December 15, 1998, Volume 1 - Number 4
The Twelve Most Influential Companies in IT
The Dozen
By David Stodder
Bring to bear the full measure of a diverse knowledge and talent base; fuse the best of information technology (IT) with creative business brainpower. These are urgent objectives that will drive competitive organizations into the next millennium. Never have businesses had better tools to meet traditional goals of keeping customers, driving down costs, and expanding markets. In the past, only the wealthiest organizations could afford powerful technology. With the Internet as catalyst, we are entering an era in which a far broader base of businesses can afford a formidable IT arsenal.
“Better, faster, cheaper,” the mantra of technology vendors, still drives product development and marketing. But now a new word must be added: integrated. This year’s Intelligent Enterprise Dozen—and our choice for number one—offers testament to the importance of application integration. Today, vendors find that customers have the parts; now they want the whole. Best-of-breed products must adhere to industry standards; packages must work together; and productivity tools, such as decision support, must become closed-loop solutions.
To become the intelligent enterprise, organizations need a hub—a platform—around which the universe of technology comes into focus. Hence it comes as no surprise that we are in the middle of the biggest platform war the industry has ever seen. Should it be the database, the operating system, or federated knowledge management architecture? The front office suite or the ERP supply chain? Should the scalable enterprise begin with the microchip and extend to the server, or should it rely on a freewheeling, cross-platform, plug-and-play model?
Welcome to our annual editorial choice of the most influential companies in IT. Along with the top 12, we identify an additional dozen “up-and-comers” and a selection of vendors in key categories. Taken together, these companies set the pace in 1998 and are poised to lead in 1999.
1 SAP AG
Walldorf, Germany
In 1998, SAP thrust forward in a bold attempt to close the loop. What was once an application “package” exploded into an ERP suite with vertical variations. SAP’s Business Information Warehouse (BW) transformed basic reporting features into a serious offering. And with its Supply Chain Optimization, Planning, and Execution (SCOPE) foundation in place, SAP is ready to link the chains. This is the information supply chain. The real deal, in real time.
Although founded in 1972, SAP received little notice until 1992 when it introduced R/3, a client/server version of its mainframe R/2 product designed to automate and integrate business processes. The company has not looked back. ERP industry growth has slowed, but SAP’s revenues haven’t. This year’s product introductions—along with deft application of fear, uncertainty, and doubt (FUD)—have positioned the company well for the future.
SAP grew to prominence during the software industry’s Cultural Revolution. Early in the 1990s, business process reengineering (BPR) was the Holy Grail. CASE was to be the evolutionary stage after the artisan phase of software development. But it didn’t stick, at least in the United States; Europe was more receptive. In the United States, the rage was rapid application development, which meant that software had to be either highly flexible (tough) or continuously new (easier) to keep up with business change. SAP, however, pushed forward with BPR, developing a rigorous, integrated solution for standard business applications.
R/3 fuses BPR with the emerging three-tier architecture for distributed systems. The middle tier, managing business rules and logic, is the integration point for SAP’s entire Business Framework: from front office sales force automation and customer service to decision support and data warehouse systems. Application integration relies on Business Application Programming Interface (BAPI) and Application Link Enabling (ALE) specifications, SAP’s answer to ODBC and OLE.
Dèjá Vu All Over Again?
SAP seems determined to do with enterprise applications what Microsoft did with desktop productivity software. That’s a terrifying thought to best-of-breed vendors. SAP’s liveCache, which uses expanded memory to allow realtime processing of data objects, might even challenge the hegemony of database software.
Of course, there’s no free lunch. ERP, with its required retinue of expensive consultants, is largely beyond the reach of all but the Global 2000. Scalability and complexity are still tough issues. Finally, not all organizations can reorient their business processes around a software edifice.
SAP knows this. With more flexibility and aggressive pricing, SAP is determined to attract a broader spectrum of customers. SAP is focused on solving the biggest problem holding back the intelligent enterprise: application disintegration. SAP’s success and vision make it a clear choice for Number One in the Dozen.
2 Microsoft Corp.
Redmond, Washington
A world without borders: That’s what the New Economy is supposed to bring. In 1998, Microsoft discovered that the world actually does have borders. To reach cyberspace immortality, one first has to ascend through a very physical world inhabited by lawyers, politicians, governments, big media, and a fabric of nervous competitors, each with a highly developed taste for the physical trappings of wealth, power, and property. When computing was the stateless province of nerds and their visionary friends, this nasty, brutish, and short world didn’t care. But with cyberspace access becoming ubiquitous and electronic commerce putting everything that’s not nailed down in play, all eyes are upon the chief nerd and his flock.
It’s hard to look at Microsoft’s technical accomplishments without first considering its courtroom travails. After all, the company might not exist in its present form 12 months from now. On the other hand, its battle with the U.S. Department of Justice could go on for years, harassing Microsoft just enough to disrupt how the company packages its next-generation products. In a volume business such as Microsoft’s, packaging is a core part of its strategy. The antitrust battles will affect Microsoft’s ability to lead customers into the enterprise frontier it has been attempting to construct for them.
