps to approach the business intelligence problem, suggesting the core domains are the first place to start.

For example, an insurance carrier might have three policy systems for different lines of business. The goal would be to bring them all together into one system.

“You can do some trending and analysis over lines [of business],” says Hussain. “If you have that situation, though, you should realize the biggest challenge is to bring together a common definition of policy because these three systems—even though they are all policy systems— have their own way of defining a policy. The tools will not help you there. You need to know the plan for that effort and put adequate resources in place so the problem can be solved the right way.”

Defining terms like policy, defining a person, and defining other concepts are core foundations that the BI platform will be built on, so Hussain points out it is critical to get them right.

“The business doesn’t want to wait,” says Hussain. “They see [BI] as a competitive advantage and they are eager to get the modern platform in place so they can get rid of ad hoc reports. Ideally there should be one policy system so there can be one definition. In my experience, large carriers typically have multiple sources and trying to reconcile them is a challenge.”

Carriers also have large amounts of data that Lucker maintains needs to be made electronic. For example, carriers want to perform BI or predictive analytics around MVRs, but very often the companies only have that data in a paper format.

John Lucker, principal with Deloitte Consulting, believes the MVR reports are easy to capture, though, and having the files in an electronic format provides tremendous value.

Manual processes can be time consuming, particularly if you are trying to extract data from disparate systems and turn it into actionable business intelligence. That’s the situation MutualAid eXchange (MAX) found itself, according to Paul Heacock, the carrier’s CIO.

“We ran our carrier system in production and then we ran our agency system, and we put them together manually in a spreadsheet to have some sense of what our income levels were as opposed to what we projected,” says Heacock. “That was the reason we really needed to get to a data warehouse.

MAX also found itself spending too much time building its reports and didn’t have much time to analyze the data and learn what was actually going on.

“A lot of our reports were spreadsheet-based and they were complicated,” says Heacock. “They were not user friendly; they were transaction oriented. So we had to repackage them into a format that was a little easier to understand. Even then, when [business users] asked a question we couldn’t drill down.”

One of the nice features of any BI tool—including the 4sight Business Intelligence tool purchased by MAX—is it is being built from the ground up.

“We predefined the reports and the dashboards that we want,” says Heacock. “When we see a number we are curious about we can drill all the way down to the coverage level to find out what created that number. We hope to spend a lot more time analyzing numbers than creating the numbers.”

QUALITY DATA

Lucker advises insurers not to obsess on data quality, either.

“This is a big topic with regard to embarking on BI projects,” he claims. “Typically, business intelligence is not a counting exercise. The phrase you hear often is ‘good is often good enough.’”

Lucker recommends to clients that they perform an exercise in which they find data they think is dirty and then design an analysis of the data before the data is cleaned. The company then cleans the data and performs the same analysis on the data after it’s been cleaned.

“What I’ve found almost every time I’ve done this with a client is they’ve seen the cleaning exercise doesn’t significantly change the trends and pattern that they were setting out to see in the business intelligence exercise,” says Lucker.

Lucker believes there are various statistical reasons why this occurs.

“Very often the dirt in the data is random,” he says. “If you think of it like a bell curve, if there is dirty data on the right of the bell curve, statistically speaking there would be another dirty data point to the left of the bell curve. If you plotted all the dirty data around distribution you would see they average out. The statisticians on my team always say dirty data typically cancels itself out. There’s some inherent bias in that dirtiness, you can understand that but you don’t necessarily have to clean it.”

Traditionally, historical trend analysis served as the business intelligence for the insurance industry. Insurers have been doing such analysis for a good number of years and Callahan feels there are three realities to deal with.

The law of large numbers has mitigated some of the issues with data quality. When insurers do a look-back it can be less effective in making predictions because the sensitivity in predictive models is much higher with data quality.

Secondly, there is a comfort level in thinking the historical perspective is sufficient. What companies need to realize is reactive strategic actions they take will not be as competitively advantageous as being able to proactively put strategies in place.

