 |
|
 |
|
 |
|
|
|
| Current Projects | |
|
|
| Press Releases | |
|
|
 |
News About Daman |
|
|
| Articles & White Papers | |
|
|
 |
|
 |
|
|
|
Transactional Data Store
Data warehouses are evolving all the time.
Sometimes they evolve into hybrids. A look at one such hybrid: Daman's first-to-market "transactional data
store."
DCI Newsletter
TUESDAY, JUNE 22, 1999: Austin, Texas-based Daman Consulting,
a provider of data warehouse and enterprise application integration solutions, recently announced the successful
development and deployment of a "transactional data store" for National City Corporation, a financial
services provider.
The TDS, as developed by Daman with the assistance of Compaq Computer consultants and National City's own I.S.
personnel, is a specialized database combining decision support capabilities with data more timely than the data
usually found in data warehouses.
It's also one more step away from viewing data warehouses primarily as historic stores, and more as dynamic, up-to-the-minute
analytical tools. "The data warehouse as the store of historic data has its limitations," said Philip
Russom, director of the data warehousing and business intelligence service at Hurwitz Group. "There is a need
for transactional data warehouses that complement the historic warehouse by capturing up-to-the-minute customer
behavior for analysis."
The TDS is described as being an architectural compromise between an operational data store and a historical repository,
like a data warehouse. "Many of our retail business areas" will benefit from the TDS information, says
Greg Komar, vice president of technology and planning for National City Corp.
Komar says the company decided to forego "the usual ODS solution," believing that costs and solution
delivery time would have been five times greater than what the TDS required.
The TDS works in conjunction with an enterprise data warehouse solution. Most data warehouses use historical data
that is several months or years old, analyzing it to reveal patterns and predict developments. The TDS, in contrast,
supplies daily transaction data aggregated on a monthly basis. This means that "who the customer is"
can be stored in the data warehouse, while "what the customer does" can be captured by the TDS. Together,
the EDW and TDS offer a more complete picture of customer information, behavior and purchase propensities.
Because this data picture includes current, timely information, users can better understand complex retail financial
markets and refine their services according to individual households. Users can also better assess the profitability
of operations and therefore reduce overhead or infrastructure redesign without costly changes to their operational
systems.
Early results from NCC report that processing performance has been higher than predicted. The process of producing
daily aggregations of data has been reduced from 24 hours - the time required by an ODS -- to three. This performance
boost has reduced overhead considerably and freed system resources for other tasks.
Plus, because transaction data now comes from the TDS instead of the operational systems, there is no need for
operational reports. This means that the TDS or data warehouse can be changed without impacting operational systems.