Microsoft delivered key products this year. Visual J++ 6.0 angered the Java community with its deep Windows integration, but received complimentary reviews and pushed Java into the mainstream. The comprehensive Visual Studio 6.0 updated Microsoft’s development tools, most significantly integrating them with Microsoft Transaction Server to produce a full Web application server platform. Visual Studio’s Component Manager put the infamous Repository to practical use. Along with Platinum Technology, Microsoft is attempting to make the Repository the foundation for sharing metadata.
Plato and Other Digital Nervousness
According to surveys, Exchange passed Lotus Development Corp.’s Notes in sales in 1998. Now, here’s a case where Microsoft is winning in the market against a strong competitor. Microsoft could be thinking: Perhaps Judge Jackson won’t mind if we drive our Digital_Nervous_System through Exchange rather than Internet Explorer.
OLAP Server, a.k.a. “Plato,” shook the business intelligence software community to its roots in 1998. The entire industry recast itself to make room for OLAP Server, which Microsoft delivered bundled with SQL Server 7.0. The release displayed both the seriousness of Microsoft’s database R&D efforts as well as its talent for consolidating features of best-of-breed products—improved with ease-of-use wizards—into a low-cost package for volume shipment. Now it will be up to Microsoft’s competitors to differentiate themselves to survive.
And NT itself (recently renamed Windows 2000)? Microsoft’s newly ordained president, Steve Ballmer, suggested that it would appear in the fall of 1999. Not so fast, Ballmer: that might require an act of Congress.
3 IBMCorp.
Armonk, New York
When a pitcher hurls the ball home, nine baseball players, each with a distinct position, are instantly galvanized to achieve a single objective: get the other team out. And when it’s their turn to bat, the unifying cause is to score runs. The beauty of baseball is that the rest of the time, the players can simply be who they are. Teamwork does not require that everyone always be on the same page—only when it matters to succeed.
IBM, with more than 270,000 employees, seems to have discovered this essential quality of teamwork. That’s not an easy task when markets are changing and once distinct product and service groups bump hard into each other. Early in this decade, IBM had to put aside internal differences and pull together to achieve one clear objective: survival. Directed by CEO Lou Gerstner, the company more than survived; it thrived.
Perhaps this experience gave IBM a thirst for “big picture” objectives: business intelligence, e-business, and knowledge management. Some might call these buzzwords, but at IBM, they are overarching goals that galvanize the company to meet customers’ business requirements. Each objective plays to the strength of essential team members, supported by others that complete the picture.
The unifying element in these scenarios is DB2 Universal Database (UDB). Rather than walk away from a difficult market posture in the early 1990s, IBM sunk tremendous resources into DB2 to make it the Mark McGwire of the team. DB2 UDB is the data management heavy hitter that ties together IBM’s emergent federated architecture for standard data, digital libraries, and other exotic content. The federated architecture “plays the data where it lays,” as IBM puts it. DB2’s role is to use extensions and advanced optimization to do what a database management system does best, except over a distributed, heterogeneous spectrum. That’s a tall challenge, but with its deep R&D resources, IBM isn’t going to back out of the box.
Knowledge Management: Collaborating on the Future
The keystone of IBM’s knowledge management effort is how well the data management group plays with its collaborative computing teammate: Lotus Development Corp. Hoping to leapfrog its competition, Lotus is about to release major new versions of its Notes groupware and Domino Web server. Lotus and DB2 are also key players in IBM’s “biggest picture” objective: e-business. In 1998 IBM introduced Websphere: a unified application server made up of its veteran transaction middleware players plus Component Broker and Enterprise Java Beans technology. The company’s Visual Age products have been a hit with cross-platform developers.
In 1999, IBM hopes to ride Tivoli Systems’ “Tsunami”—code name for TME 10—into the distributed systems management promised land. The new year should also bring IBM’s resilient AS/400 group more tightly into the fold. When the big picture is in focus, IBM has the players to field a strong team.
4 Intel Corp.
Santa Clara, CALIFORNIA
Intel lives in a shrinking world. Soon, Intel processors will break one gigahertz: an astounding feat made possible by defeating space and time at practically a nanotechnology level. With each barrier broken, Intel makes it possible for the entire industry to pack more computing power into less space. Software developers breathe a sigh of relief; their billions of lines of code will run as speedily on a notebook as they once did on a computer the size of a refrigerator.
Since the introduction of the IBM PC, computer users have been on an extended joyride. At the same time, the world’s economy, with occasional bumps and bruises, has enjoyed a continuous surge of historic proportions. Moore’s Law has become the law of the economy. More than any other company, Intel defines the mantra of the volume business: Deliver your product at an affordable price and fight like mad to drive down costs. Intel’s customers use computer technology to do just that. To a remarkable degree, the world’s economy depends on Intel’s ability to defeat Father Time and extend Moore’s Law into the millennium.
Like its Wintel companion, Intel leverages dominant market share to pay for a massive, continuous reinvention of its products. New manufacturing technology lets Intel extend Moore’s Law with a succession of faster chips: Pentium II, Xeon, Merced, and beyond. To produce these at an economic level the market can bear, Intel has had to redefine its technology.