There needs to be an awareness of importance moving forward. You are watching through the rearview mirror, but what you would like to be able to do is look through the windshield and drive to where you want to be—achieving profitability in the first pass.

“There’s a sense of comfort some companies have with their approach to analyzing the data,” says Callahan. “They are thorough. They have dashboards, scorecards, a whole department that does historical analysis, and they can become comfortable in that perspective. When they move to the next level they have to move forward with building the predictive side where they can target their next steps and strategies rather than just analyze previously taken steps and strategies.

Sticking to the 80/20 rule is important, according to Lucker, to avoid scope creep, which inevitably takes place when carriers begin these BI projects.

“I’ve been in project planning meetings and clients will say they understand the 80, but what about focusing on a small amount of the business,” says Lucker. “They end up focusing an inordinate amount of time on the 20 percent and completely lose track of the business value and the value of time in accomplishing the 80 percent sooner rather than later.”

Organizing the staff is an important initial step because companies are not always thoughtful around determining who the right people are to work on these BI or analytics projects.

“My experience is it takes a certain aptitude and a passion for BI and analytics,” says Lucker. “Not everybody is good at the mechanical aspects of working with data or working with the software tools or passionate about the level of granularity that’s required. Find out who the go-to people are and make sure they are involved from the onset.”

The initial delay in adopting a business intelligence solution can be traced to long project delays, according to Pauli.

“These things seemed to go on for years and years,” she says.

But technology providers realized if they couldn’t help the process along and help the carriers divide it up into smaller consumable segments, potential BI implementations were going to go down the tubes.

“The business side doesn’t have an appetite any longer for an 18-month or a two-year implementation,” says Pauli. “If [the business side] is part of the initial process and has some stated goals, that is when you can chunk up the deliverables.”

Heacock believes MAX’s data is in better shape than most insurance carriers because MAX has been in business for just a decade.

“One advantage we have is we have only 10 years of data,” says Heacock. “It’s been in the same legacy system, so it hasn’t gone through two or three iterations of conversion and changes in that regard.”

MAX business users have found some bad records that don’t make sense. But because the carrier is running the BI solution parallel with its legacy system, when two numbers don’t match it has been easier to determine where the problem lies.

“When we start to drill down it’s sometimes something incorrectly calculated or presented in the BI tool, but often it’s just the legacy data,” says Heacock. “It’s not anything that I feared it might be.”

THE RIGHT STUFF

One of the biggest challenges facing companies trying to move to the next level of business intelligence is what Callahan calls management bandwidth.

“The ability to develop models requires senior-level management insight, perspective, and collaboration to discuss the specifics of what they are looking at,” he says. “Management is overburdened now, so finding the management bandwidth to dedicate to a project like [BI] is difficult.”

The second major challenge for insurers is finding the expertise to understand how the different components of data inter-relate. Callahan maintains it’s an expertise in which a lot of companies don’t have much depth.

“The bandwidth issue is truly a matter of prioritization,” he says. “Companies that are moving forward in predictive analytics and business intelligence are recognizing [those tools] can provide them with competitive differentiation and they are prioritizing their efforts and allocating resources to focus on a particular area.”

The ones that have been successful with BI are ones that are picking an area to focus on. In the property/casualty area there has been a lot of focus on claims because the leverage of the claims dollar is so high. In other areas the focus may be on customer segmentation because insurers are moving into a differentiated market.

“They realize profitability varies with each customer so they are going to focus on the most profitable customers,” says Callahan.

In terms of developing the expertise from the existing staff, Callahan feels there are people within organizations that have worked with the data, particularly in the actuarial department or the information reporting department that have the necessary awareness.

“You can supplement that if you get toolset expertise that comes with the [software] installation and development of the first project, so you need knowledge transfer to your internal experts,” he says. “If you are trying to leapfrog the process you can contract the expertise if necessary to get your first project in house. The goal would be to get the knowledge transfer in place so you have your own staff develop the data awareness to take you to the next level.”

