| October 20, 1997 |
OLAP SpreadsVendors are integrating online analytical processing with enterprise ap
plications
By
John Foley
Another sure sign that OLAP is becoming mainstream: Microsoft is entering the market.
OLAP products arrange data in multidimensional hierarchies-or cubes-of interrelated subjects, rather than in the two-dimensional tables of relational databases. Analysts say this way of organizing data relationships is more intuitive for users. It also
yields subsecond answers to some calculations.
Blue Cross and Blue Shield of Rhode Island, for example, uses Arbor's Essbase server to analyze financial proposals for its 1998 budget-analysis that previously was done on a mainframe. "What used to take days can be turned around in an hour," says George Trudel, a consultant with the health insurer's business and technology office in Providence, R.I.
KFC Corp. uses Informix's MetaCube OLAP suite for strategic planning and sales and marketing, including the tracking of product introductions and the effectiveness of ad campaigns. Employees "are anxiously using the system to find out how sales are going," says Tom Rosing, manager of decision support with KFC's restaurant support center in Louisville, Ky. "Once users get a taste for this very granular data, it's not a push for IT. It's more of a user pull."
Like many other organizations, KFC and Blue Cross and Blue Shield of Rhode Island are deploying more OLAP servers and making them available to new applic
ations and more users. Fueled by this trend, the OLAP market is expected to grow 40% this year to $1.4 billion, according to Richard Creeth, a consultant and author of
The OLAP Report
.
Vendors Rush In
Just last week, Corel Corp. announced that it, too, is developing an OLAP tool, called Corel Resero. An early user of the product, Otis Elevator Co. in Farmington, Conn., has purchased 1,500 licenses to access and analyze financial systems in offices worldwide. "It's very natural to do reporting on OLAP," says Omar Fathi, director
of IS at Otis. "It presents data in a way that users understand."
Essbase 5 also supports what are called dynamic calculations, which make it possible to draw data relationships at query time, rather than in advance. Ralph Hudack, a business systems developer for Xerox Corp.'s U.S. operations in Rochester, N.Y., says this feature alone should help Xerox cut the time it takes each month to load and precalculate its Essbase server-which includes more than 10 million records a month-from 26 hours to about 10.
Oracle next month will ship a sales and marketing data mart that bu
ndles analysis applications with its ROLAP (relational OLAP) option, which extends the functionality of Oracle's Express OLAP server to data stored in a relational database. In December, Oracle is slated to deliver an upgrade to Oracle Applications Data Warehouse, software designed to populate data warehouses with data from Oracle's packaged applications. Version 1.1 comes with improvements in data modeling and filtering, the ability to schedule data extraction and loading, and certification for Oracle8. More than 80% of companies using the product have included Oracle Express in the warehouse environment, says Jesus Castillo, director of product marketing with Oracle.
Independent software vendors see an opportunity in OLAP. Baan, for instance, is linking its enterprise applications to Hyperion Software's data-analysis products, including Hyperion OLAP. On Oct. 7, i2 Technologies Inc. announced plans to integrate its supply-chain management applications with Arbor's Essbase. PeopleSoft is integrating the m
etadata layer of PeopleSoft 7 with Essbase. "There is now widespread recognition that a new class of applications, analytical applications, require a different platform from transactional applications," says Jim Dorrian, chairman and CEO of Arbor. "Now, it's a question of who the winners will be."
Sybase and IBM want in on the action. Sybase holds a minority stake in WhiteLight, which will ship OLAP software for relational databases this quarter. IBM's DB2 OLAP Server was to ship this quarter but has slipped to the first quarter of next year, the company says.
An established supplier, MicroStrategy Inc. in Vienna, Va., next week will announce DSS Objects 5.0, an upgrade to its OLAP programming interface that includes new drivers for major databases and the ability to build distributed data marts.
But all eyes are on Microsoft, which a year ago acquired its OLAP technology from Panorama Software Systems in Tel Aviv, Israel. In September, Microsoft announced the beta release of OLE DB for OLAP, a softwa
re specification for sharing data among different vendors' OLAP products. More than a dozen companies announced their support.Microsoft is expected to release its OLAP server, code-named Plato, next year. Plato will run on Microsoft's Windows NT-based SQL Server database, also due in 1998. Analysts expect Plato to lower the cost of OLAP, making it available to an even broader base of users and changing the market's dynamics.
--with additional reporting by
Karen M. Carrillo
|
This Week's Issue
Technology Whitepapers
- Mobile BI: Actionable Intelligence for the Agile Enterprise
- Creating the Enterprise-Class Tablet Environment - by Yankee Group
- How To Regain IT Control In An Increasingly Mobile World - by BlackBerry
- Red Alert: Why Tablet Security Matters - by BlackBerry
- New Visual and Wizard-Driven Paradigms for Exploring Data and Developing Analytic Workflows
nline analytical processing is breaking out of its niche. Used mainly by power users in financial departments for intense number-crunching, the data-analysis technology is set to become widely available. Arbor Software, IBM, Oracle, and others are readying products that are better suited for enterprise deployment, and independent software vendors are integrating OLAP engines with enterprise applications.
At Arbor's annual user conference in Orlando, Fla., next week, the company will announce Essbase 5, an upgrade to its OLAP server. Major improvements include the ability to partition applications among multiple OLAP servers for greater scalability and to run queries against data stored both in Arbor's multidimensional engine and in a relational database.