At the enterprise level, “Intel Inside” no longer means simply a microchip: it refers to a computing architecture. IA-64 Explicitly Parallel Instruction Computing (EPIC) technology, defined jointly by Intel and Hewlett-Packard, will replace traditional sequential execution with, in the words of Albert Yu, Intel’s senior vice president, “the innovative use of predication and speculation uniquely combined with explicit parallelism.” In other words, IA-64 depends on software—specifically, a more intelligent compiler—to get the job done. Intel can’t keep Moore’s Law humming unless the software community does its part.
Wintel…or Telwin?
To succeed, Intel has had to exert a more direct influence over the software industry. Intel’s ambitious plans have put heat on Microsoft. In September, Intel held a “scalability day” of its own, demonstrating a 16-node, NT 4.0-based Pentium II Xeon cluster doing data mining against three terabytes and a two- billion-record DB2 database. It showed the muscle behind Intel’s Virtual Interface (VI) architecture specification, which defines a common server-interconnect technology. With serious database and application vendors on board, Intel is picking up momentum in its battle of the enterprise.
It may not view the Internet as a “strategic inflection point,” but Intel has seen enough to recalibrate its products for specific market levels. Unless Intel jumpstarts desktop multimedia, thin clients will mean thin margins. Intel can’t afford to wait for 64-bit computing before it marches its flag into corporate IT.
5 Oracle Corp.
Redwood Shores, California
Long ago, Oracle identified its quest as bigger than the database. The company worked hard to become an established player in vertical applications—not to mention the DBMS of choice to support competitive packages. As its DBMS cleared the field, Oracle’s development tools strengthened their grip. And as market share increased, the company leveraged this asset; now services deliver more than half its revenue.
Oracle has always faced tough competition—and hammered back with steel and swagger. Oracle’s closest database competitors are wounded and struggling. But now, the challenge has become bigger than the database—and much more threatening. Microsoft, with its billions, is coming after Oracle’s base with SQL Server 7.0 and is using its resources to tear away at Oracle’s mindshare, which jeopardizes the services revenue. Meanwhile, in 1998 SAP launched its full-blown ERP information supply chain, overshadowing Oracle’s Applications release 11.
How does Oracle respond? Ellison put away his jet and yacht, and rolled up his sleeves. As part of a thorough management shakeup, Oracle’s president, Ray Lane, overhauled the Applications sales force and put Peter Dunning, SAP’s former U.S. sales chief, in charge. Then, the company produced two key releases: Application Server 4.0 and Oracle8i. Much derided when network computers themselves didn’t take off, Oracle’s Network Computing Architecture (NCA) suddenly looks very real.
Application Server 4.0 has become the defining product in a crowded, nebulous field that includes Web servers, middleware, object brokers, and security management. Will application servers evolve in the same way that DBMS software did 20 years ago—that is, from a collection of related functions into an integrated whole? Oracle thinks so.
You Can Take the Oracle Out of the Database, But…
Hard as it has tried to redefine itself, Oracle’s continued success depends on the database. Migration to Oracle8 has been sluggish; 8i may be just what the doctor ordered. “Windows is dead,” Ellison and his commanders are proclaiming, “Long live Oracle 8i.” Such hyperbole may not scare Bill Gates, but the renewed platform debate has Oracle back in the news.
Oracle 8i is truly Internet-infused. Along with putting a Java Virtual Machine inside, Oracle introduced the Internet File System (IFS), interMedia, and WebDB. What were once clumsy object/relational cartridges now look like a sensible NCA idea that counters Microsoft’s OLE DB. With 8i, Oracle has finally offered a clear statement of its vision for the future of the database (or rather, the “Internet information management platform”). Will it render operating systems irrelevant? Alas, the deciding factor remains the same: performance.
Ellison wants to expand NCA beyond technology. He has proposed that Oracle host enterprise applications online as an “application service provider.” Bold vision gives Oracle its sustaining momentum. Now the company has to come through—and make sure it doesn’t race too far ahead of its customers.
6 Siebel Systems
San Mateo, California
Client/server was a sea change. The Internet is a sea change. Reusable distributed components…well, the jury is out. Technology, however, is only part of the story behind the biggest sea change in business computing. After decades of intense focus on automating back-office “cost side” operations, such as billing or inventory management, businesses are in a frenzy to streamline the revenue-generating side of the ledger. Organizations are snapping up sales force automation (SFA), call center management, and customer relationship management (CRM) tools with hopes of driving down costs per sales lead and integrating sales and service into an enterprise-wide whole.
Siebel Systems, founded just five years ago, is already a leader in SFA. Revenues have been growing at an obscene rate. Siebel now has its sights set on long-range ambitions of becoming the dominant player across the spectrum of front-office “enterprise relationship management” (ERM) software. In 1998, Siebel merged with Scopus Technology, a leader in call-center and customer service solutions.
