Seattle-based PEMCO Corporation is one of seven independent companies that form PEMCO Financial Services which provides one-stop insurance, banking and brokerage services. But because PEMCO Corporation is a business-to-business transaction processing clearing house for the umbrella company, its customers are all internal. Through PEMCO Corporation, the parent company shares common functions such as data processing, inventory and corporate accounting services.

PEMCO's Challenges

"Two years ago, PEMCO was faced with two simultaneous challenges," according to programmer analyst Dianna J. Brinkley. "We wanted to build a Customer Information System (CIS) with a tool that would use our data as well as perform both traditional and non-traditional scrubbing--in other words, not just names and addresses. The CIS would function as a hub of various legacy operational systems. We also wanted to build an information or data warehouse, but our data of several million records was all in free-form fields--not heavily codified or edited."

The unformatted records contain information about the insurance policy and the customer such as vehicle identification numbers (VIN) and miscellaneous discounts. Unlike typical names and addresses, this type of data requires sophisticated data re-engineering tools that go beyond scrubbing and traditional cleansing.

Why Vality?

PEMCO researched a variety of software tools in the market and chose Vality Technology's INTEGRITY Data Re-engineering Environment to build the CIS and the information warehouse. "Most of the tools on the market were not comprehensive enough. We are mainframe based and the data needed to be accessed with query tools. Vality's INTEGRITY was the only tool in the market that was robust enough for our volume and type of data. Other tools were limited to scrubbing names and addresses," continued Brinkley. "For example, when PEMCO customers correct or change VIN numbers, it may appear as if the customer has two different vehicles; yet it's the same car. This bad or inconsistent data contributes to incidences of internal errors," explained Brinkley. PEMCO subjected the multiple representations of data to INTEGRITY which corrected, standardized, reformatted and recognized the duplicate and inconsistent policy holder data.

The Vality tool has also been used for additional applications at PEMCO where the information was in free-form fields and the staff needed to search for key words of insurance policy riders (e.g., jewelry and camera equipment). PEMCO also deployed the INTEGRITY tool for the homeowner and renter policies.

While the CIS is about a year away from completion as additional processing is still in progress, the information warehouse has been implemented; and it is being used for a variety of tasks.

Approach to Cleansing Data

The four-pronged Vality methodology first looked at the condition of the data through a data investigation process that includes meta data mining and data parsing. This phase helped PEMCO to identify and correct data so as not to corrupt the new system. In the second phase, called data transformation, the data (which can come from multiple sources) is conditioned and standardized. The third phase focuses on data integration/matching which ensures that data is consolidated according to PEMCO's individual business rules and culminates in INTEGRITY's final phase involving data survivorship.

Cost Benefit Analysis

As a senior decision support analyst at PEMCO, Donald M. Thompson's job is to answer business questions posed by management. Since the information warehouse was implemented, Thompson has judiciously analyzed the return on investment of the Vality solution in combination with the information warehouse. So far, he's catalogued at least 20 instances where PEMCO can show an unsurpassed return on investment. From this list three examples of high data quality are worth noting. PEMCO was able to:

  • Identify supplemental dollars paid on closed claims where another insurance company's insured was at fault, saving approximately $1 million every three years.
  • Realize annual savings of $600,000 from volume discounts it obtained by entering into agreements with glass and window repair shops most frequented by PEMCO customers.
  • Using the more detailed and accurate customer and vehicle data contained in their information warehouse, PEMCO was able to segregate a significant population of insureds more likely to have adverse claims experience. This group was paying $3 million dollars less in premiums than PEMCO was paying to cover their claims.

The results of these types of studies and many others have enabled PEMCO to reduce the frequency and magnitude of rate increases and determine more equitable rates for those they insure.

Implementing the Information Warehouse

PEMCO's information warehouse operates on an MVS platform employing a DB2 database; it was originally deployed on an IBM ES/ 9000 mainframe using Iceberg Storage Subsystem. INTEGRITY operates on MVS-based platforms in addition to UNIX, Windows NT and AS/400-based systems.

Departmental analysts at PEMCO use the desktop information software tool BrioQuery which accesses the DB2 tables that reside on the mainframe. "It gives it a nice front end to do the less complex analyses. The vast majority of data warehouses are too big and the more complex analytical work is done by the analysis staff," explained Brinkley. "It provides access at the actual analyst level. They don't need as much technical expertise," added Thompson.

"The implemented software has greatly reduced the need for programming sources," said Thompson. "Infor-mation requests used to take programmers weeks, if not months, to answer. Information had to be extracted, coded and tested before it could be produced. And requests for one more report would get to the bottom of a programmer's pile," he concluded.

If Thompson could change anything about the information warehouse it would be its operational system. "Our information warehouse (IW) and our policy processing system use the same computer. I'd rather not have to fight for that CPU," adds Thompson.

Brinkley prefers to call PEMCO's data warehouse an "information warehouse," because business logic makes data more valuable. So valuable, in fact, that PEMCO's actuarial staff and statistical department conduct profitability analyses that could not be done before. Underwriters can also identify areas of concern such as rising accident frequencies and concentration of risks in small geographic areas.

"With the information warehouse we can ask questions and conduct data comparisons we never could before, such as compare property losses of under $50,000 with losses of more than $50,000. We can get answers in hours and often in minutes," explained Brinkley.

"Our new system doesn't just answer questions, it makes us ask new ones. It builds a train of thought and momentum. When you have an information warehouse, the synergy of the thought process is captured and capitalized, not lost," concluded Brinkley.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access