THE DIRECTION

Callahan points out the economy hasn’t been kind to major projects such as business intelligence and it’s not bouncing back enough to give insurers excess profitability where they have the luxury of exploring solutions at their leisure.

“We have the pressure of achieving sales growth and profitability in an environment where expense reduction doesn’t contribute as much as it could in the past,” he says. “[Carriers] have to find out how to be more profitable. To do that, they have to look more closely at their products, their market segments, and their risk exposures. That’s going to drive the greater use of tools to achieve profitability. They have to look at their business with a magnifying glass. They can’t cut any more expenses; they have to become smarter in the marketplace.”

Lucker maintains it’s important for carriers to be thoughtful and holistic in determining what they want to accomplish with the business intelligence project. Once they design the docket of work to determine what specific BI findings they hope to discover, it’s important to anchor that work to a cost-benefit analysis.

“Look for ways to create the best short, medium, and long term project dockets so you can provide the organization with some tangible return on investment to help the executives and the rest of the team realize value throughout a timeline,” says Lucker.

Lucker finds some companies—at the end of the two-year effort—hope they finally get something for their effort.

The project dockets enable insurers to accrue benefits throughout the course of the project and this gives them the opportunity to achieve business satisfaction and a return on investment.

The dockets also give the carrier the opportunity to alter their course, which is something many companies need to do.

“They go so far along they make [the project] bigger than they need to,” says Lucker. “If it doesn’t go as well as they had hoped there can be a tremendous emotional hurdle to overcome.”

Heacock is excited about the new dashboard capabilities available to MAX that will present a high-level summary of the data and also enable the what-if modeling.

“The ability to play with the components of the combined ratio, for instance, is important,” he says. “It’s easy to play with the losses, but what if our average premium went down 10 percent or up 10 percent and everything else remained constant? What would that do to our combined ratio?”

The modeling capabilities are something MAX intends to spend more time on once the carrier gets the more germane tasks completed, such as new production reports and exposures reports.

“We still have to get that nailed down, but we’re looking forward to the ability to model the data,” says Heacock.

Heacock reports MAX is working on some basic issues for now, but also is developing a model that patterns the sales pipeline.

“We have a goal of X written premium for next year,” he says. “There’s an element involving retention of current business and the new production. If you follow new production down the chain, how many quotes do we have to make to get that level of production? In order to get that many quotes, how many sales calls and proposals do we have to make. We’re trying to trace it all the way down to how many hits we need on our Web site. That’s been defined conceptually, but not practically.”

BPM IN PLACE

Pauli believes the business intelligence initiatives that fail are the ones that take on a life of their own from whatever group is managing the project. At some point, misdirected projects are trying to get information for the sake of information.

“You really need to have a predefined goal of what you are trying to accomplish once you get your data in the BI software and you start to get results, particularly if you are combining business process management (BPM) with your BI initiatives,” says Pauli.

Most of the technology providers in BI or BPM either started off as pure-play BI or pure-play BPM, explains Pauli. Today, though, many of them have capabilities that cover both areas.

“Once you get your business intelligence to the point you are producing results then you’ve got to be able to do something with it,” she says. “If you don’t have the BPM capability, then it’s BI for the sake of BI and that’s kind of pointless.”

Industry Specific

Pauli feels there are some really good business intelligence systems that have great technology, including some cloud solutions. Still, she believes insurance companies are looking for software providers that have an expertise in their field, so they don’t have to explain themselves, which is why she believes for most insurers the BI solution has to be somewhat industry specific.

Heacock believes there’s an element of BPM in its business intelligence efforts. Currently, MAX is looking at all business processes and trying to optimize them.

“From the top or from the bottom, we need to get those views optimized and then measured before the metrics,” says Heacock. “If I don’t know how many Web site hits that I can trace through to a sale, I can’t build a model. I need to understand what drives those Web site hits? How many unique visits do we need? There’s an element of BPM in there as well.”