However, it’s no rumor that ERP vendors—SAP and Oracle in particular—are determined to get their share of the ERM market. In 1998, each made real and FUD news: The latter perhaps intended specifically to slow Siebel down. “Best-of-breed” has suddenly become a favorite epithet. Best-of-breed, indeed, ERP vendors say: Why expose your company to further incompatibility? One-stop shopping is where it’s at. How can Siebel compete?
Siebel’s weapon is its n-tier Internet Component Architecture: the foundation of the Enterprise Applications suite. Siebel has bet its business on what it sees as the only workable architecture for next-generation Internet-based computing. While the architecture can accommodate Java, Siebel’s products, tiers, and interfaces offer a full-blown implementation of Microsoft’s distributed object computing, from Visual C++ and ActiveX down to OLE and DCOM. The architecture is open at the user interface level, allowing developers to create custom interface components—and potentially let users download myriad industry-specific software products off corporate intranets or via external Internet connections.
Doing Business with Objects
The focal point is Siebel’s implementation of business objects. These let developers work at a level of abstraction separate from both user interfaces and internal “operational tiers,” which consist of Internet, security, and other middleware managers. Business objects support modern process integration; using OLE interfaces, Siebel tools can interact with other applications and legacy systems.
Components, once the future, have become the present as far as Siebel is concerned. The company’s architecture could let a “Siebel-centric” selection of best-of-breed software counter what the ERP vendors offer.
With the Scopus acquisition, Siebel has homegrown integration issues to solve. But once the full-blown ERM of the future comes together, look out: Salespeople will tell you things about yourself that even your dog doesn’t know.
7 NCR Corp.
Dayton, Ohio
Lars Nyberg, chairman and CEO of NCR, turned a nice phrase during his keynote at the NCR Partners user group event in October—a phrase that we will hear often from NCR. “Data warehousing,” he declared, “is about turning transactions into relationships.” A data warehouse can offer an organization many benefits, but the biggest is the ability to use its information to learn about customers. In our Consumer Age, knowing your customers is not a touchy-feely concept; it is the first principle of successful competition.
CRM exploded in 1998. Our industry is forever regrouping product categories to create something that sounds new. CRM encompasses a vast range of existing products: from sales and marketing automation software, to demographic content, to call center and help desk applications, to OLAP, data mining, and data warehousing systems. CRM becomes significant when organizations can close the loop around these technologies to discover information about customers, interact directly with them, and offer appropriate products and services. This “positive feedback loop” helps companies retain their most profitable customers.
NCR is determined to be first among product and service vendors when it comes to the data and knowledge discovery side of CRM. With its Teradata relational DBMS built specifically for scalable data warehousing, NCR already offers one of the best CRM engines in the business. Having struggled for many years to get corporate attention while NCR headquarters sorted out other matters, the Teradata group had to be gratified to hear Nyberg call the DBMS the “cornerstone of the company.” With electronic commerce and the use of customer interactive devices vastly increasing the amount of data and intensifying the need to explore it at a detail level, Teradata's brightest days may lie ahead.
A One-to-One Relationship
1998 saw the general release of a new version of Teradata that expanded support for OLAP and data mining through a series of SQL extensions. NCR pumped up its OLAP Services by forging a licensing agreement with Geppetto’s Workshop to resell its dynamic query generator, The Ant Colony, as Teracube. But perhaps most important, Teradata is now running on Windows NT. One of Teradata’s selling points is that it is essentially self-maintaining, requiring little DBA intervention. This could be a serious selling point in the price-conscious NT market.
Unlike its competitors, Teradata is designed specifically for data warehousing, so DBAs don’t have to do all the messy design work necessary to get good performance out of OLTP-oriented software. The product’s shared-nothing parallelism also fits well with Microsoft’s vision of NT scalability; in fact, it helps Microsoft fend off critics by offering proof that with the right architecture, NT can scale.
One-to-one marketing and mass customization, two CRM goals, require a system that can carry its weight in data and provide fast ad hoc analysis. If NCR can flex enough marketing muscle, Teradata could lift the company to new heights.
8 Sun Microsystems
Mountain view, CALIFORNIA
Sun can rest assured: The sky is not falling. Windows NT will not kill Unix, which has been unable to kill the mainframe. Sun Microsystems stands alone as the only major vendor selling exclusively Unix-based systems. What sounds like weakness—given the popularity of NT—is looking more like strength. First, Sun now dominates Unix. And second, just as IBM brought down prices and focused on modernizing the platform, NT’s low-end success has forced Sun to make Unix servers enterprise-ready. Meanwhile, Microsoft, Intel, and their partners don’t have NT ready to compete with Sun’s high end.
As information systems grow larger and perhaps more centralized, the mainframe and its professional supporting cast will remain tough to dislodge. But with scalable systems running SunSoft Solaris on Ultra-Sparc processors from its Darwin workstations up to the mainframe-class Starfire, Sun offers a solid alternative for new-age, Internet-based distributed systems.
