Teradata Magazine Cover Teradata Magazine Online  
Register Help Password
Password:
Quick Links
Current Issue
Archives
Teradata.com
Teradata Magazine Rss Feed
ARCHIVES Search Teradata Magazine Online:  


Query? Got Questions about Teradata? Send your technical inquiry to the experts in Teradata engineering. From the architects to the developers, we'll get your question to the appropriate certified Teradata professional for resolution.
teradata.query@
teradata-ncr.com



HELP!

Is your data warehouse making you money or costing you money? By consolidating out-of-control data marts into an enterprise data warehouse, you will tame escalating cost and management headaches—and realize far greater return on investment.

IN THE WAKE OF the dot-com boom’s irrational exuberance—and in the midst of economic uncertainty and global instability—businesses are struggling to contain costs. IT managers are under increasing pressure to spend less and focus on projects that will deliver a clear return on investment (ROI). How dothese urgent priorities impact data warehousing projects?

In truth, many data warehousing projects have run amok. Few have ever been measured for their effectiveness and ROI. The time has come for IT managers to take stock: What’s the best way to deliver higher business value—and at the same time, rein in and reduce the spiraling costs of data warehousing?

For many organizations, the number one problem is data mart proliferation. Years of creating a new data mart every time a department needs to do some analysis are coming back to haunt them. According to research done by Teradata®, a division of NCR Corp., the average annual support cost per data mart is $1.5 to $2 million. The aggregate cost of all the data marts being funded independently at the department level would shock most CIOs and IT managers; they are just now coming to realize how costly data mart proliferation has become. With 59 percent of the companies in the survey maintaining up to 30 data marts—and some as many as 100—annual support costs easily dwarf the business benefits the data marts were intended to bring.

But cost is only part of the story: The real ROI killer is lost business value. When data is spread everywhere, it’s difficult to find answers to questions critical to understanding current business performance and making decisions that affect future enterprise competitiveness. Data mart proliferation leads to timewasting“corporate intramurals,” as one CIO calls it, in which different departments vie over who has the correct view of the data. Added to these difficulties is the unproductive and frustrating activity of developing and maintaining new data marts over and over again.

Is there a reasonable solution—one that doesn’t itself cost an arm, a leg, and multiple years to implement? For a growing number of organizations, the answer is data mart consolidation. It may not be the Holy Grail, but for many IT managers, it’s the next best thing. Data mart consolidation allows you to save millions and give business decision makers a more accurate, comprehensive, and timely view of their data.

COME TOGETHER
What is data mart consolidation? It involves integrating data from multiple department, division, or agency data marts into a consistent whole that is better able to answer all of an organization’s needs, rather than just a subset. Using a shared data model, a consolidation effort merges data from multiple channels and functional areas to produce a more valuable resource that defines data once for use by everyone. There’s no redundancy: Data is stored in one place only, delivering a single version of the truth. Imagine: Fewer systems. Fewer disk arrays. Lower maintenance, thanks to fewer processes and systems to support. You won’t suffer the lag time that comes with loading data into multiple locations. You will eliminate data inconsistencies. Your users will enjoy direct access to all the highquality data they need.

Data mart consolidation brings operational efficiencies that directly improve the bottom line. But even more important is the impact on overall business success. Many of today’s leading organizations credit their success to the development of an integrated, enterprise-level data warehouse. Some were fortunate to have begun with that objective but others, frustrated by data-mart architectures that did not allow them to achieve business benefits, have migrated. Here are some examples:

Harrah’s Entertainment. Harrah’s has saved over $20 million per year in overall costs since implementing its Total Rewards™ customer relationship program, which is based on Teradata database technology. The enterprise data warehouse has given Harrah’s a much clearer picture of its customers, enabling the company to create targeted marketing campaigns that have nearly doubled the profit per customer. In the first year, same-store sales have grown 14 percent over the $242 million the company achieved in 1998. Even more impressive: The data warehouse enabled Harrah’s to craft incentives that encourage select customers to visit Harrah’s properties nationwide, leading to 72 percent growth in the number of customers who play at more than one Harrah’s property.

Travelocity.com. With an enterprise data warehouse, Travelocity improved gross bookings and earnings by 100 percent in 2000. Gross profit margins improved 150 percent, and booker conversion rates rose to 8.9 percent, the highest in the online travel services industry. After TWA announced a special, $360 round-trip air fare between Los Angeles and San Juan, Puerto Rico, Travelocity used its Teradata enterprise data warehouse to identify customers most likely to take advantage of the offer. Approximately 25 percent of Travelocity’s targeted customers either accepted the offer or booked a different flight to the Caribbean.

Texas Comptroller of Public Accounts. The enterprise warehouse gave the organization intelligence that led to the collection of $94 million in unpaid or underpaid taxes. Running on Teradata, the warehouse enables the Comptroller to uncover an average of $1 million per week in additional revenue.

