CATEGORY: Data Quality, Profiling & Augmentation
REVIEWER: Michael Sykes, manager of data warehousing for ALSTOM Power.
BACKGROUND: ALSTOM is the global specialist in energy and transport infrastructure. ALSTOM serves the energy market through its activities in power generation, power transmission and distribution, and transport through its activities in rail and marine. ALSTOM has annual sales in excess of $20 billion and employs 120,000 people in more than 70 countries.
PLATFORM: Windows 2000.
PROBLEM SOLVED: ALSTOM Power has multiple legacy systems as well as business transaction systems that contain customer information. This information is not standardized and contains duplicate data within and across these systems - creating a variety of reporting problems. Reporting accurate customer revenues has been a significant challenge that often required more time to gather data than analyze information. ALSTOM's strategy of "One Face to the Customer" drove our data warehouse team to develop a concept of "Information Brokers." These brokers provide a single representation of each attribute of the broker (in this case the customer) as well as a cross-reference to each source record. The goal of these brokers is not to create new information, but to allow our business associates to choose how they want each broker attribute represented. For example, they can choose to have the customer name come from one ERP system, while the street address comes from another system. Our first iteration of our Customer Information Broker utilized approximately 25,000 customer records from four business systems. The challenges were to: standardize each customer record while maintaining the original value, group similar customer entries to remove duplication and define parent-child relationships so that we could report on any customer hierarchy level (such as holding company). With dfPower Studio from DataFlux, we have successfully created a predictable and repeatable process for generating multiple matching scenarios. By establishing a priority for these matching scenarios (i.e., low tolerance to relaxed tolerance), we are able to report customer revenues in the 90- to 95-percent accuracy range with minimal business associate participation.
PRODUCT FUNCTIONALITY: dfPower Studio has performed beyond our expectations and is an integral component of our Information Broker infrastructure. Although dfPower Studio is a standalone product, it is extremely flexible and easy to use. The software has a very intuitive GUI and utilizes its own data transformation and movement engine. In addition, dfPower Studio allows us to customize the underlying standardization and matching logic. Another key functionality is the ability to evaluate information based on multiple "and/or" conditions. Without this feature, I'm confident that our results would have fallen well short of our goals.
STRENGTHS: dfPower Studio provides a user-friendly interface that makes it easy to use for both beginning and experienced users. Its layout is process- centric, making job creation a step-by-step activity. Processes can overwrite or append information to the source data. The product has also reduced our cost and usage for third-party services. D&B data, for example, is now used for support information rather than driving the definition of our customer entities.
WEAKNESSES: The match code process generates a textual string which, if used for multiple field evaluation, can degrade grouping performance. The only other weakness is the limited set of prebuilt standardization rules.
SELECTION CRITERIA: We reviewed a number of vendors during our evaluation process. We chose DataFlux for several reasons including a positive price/benefit ratio as well as the scalability and flexibility of the product architecture. In addition, features such as data profiling, USPS address verification and "out-of-the-box" rules logic expedited our implementation. The transportable rules logic allows us to carry our rules to other dfPower Studio installations. Another differentiator was the DataFlux organization. They are personable and approachable and do not tout "overkill" features with prices to match.
DELIVERABLES: dfPower Studio allows ALSTOM to view customer information with significantly more accuracy than before. This translates into real business benefits for ALSTOM, our customers and our vendors. With DataFlux's guidance, our approach was simple, manageable and low-risk. By incorporating dfPower Studio into our Information Broker infrastructure, we are able to build and manage modular processes. This structure helps us add and modify standardization or matching criteria quickly and easily, significantly reducing the costs of future development.
VENDOR SUPPORT: Support for the product has been good. We have been able to resolve all issues over the phone through the support line. Our account representative keeps us current in the product evolution.
DOCUMENTATION: The documentation is clear and adequate for our needs.
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
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access