In 1998, Sun celebrated more than 500 licenses for its Ultra Enterprise 10000 Starfire, acquired from Silicon Graphics Inc. in a transaction that SGI may now regret. Starfire, a true-blue symmetric multiprocessing (read: not ccNUMA) server, was the missing piece in Sun’s scalability puzzle. It has become the focus of Sun’s Full Moon clustering strategy, allowing Sun to offer its Cluster 2.0 package as a high-availability solution—and attack Microsoft’s distributed cluster approach as unreliable for enterprise systems. Sun is also working scalable data sharing and intelligent storage strategies into the picture, using technology from its 1997 Encore Computer acquisition.
Solaris: The Center of the Universe
In October, Sun refreshed Solaris with a new version that put the operating system at the center of a tightly integrated development platform. Solaris will ultimately command an array of software: Java, Enterprise JavaBeans (EJBs), Jini, and NetDynamics, which Sun acquired in 1998. With NetDynamics, Sun has a strong player in a hot market—and a potential integration point with NT-based applications. Sun also announced Project Cascade, focused on cleaning up interoperability between Solaris and NT.
Java had the most confused year in its young life. Working with Symantec, Sun promised major performance improvements in its Java Virtual Machine (JVM), but these were postponed until 1999. Client-side Java ran into snags, while server-side Java and EJBs showed potential, particularly for legacy integration. Sun officially submitted Java to the International Standards Organization, but didn’t seem at all ready to relinquish control over the language’s implementation. Nasty legal difficulties with Microsoft cast a cloud over Java’s potential.
Will the Unix bug bite Java? In absence of a standard, Hewlett-Packard and others are building proprietary JVMs. Just like the old Unix days; proprietary JVMs could end up as part of each vendor’s enterprise solutions arsenal. As Sun remakes itself into more than just a hardware vendor, it must decide where to draw the lines of competition.
9 Information Advantage
Eden prairie, minnesota
Will information portals be the doors of corporate perception? What will happen when Yahoo! meets OLAP? The result will be a tremendous upswing in demand for decision-support analysis tools, as well as data warehouses that can manage and deliver information. Through a browser interface, corporate users could perform advanced text and multimedia searches; via the data warehouse, query a heterogeneous collection of databases; and using agents, OLAP, and data mining, obtain timely and precisely defined slices of information.
Information portals were the buzz of the data warehouse community in 1998. Vendors jumped on the concept, but none defined the opportunity better than Information Advantage. In September, the company announced DecisionSuite Eureka!—an Internet-based product that promises to fuse the portal concept with business intelligence. The announcement culminated a momentous year. Then in July, Information Advantage acquired IQ Software, a leading enterprise reporting tools company. Due out in December, Eureka!, based on IQ technology, will be a stand-alone product. It will also serve as the unification point for the two companies’ product lines.
The acquisition was a major event in a year that saw lots of movement among OLAP and reporting tool vendors as they braced for Microsoft’s entry into the market. The IQ acquisition gives Information Advantage a bigger product line, customer base, and presence on business users’ desktops. The company plans to customize Eureka! for R/3 users, hoping to fill the void before SAP does.
Eureka! There’s My Data!
Most important, Eureka! brings the business value of a data warehouse into better focus. Data warehouse “value” was a fiercely debated topic in 1998, particularly after the mainstream business press ran negative stories reporting expensive failures. Bill Inmon and other gurus fired back that the stories were based on questionable data. However, the bad press served notice that at the end of the day, all the methodologies, metadata, and star schemas had better boil down to tangible business value.
Rick Tanler, chairman and founder, has long been an intellectual leader in this fight. Like Red Brick Systems and Brio Technology, the company was founded by former Metaphor Computer Systems people. Tanler and his colleagues based the company’s three-tier relational OLAP technology on their experience in working with big companies doing big DSS and data warehouse projects, particularly in the retail industry.
In Tanler’s view, personal computers have given businesses an “easy-to-use GUI created for every application, which more appropriately is a UFO, an Unwanted Feature Organizer.” With browsers, content takes precedence over application features. And by fusing the Internet with data warehousing, Information Advantage wants to open the door to an unparalleled opportunity: customized information that raises the value of an organization’s intellectual capital.
10 MicroStrategy Inc.
Vienna, virginia
Go west, young men: As cities and towns across 19th century America settled into the established way of life, this command drove the adventurous deep into the frontier. In the land of business intelligence, OLAP, once the wild frontier, is settling down. Microsoft has arrived. Oracle, IBM, and other leading database companies are bundling OLAP with their DBMS engines, where it could become simply a standard database function. The early OLAP settlers now face difficult choices: Consolidate and become part of the establishment? Specialize and try to dominate a profitable vertical market? Or find the next frontier?
MicroStrategy is heading west, at least in a figurative sense. With founder and CEO Michael Saylor at the helm, it’s doubtful the company will ever abandon its quest. From the earliest days, Saylor and MicroStrategy have captured the imagination of IT professionals. The company’s leading customers dare to go beyond cost-effective data access and analysis: They want to achieve business objectives that would be unattainable without advanced decision support. Having invested in huge data warehouses, many are ready to develop data-rich products and services that not only bring a better return on investment, but also make a profit.