Anthem Blue Cross Blue Shield. A centralized, enterprise data warehouse helped Anthem achieve the ultimate in customer service: saving lives. Using Teradata for trend analysis, Anthem conducted a study to identify “best in class” coronary bypass surgeons who had a patient mortality rate lower than 1 percent, compared with the industry average of 4 percent. Armed with information, Anthem was able to route its policy holders to the most skilled and successful surgeons.

Lucy in the Sky with Diamonds
Brent Lautenschlegar, one of Delta Technologies’ enterprise data warehouse consolidation project leaders and now principal consultant with Reflection Technology, says that Delta Air Lines consolidated 27 standalone data marts into a centralized, integrated enterprise data warehouse based on Teradata. Previously, analysts struggled to gather data manually from multiple, disparate sources just to get answers to common business questions. A good example is flight profitability. Analysts tackling this business problem required data from three key subject areas: Operational Statistics, Ticket Revenue, and Cost. Making things even more complicated, the Cost subject area was itself pulling data from upwards of eight different sources. The process often took weeks, with numerous analysts gathering, cleansing, and analyzing data from each of these sources to determine profitability by flight for the previous month. Today, Delta can analyze flight profitability daily. Delta can populate its enterprise data warehouse nightly and make it available for access by all users. “Delta was able to get answers to important questions that it just couldn’t do before in a timely manner,” said Lautenschlegar. “The monthly ROI for flight profitability alone is measured in millions—and this is just one of dozens of areas with similar results.”

CAN YOU GET THERE FROM HERE? Sounds good. Looks difficult. How do you get there from here? Teradata offers a Data Mart Consolidation program that begins with a Business Impact Model (BIM) assessment, which uses your data to calculate what you’re currently spending and demonstrate what and where the cost savings will be once you’ve consolidated. The patent-pending process identifies data marts that are candidates for consolidation and typically recommends a phased approach over a period of months. To eliminate guesswork and minimize risk, it measures your net cost savings, after the cost of installing and migrating to the consolidated Teradata system, in terms of Internal Rate of Return (IRR), Net Present Value (NPV), and estimated payback period.

The program is delivering visible results for Teradata’s customers. After going through the process, customers are surprised to discover:

• How much they’re spending today

• The cost savings they can realize in a relatively short period

• The business value of having an integrated, enterprise data warehouse.

Even after assessing the projected costs of consolidation itself, the savings remain so significant that customers often accelerate the consolidation process. They want to begin getting expenses in line as soon as possible. And users are anxious to gain the ability to find answers to enterprise- level questions that they previously couldn’t even ask.

That’s because Teradata can do for data warehouses what competitive database products cannot: store data at the detail level and still provide industryleading performance. With this powerful combination, your users can drill down to get a real picture of their business. They can do predictive analysis based on real data—a critical advantage when it comes to reducing attrition, detecting fraud, or increasing customer satisfaction through better customer knowledge.

TICKET TO RIDE
Teradata is uniquely positioned to support the consolidated environment. Other relational DBMS products, built to support online transaction processing (OLTP) applications, require a lot of bending and shaping to handle decision support. From day one, Teradata database technology has had a single focus: to support complex and demanding data warehousing workloads at an enterprise level. As they grow in strategic value, data warehouses have to support rapidly rising data volumes and increasing numbers of concurrent users, who ask demanding questions, both planned and ad hoc. A database that requires excessive indexes, optimizer hints, and modification of initialization parameters just to handle “known” queries can’t support the unpredictable nature of an enterprise data warehouse.

Attempting to keep up with the workload and maintain acceptable performance, other DBMS products use a variety of workarounds:

• Constraining queries to a defined set of parameters, even though this limits the variety of questions users can ask;

• Forcing users to work with summary tables instead of detail-level data, even though this compromises the accuracy of analysis;

• Limiting data warehouse update frequency, even though this means forcing users to work with data that is out of date.

From the point of view of business users, these workarounds represent major concessions, preventing them from gaining the full ROI from their data warehouse environment. Databases that require IT managers to tune queries, frequently reorganize data, and apply parallelism only intermittently—instead of on all functions and queries—simply don’t have the flexibility to deliver high-value, timely analysis. Isn’t that what you wanted your data warehouse to provide in the first place?

Other DBMS vendors claim to support the tough demands of enterprise data warehouses, but have trouble coming up with real-world examples. Teradata is proven: For over 20 years, Teradata has been the foundation of real-world, production systems, ranging in size from very small to hundreds of terabytes and thousands of users. With Teradata, there is no diminished rate of return once the system’s size and complexity grows.

