The big dig – data-warehousing case study of EuroTunnel – Company Operations
How EuroTunnel blazed a shortcut to data warehousing.
As a company that transports thousands of vehicles and passengers through the Channel Tunnel each day, EuroTunnel knows the importance of keeping traffic moving at a rapid pace. So, when EuroTunnel’s computer systems were faced with an information roadblock, its information services (IS) department knew the fastest shortcut: the development of a data warehouse using an extraction, cleansing, transformation, and loading (ECTL) tool.
TRAFFIC TIE-UPS: AN OVERBURDENED DATABASE
Initially, EuroTunnel, based in the U.K. and France, only employed a Sequent-based booking system called Travellog for its Le Shuttle ticketing and marketing operations. As tunnel usage grew, Le Shuttle’s sales staff was processing over 5,000 individual ticket and account sales each day.
At the same time, EuroTunnel’s marketing department was accessing the system to generate reports to analyze promotions and forecast sales. The result was data gridlock. As large marketing inquiries were running, short on-line transactions–such as those for ticket sales–were often delayed.
“What should have been a two- or three-minute customer call ended up as four or five minutes,” explains Frank Barber, EuroTunnel’s technical services manager. “Customers did not wish to wait. In addition, some marketing people wanted to integrate Travellog booking information with revenue data from our Oracle systems.”
Consequently, EuroTunnel’s IS group decided to develop a database specifically for marketing reports. To do so, information would have to be transformed and migrated from the Sequent-hosted Travellog system to a new departmental data warehouse.
THE COST-EFFICIENT ROUTE
Barber and his team spoke with several data warehouse vendors but found their recommendations to be expensive and complicated.
“With one company we spoke to, data warehousing meant spending millions of dollars. We stopped talking to them at a very early stage and began looking at doing the work in-house,” says Barber.
When EuroTunnel IS was about to begin months of complex, low-level development, consultants informed them about Ardent Software’s DataStage. This tool lets users extract, cleanse, transform, and integrate data from a variety of data sources to a wide range of target databases, simplifying and accelerating the most critical stages of data-warehouse development.
Initiated in November 1996, EuroTunnel’s DataStage project involved migrating over half a million records from the Travellog OLTP system and Oracle databases to a new Alpha-based Oracle data warehouse. EuroTunnel’s IS team was now able to more easily and efficiently cleanse, transform, and manage the data.
EuroTunnel’s data administration manager, Graham Barnes says: “Despite the hidden complexities in our data, DataStage performed well.”
In just over 12 weeks, the Le Shuttle data warehouse was up and running. Today, DataStage delivers information from the OLTP system to the data warehouse via unattended, overnight operation. When the marketing analysts arrive in the morning, Barber notes, “It’s all there for them.”
Already, the benefits of the project are emerging. Marketing is using the data warehouse to target specific market regions and manage “yields” (i.e., forecasting full-price ticket sales to determine the need for discount fares).
Equally important, claims Barber, is the elimination of conflicting interpretations of data. “By having a single data warehouse, we can enforce the use of only one set of data. If you want `sales,’ everyone’s concept of `sales’ is the same. This consistent reporting is very important.”
THE JOURNEY AHEAD
With reservation data in place, EuroTunnel’s IS team is working on bringing revenue and account information into the warehouse. In addition, other departments have begun requesting their own data warehouses because, says Barber, “They’re beginning to realize this is something worth having.”
In fact, Barber’s vision is to establish a number of diverse data marts over the next two years, eventually uniting them as a single large data warehouse–with extraction and transformation tools continuing to play a pivotal role.
“In any project, the more variables you have, the more difficult it is to produce a good plan,” Barber concludes. “A product like DataStage removes some of the variables. It does things in a well-managed way. And it works. DataStage has allowed us to actually move into data warehousing without a huge investment.”
Circle 250 for more information from Ardent Software, Inc.
COPYRIGHT 1999 Nelson Publishing
COPYRIGHT 2004 Gale Group