A good example is National Data Corp.’s Health Information Services subsidiary. With MicroStrategy’s DSSWeb, Health Information Services has created Intellect Q&A, a Web-enabled subscription service that offers access to its multiterabyte warehouse of healthcare market information. Intellect Q&A takes the traditional goal of wider data access to a new level. Along with cost-effective data access for healthcare professionals, the service opens up an incredible resource for a wider audience, such as investment analysts, advertising agencies, and healthcare researchers.
Mom, Your Database is on the Phone
Intellect Q&A also shows Saylor’s “Query Tone” concept at work. A Query Tone, like a dial tone, enables access to databases anywhere in the world. Data warehouses become the crown jewels of new Web-based data syndication or subscription services. In 1998, MicroStrategy introduced DSS Broadcaster, which turns the “passive” Query Tone into an active service. Using personalized alert triggers in the database, DSS Broadcaster can inform mobile service customers—via their pagers, Palmtops, cell phones, or fax machines—instantly with critical information. A financial services company could keep not only its own salespeople and traders better informed, but also its customers. These days, it’s not just professionals who need to know the price of pork bellies.
MicroStrategy went public in 1998—and investors rewarded the company for its market success and compelling “consumer DSS” vision. Even in our wacky Internet age, Wall Street will expect MicroStrategy to settle down and offer nice quarterly returns: a pressure that is not always easy for the adventurous to bear. But if MicroStrategy is right, the DSS frontier will become a boomtown.
11 BEA Systems
Sunnyvale, California
Transactions are about getting something from one place to another, safely and intact. BEA Systems, built on transaction systems, is in the middle of a journey of its own. And a rapid one: BEA is determined to land in the top tier of enterprise middleware providers.
The company began its quest in 1995 by acquiring the venerable Tuxedo OLTP system, using cash from a stunning $50 million investment by Warburg Pincus. While dwarfed by IBM’s mainframe IMS and CICS, Tuxedo held, and still holds, a serious share of the TP market and is the leader in distributed Unix. As IBM could attest, people don’t swap out their OLTP systems easily; the purchase gave BEA a solid base of license revenue. This year, in a Computer Associates-like move, BEA stoked its license engine by acquiring Top End from NCR.
But beyond just a foundation, Tuxedo and Top End have given BEA a leadership position in open systems OLTP and a valuable cadre of experienced engineers. In the 1980s, OLTP was considered a mind-bendingly tough problem. The major hardware vendors employed some of the best brains in computing to solve OLTP for their proprietary systems, and they succeeded. AT&T, then owner of Unix, got Tuxedo rolling as the first open OLTP system. Soon, the industry’s OLTP brains were consorting to create Distributed Computing Environment, which became the basis for CORBA. In other words, while BEA is new, it is actually keeper of an open systems flame that has been burning for some time. Led by CTO Alfred Chuang, BEA’s brains are concentrating on building the next generation.
OLTP: From Mission-Critical to Business-Critical
Distributed object (or component) transaction middleware is the 1990s mind-bending OLTP problem. After a brief period of stability, OLTP has become ground zero for e-commerce, ERP supply chains, shared data transactions, and mobile computing. CORBA, Microsoft’s DCOM, and now Sun’s EJBs simplify specifications, but also create integration problems. Fusing compliant applications with OLTP systems still requires expert handiwork. In competitive markets, it has become business-critical to solve OLTP application integration quickly. BEA is betting that customers would rather buy a plug-and-play OLTP system than build their own or cast their lot with Microsoft and its evolving NT solution.
BEA made strides in July when it shipped M3, an object transaction monitor. But even as it released M3 to great fanfare, BEA found itself in danger of being left out of the hot application server market. M3 did not come out of the gates with Java support. To remedy matters, BEA acquired WebLogic, provider of the established Tengah Java-based Web server. It was the biggest, most strategic move since the company’s founding.
Suddenly, BEA is closer to its destination. If it can integrate its products into a seamless architecture, it could emerge with the most comprehensive solution of all.
12 Hyperion Solutions
Sunnyvale, CALIFORNIA
Analytic apps was one of the loudest rallying cries of 1998. Most plainly, the phrase means building applications specifically to enable business and data analysis. OLAP systems and data marts quickly come to mind, along with systems built with feature-rich reporting tools. But just like in the Merriam Webster Dictionary, secondary and tertiary definitions are more descriptive of what the phrase really means.
As technologies mature and address a wider customer base, packaging becomes critical. Arbor Software, which this year completed a merger with Hyperion Software, started the ball rolling five years ago when it defined server-based multidimensional analysis as OLAP with Essbase. OLAP has been hot, but all buzzwords lose their buzz eventually. Analytic apps is a nice new phrase that sounds like standard equipment in its business-critical application portfolio, rather than some acronym-infested technology that requires an IT organization to run.