BRITISH AIRWAYS: All Together Now
British Airways wanted a single view across the airline of each transaction with the customer. This will lead to improved customer service, increased revenue and reduced costs, and help increase returns from yield management. It will also enable British Airways to design and implement higher impact marketing initiatives. Prior to this, the company stored its entire customer and commercial data on multiple, disparate databases supplied by Teradata, Oracle and IBM.

REVOLUTION
One Teradata customer uses a newly consolidated data warehouse to reduce losses on perishable goods through more accurate demand forecasting. Another customer, battling competitors in a newly deregulated market, has been able to minimize attrition to near zero, thanks to far more complete knowledge of customers. And in one dramatic case, a Teradata customer consolidated 22 of 76 data marts over a period of eight months, delivering actual cost savings of over $6 million per quarter. Due to the deliberately conservative design of the BIM, customers’ actual savings are often greater than the initial estimate.

With news of such success spreading quickly, consolidation has become a major trend in data warehousing. “In the last 18 months, I have spoken to over 100 companies,” said Allen Messerli, principal of Allen Messerli Enterprise Systems. As CIO of 3M Corp., Messerli directed a massive consolidation project. Now, he spends his days guiding other large organizations through consolidation. “It’s tough to keep up with the demand. The credibility of Teradata’s Data Mart Consolidation program is super.”

For many organizations, the consolidation process is an opportunity to correct other shortcomings with their firstgeneration data warehouse and data mart environments. A single, integrated data warehouse allows them to finally establish measurable, business-driven objectives—and develop business-level confidence in a platform that can meet not only present needs, but also future expansion.

Following its own internal review, British Airways recognized the importance of a single, consistent enterprise-wide data source to improve its commercial decision-making through a complete knowledge of its customers’ interaction with British Airways, both in buying and in service delivery. As a result, they are consolidating the airline’s customer and commercial data onto a single enterprisewide Teradata Warehouse.

American Eagle Outfitters recently announced plans to implement its first enterprise data warehouse after consolidating multiple databases running with IBM DB2 and Oracle. Now, with a single system based on Teradata, the company will finally enjoy a comprehensive, integrated view of its business. “By consolidating our information with this technology, we can better manage our large amounts of data,” said Ken Watts, American Eagle’s CIO and vice president of Information Services. “This data warehouse will be a key component in supporting our future growth.”

sunrise: Here Comes the Sun
TDC Switzerland (sunrise), Switzerland’s second largest telecommunications provider following the early 2001 merger of diAx and sunrise, has announced that it will replace its existing Oracle data marts with a single Teradata database. The new architecture will allow sunrise to consolidate multiple, costly data marts into one Teradata system. Decision makers across the enterprise will be able to use and analyze transaction data from various business units for the first time.

TERADATA: PROOF POSITIVE
In response to strong financial as well as competitive pressures, pencils have gotten sharper. Savvy buyers are more thorough and careful than ever as they assess potential platforms. Benchmarks using their own applications and data; detailed total cost of ownership assessments; proof required of vendors’ statements; and lots of good references: These kinds of requirements are now the rule rather than the exception.

For example, Highmark Blue Cross Blue Shield recently migrated its fiveyear- old, DB2-based data warehouse to Teradata—a process that the company accomplished in 90 days. Highmark decided to migrate after reviewing the results of an internal head-to-head, full-warehouse benchmark that revealed Teradata’s 300- to 700-percent advantage in query processing performance over the dedicated DB2 environment, according to Darren Macioce, vice president of Healthcare Informatics at Highmark. The new Teradata-based system will support numerous strategic applications, including data mining, advanced analytics, measurement of physician and hospital quality of care, and mandatory reporting for accreditation. Consultant Messerli says that for most clients, the decision to migrate from other database environments to Teradata is an easy one. “You can’t do an enterprise data warehouse on any other platform.”

WE CAN WORK IT OUT
Let a Teradata data warehousing expert help you quantify your company’s potential data warehouse cost savings and business benefits with a FREE consultation and ROI report.

*With apologies to The Beatles

©2002 NCR Corporation, Dayton, Ohio U.S.A. Produced in U.S.A. , All rights reserved.

Teradata and NCR are registered trademarks of NCR Corporation. NCR continually improves products as new technologies and components become available. NCR, therefore, reserves the right to change specifications without prior notice. All features, functions, and operations described herein may not be marketed in all parts of the world. Consult your Teradata representative for further information.

British Airways is a registered trademark of British Airways.
Delta Air Lines is a registered trademark of Delta Air Lines , Inc. sunrise is a registered trademark of TDC Switzerland AG.
Harrah’s is a registered trademark of Harrah’s Entertainment, Inc.
Total Rewards is a registered trademark of Harrah’s Entertainment, Inc.

www.teradata.com

PDF file of this story




Copyright by Teradata Corporation 2001-2007.