CATEGORY: Business Intelligence

REVIEWER: Marty Ellingsworth, director of operations research, customer research and strategies for the Fireman's Fund.

BACKGROUND: The 140-year-old Fireman's Fund, based in Novato, California, holds an "A" rating from the A.M. Best Company. In 2002, gross premiums written topped $4.1 billion. Fireman's Fund is a subsidiary of Allianz AG of Munich, one of the world's largest financial services firms with almost $1 trillion in assets under management.

PLATFORMS: Windows NT. Some data services are also in AIX, and we have mainframe SAS applications as well.

PROBLEM SOLVED: Fireman's Fund competes for a share of the U.S. property and casualty market in a dozen lines of business, many heavily regulated at state and federal levels. To succeed, we need a clear understanding of what makes us profitable in order to make intelligent and timely decisions that are based on good information, allowing us to consistently meet customer expectations. We use SAS for virtually all financial and operational reporting. The operations research group, in particular, uses SAS to fuse data from all corners of the enterprise, creating intelligence to improve profits and customer satisfaction.

PRODUCT FUNCTIONALITY: SAS enables us to aggregate data from source applications and various data warehouses on a "sandbox" server for data fusion and analysis. The implementation architecture depends on situational issues unique to each project; however, models are typically piloted to provide real-world results that business sponsors can appreciate. Sometimes we hand SAS models over to users who are familiar with SAS but do not have statistical modeling skills. Other times, we maintain and run the scoring code, distributing results periodically. Future plans include using SAS to support both price policies and studies run by our actuarial groups.

STRENGTHS: SAS Enterprise Miner is extremely intuitive. Its graphical features are helpful for non-statisticians and shorten model-building time. With SAS, we can build models 20 to 30 percent faster than was possible before implementing this solution. With SAS Enterprise Miner, it's easy to store projects to pick up later ­ even for a different programmer. The result is better use of staff resources, less duplication of efforts and significant time savings. Whenever we need to refresh or tweak a project, another analyst can pick it up and understand it exactly. Underwriters and adjusters echo these sentiments. They are able to focus their activities with greater accuracy because we can quickly build statistical models that have highly relevant and useful predictions.

WEAKNESSES: SAS' structure of licensing its products separately may limit the use of alternative approaches for small shops like mine.

SELECTION CRITERIA: We focused on vendors' ability to use technology, particularly statistics, to uncover answers to help drive our business decisions. In SAS, we found a programming environment where information can be merged into data sets to which analytics can be applied, enabling us to discover emerging trends and issues that could affect our business. Additionally, we found SAS to be the most capable of handling data integration issues arising from complex project interactions.

DELIVERABLES: With the models created by SAS Enterprise Miner, we can fully leverage the information stored in our data. We've been able to note suspicious correlations and flag anomalies for follow-up by specialized investigators, helping to root-out fraud and increase employee productivity. We've also used SAS to uncover premium fraud and identify cases that are candidates for subrogation (claims we service and pay, but for which others are ultimately responsible), reducing our claims liability. SAS models help us rank our customers so we can better meet their unique needs. For instance, we can move higher-claims customers into different programs while simultaneously focusing on retaining lower-claims clients.

VENDOR SUPPORT: SAS is a very friendly, user-focused company. SAS technical support has always been there to provide answers to procedural and technical questions. In several instances, they've even supplied cutting-edge macros that are being prepared for release in future versions of SAS. We have also used SAS training to bring our staff up to speed on the product, advance their particular competencies, help them build a real-world knowledge base and network with other experts in the field. SAS trainers are extremely knowledgeable in both theory and real-world applications, making the courses excellent investments for us.

DOCUMENTATION: I have used SAS for so long that it is difficult to differentiate classroom and on-the-job learning from information gleaned from documentation; however, when we have had to turn to documentation, the pertinent information has been easy to find and digest. We also use SAS-l and (SAS Consultant Special Interest Group) to keep us up to date and solve some pointed issues.

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