A tertiary definition calls more upon context than etymology. With growing concern and envy, the DSS/data warehousing industry has been watching ERP explode. Hundreds of vendors and best-of-breed toolboxes look awfully chaotic next to tightly integrated packaged apps and desktop suites. ERP has raised the integration expectations of users; survivors in a maturing market will either have metadata integration at the heart of their architectures or offer a seamless hook into SAP’s BW, Microsoft’s OLE DB for OLAP, or another industry standard. Of course, that brings up another couple of contexts: SAP’s developing interest in business intelligence and the arrival of the 500-pound Redmond gorilla in the marketplace.
All for One, One for All
Arbor and Hyperion, a leading provider of vertical business analysis tools, didn’t need a dictionary to read the writing on the wall. Now Hyperion Solutions, the merged entity is poised to become a powerhouse in analytical applications. With the company’s management and services sorted out, it’s clear that Hyperion’s emphasis will be on applications enriched with appropriately transformed data. Essbase has become Hyperion Integration Server. This middle-tier server provides the vertical tools with metadata integration, automated management, and a multidimensional data transformation engine fed by data warehouses, ERP systems, and other resources.
In other words, all the relational OLAP vs. multidimensional OLAP and data mart vs. data warehouse debates are side issues; the starting point is the application. For both Arbor and Hyperion, the number one customer was always on the business side: As a single entity, the company hopes to have the resources to rise above ERP competitors that are just now developing their analytical application strategies. It hopes customers prefer working with a vendor that has a core competence in value-added data analysis.
Through Arbor, Hyperion picks up a close relationship with IBM that might be worth keeping an eye on. Challenges lie ahead, but Hyperion offers a compelling definition for analytic apps.
13 Computer Associates Inc.
Islandia, N.y.
Things did not go exactly as planned for Computer Associates in 1998. Judging services as the critical ingredient separating it from the likes of IBM, Digital Equipment, and Hewlett-Packard, CA tried to acquire Computer Sciences Corp. “Hostile” is the only word adequate to describe CSC’s reaction, or at least that of its chief executive, Van Honeycutt. The acquisition would have dramatically changed CA. However, given the difficulties the acquisition presented, it may have been the best deal CA didn’t make.
High-margin revenue streams that come with services are certainly tempting. Perhaps CA saw in CSC something similar to what attracted it to Ask/Ingres or Legent—companies with strong license revenue coming from mature products, but facing increasingly commoditized industries and in need of radical cost restructuring. Was CA going to use CSC to challenge the service pricing of the Andersens, KPMGs, IBMs, and the rest? We’ll never know. In any case, CEO Charles Wang did make good on his goal of building up CA’s Global Professional Services by shifting 1,000 employees to the new division and acquiring Realogic Inc.
CA needs more “feet on the street” to accelerate the adoption of its two advanced products, Unicenter TND and Jasmine. The dynamic ERP market, along with the application integration demands of e-commerce, ought to be driving a similar explosion in demand for integrated systems management suites. However, the new terrain creates very tough technical challenges; Unicenter’s early adopters need help. Similarly, object databases suffer in the marketplace because so few people know how to design and manage them. In 1999, perhaps we’ll see CA’s services arm hoist these products into the mainstream.
14 EMC Corp.
Hopkinton, Mass.
From junior salespeople to the executives in the boardroom, EMC is a driven company. EMC’s takeover of the large-system storage business is the stuff of business legend. EMC’s growth has paralleled that of data warehousing. With the world’s biggest companies likely to be storing upwards of a petabyte of data early in the next century to support data mining, digital libraries, and other applications, EMC’s business can only grow. Its technology sets the standard in data warehousing. This year, EMC significantly added to its Symmetrix array storage capacity.
Now that EMC is top dog, however, its competitors are gunning after it. Network-centric computing and scalable clusters are changing the landscape and will open the market to different concepts of how storage should be managed. Compaq Computer is now behind what was Digital’s StorageWorks, and Sun has its Intelligent Storage Network architecture. StorageTek, IBM, and Hitachi are back in the game.
EMC’s response is to emphasize its open systems storage flexibility and position itself for next-generation, enterprise storage area networks (SANs). It is investing heavily in Fibre Channel R&D and software development. EMC is also fighting hard to be Bill Gates’ favorite distributed data storage provider for Windows 2000 clusters. Intelligent storage management is no longer an afterthought; it is critical to scalable performance. Microsoft plays its cards close to the vest, but chances are EMC will be a key factor if 2000 is to succeed in the enterprise.
15 Business Objects
San Jose, Calif.
Web-enabling decision-support tools sound great, but implementing a production release that takes the Internet to heart is a challenge. The first step is to understand what customers are really going to do over primarily intranets and extranets, and then adjust OLAP and enterprise query and reporting appropriately to work in a browser-based, thin-client environment. Second, security and firewalls must be addressed up front, without adding pain to an already complex administrative issue. Finally, the tool’s interface must meet users’ new expectations, which are based on their experiences with Yahoo! and other portals and Internet search engines.
Business Objects met these challenges with the release of WebIntelligence 2.0, arguably the most comprehensive Web DSS tool release of the year. The product shares a common metadata infrastructure with Business Objects, the company’s flagship integrated OLAP query and reporting product, which is itself scheduled for a major new release in 1999. Java-based WebIntelligence also sits on Inprise Corp.’s CORBA-compliant distributed component architecture, allowing it to become part of larger enterprise software environments.