MAX is trying to get the operational reports finalized and accepted. Business users are more familiar with tabular numbers, but Heacock hopes to give them graphics along with the tabular numbers.

“The real payoff is down the road where we can be more proactive rather than just reacting to the numbers,” he says. “Maybe we can figure out how we can make a more direct impact with the numbers. Our current reporting is looking through the rearview mirror. We really want to start looking through the windshield. But you can’t go forward until you know what’s behind you. Our goal is to have these things done by the end of this year. That’s an aspiration. We may not totally be where I’d like us to be, but it’s incremental. We’ll be doing some things along the way.”

Kelly Donaghy explains Pan-American Life’s first step is to work on strengthening the data around its producers. The carrier will then use business intelligence to create key performance indicators across the company.

“Our business areas are already developing metrics around sales and customer service,” says Donaghy, business technology consultant for Pan-American Life. “We look forward to bringing these pieces into the fold as the next steps.”

Pan-American Life currently is in the implementation stage with its BI solution.

“Strong senior management support and patience are needed as we align our data sources together, but I believe once we show our initial success, the adoption will be quick,” she says.

HAMPERING EFFORTS

Lucker believes BI has become universal throughout the insurance industry.

“Virtually every company would say they do it,” says Lucker. “Obviously there’s a continuum of what that means, but I do think a lot of companies are struggling to take business intelligence to a higher level. Companies are struggling in how they get enough people to understand what advanced analytics mean, what the art of the possible is.”

Lucker continue to see significant organizational inertia issues—a fear of change—in how companies can become more information driven and fact-based and less intuition driven and allowing things to happen to them by the market rather than being more forceful in the market.

“I’m seeing plenty of success stories,” says Lucker. “The way I always think about success is not everything translates to dollars—very often it does and should—so how do you articulate your ROI. That’s a very basic question, but it’s still a vexing problem for insurers in terms of how they calculate the value. You have to be comfortable with the idea you can’t always measure everything in dollars and cents.”

Pan-American Life realized that, due to multiple administration platforms and various lines of business, it did not have a holistic view of its multiple distribution channels throughout the U.S. and Latin America. For instance, if the carrier has a producer who sells in multiple regions, there was no systemic way to tie that information together. 

“We initiated a project called ‘know your producer’ to bring all relevant data together, including key performance indicators, to truly understand the overall picture,” says Donaghy. “In order to implement this project quickly, we engaged CSC to build a framework of key data points.”

Pan-American Life chose CSC not only for the vendor’s knowledge of operations, but also for their vast experience in creating metrics specific to insurance.

“CSC used data points from their insurance industry data model coupled with data points requested by our senior management,” says Donaghy. “We were able to rely on their expertise to improve our structure. Once the mart is implemented, we will be able to have a single view of our producers, regardless of the region, country or product line being sold.”

Business intelligence was one of the technology areas TowerGroup saw as having the most traction in 2010. Carriers have slowly learned that the initial hang-up—thinking they hade to have their data in perfect order before they start—is not necessarily true today.

“There’s no such thing as perfect order for data,” says Pauli. “Once [insurers] understand that not all data elements actually turn into information you can look at pulling out some of the data elements in the first initiative, get the data warehouse loaded with the important data elements that lead to decisioning, let the BI technology bring it all together and turn it into information, and then route it into the BPM software. Most of the carriers who have gotten away from BI for the sake of BI are the ones who can move forward.”  TD

Top Story

Superstorm Sandy: 2 years later

Many residents on the East Coast are still rebuilding as the insurance industry and FEMA work to pay off claims two years after Superstorm Sandy hit.

Top Story

6 ways to improve producer recruitment success rates

A new Reagan study shows that only 56% of producer hires are successful. Here are six tips to beat the odds.

More Resources

Comments

eNewsletter Sign Up

Tech Digest eNewsletter

Technology related insights for insurance professionals including key developments, solution providers and news briefs from the carrier front – FREE. Sign Up Now!

Mobile Phone
         
Close

Advertisement. Closing in 15 seconds.