Business Objects is also moving aggressively to position its products for ERP business intelligence. The 5.0 release will employ Microsoft’s VB as its scripting language, which could become interesting if the Redmond giant succeeds in convincing leading ERP vendors to make VBA their development environment of choice.
16 SAS Institute Inc.
Cary, N.c.
A pillar of the Raleigh-Durham community, SAS Institute is consistently voted one of the top 100 places to work in America. Perhaps that is the secret behind its 20-year plus run as the leading statistical analysis software company—something the company is working hard to extend into the OLAP, data mining, and enterprise reporting territory. Mirroring its work environment, SAS’s goal is to build tools that help people work collaboratively to get the job done.
As it has developed each of its enterprise tools, SAS has thought through how the entire DSS process should work and organized its products to follow a methodology. Its multidimensional server is a hybrid OLAP (HOLAP) product, giving users the choice of working directly with relational databases or off of cubes or aggregates. The SAS Enterprise Miner suite offers an interlocking set of algorithms, enabling data mining efforts to fit the business goal. In sum, SAS’s technology is not in love with itself; it exists so that people can get information and accomplish their objectives, without making mistakes.
The SAS approach is refreshing in an industry where hype tends to drive IT customers down a few dead ends in the name of higher technologies. SAS might not lead the charge, but neither does it get in the way when it’s time to use technology to meet business objectives.
17 Compaq Computer Corp.
Houston
The origin of Compaq’s fame and fortune was the PC clone, which it essentially invented. In the 1980s, its strategy devastated IBM’s PC business, which meant that IBM couldn’t control the PC’s growth, thereby opening the door to the PC revolution. Sensing well before its competitors that PCs were becoming commodities, Compaq became the first vendor to market PC servers. Now the company is betting its business on commodity clusters—and once again is positioning itself to knock off IBM and the other enterprise server vendors.
To get there, however, the company has chosen to play Dr. Frankenstein. In 1997, Compaq bought Tandem Computers; this year, it completed the biggest deal in industry history, buying Digital Equipment Corp. Although very different, Compaq, Tandem, and Digital all prospered by beating IBM to the punch. The latter two did it largely through distinctive engineering, of which both were justifiably proud—too proud. Now the big question: Can Compaq breathe its hard-driving entrepreneurial spirit through this giant, grafted-together body, so that it springs to life as the dominant enterprise solutions vendor?
Many analysts are skeptical and worried that Compaq can’t concentrate with Digital’s problems in its lap. However, Compaq has stated its strong commitment to the parts of Digital that matter: AlphaServer, StorageWorks, and its Alta Vista Internet endeavors. If Compaq can integrate these technologies with its own, plus Tandem’s popular ServerNet interconnect and NonStop SQL/MX running on NT, it will have an awesome enterprise clustering solution.
The toughest challenge will be reconstituting services. Widely called the “crown jewel” of the acquisition, Digital’s services division succeeded partly because customers viewed it as a neutral party. With Compaq as owner, they will undoubtedly adjust their view. However, building cluster solutions requires the heterogeneous talents that were Digital’s strong suit. That kind of talent is no commodity.
18 PeopleSoft Inc.
Pleasanton, Calif.
After going gangbusters, the ERP market lost some steam in 1998. Some blamed it on the Y2K obsession, others on momentary market saturation, and still others on a more ominous conjecture: SAP is taking over. PeopleSoft, Baan, and others that have had enormous success in specific packaged application markets are now in SAP’s crosshairs. In particular, the German concern wants a big helping of PeopleSoft’s lucrative human resources business.
PeopleSoft probably has SAP’s attention in part because of its concerted effort to extend its 7.5 suite from human resources into retail, logistics, financials, and supply-chain manufacturing. In addition to developing its own business intelligence/data warehouse strategy, PeopleSoft is in a race with SAP to transition its products away from client/server to a tiered Internet-based computing model. Revenues have been riding a roller coaster, but PeopleSoft is still on the rise.
19 Informatica Corp.
Palo Alto, Calif.
In 1998, SAP tapped Informatica’s PowerCenter integration server as one of the first data extraction and transformation tools certified for its BIW architecture. Informatica is determined to make the most of its relationship. Not tough: The SAP BIW badly needs tools like PowerCenter for it to be a credible data warehousing solution. Informatica is integrating its metadata layer with SAP’s Business Object Repository. The metadata layer has also received strong support from other BI tool vendors trying to become integrated with SAP BIW. In October, the company reintroduced PowerCenter as the core of its Decision Processing System, a blueprint for integrating operational data with analytical applications.
With a modular, component architecture, Informatica’s flagship PowerMart data mart is growing into an enterprise tool suite. After phenomenal, departmental demand-driven growth, standalone data marts are now viewed as part of IT’s “islands of data” problem. As a solution, Informatica has been promoting a more IT-centric strategy that hinges on a global repository, which maintains the metadata and handles user access to it. This, Informatica believes, will give organizations the best of both worlds: business flexibility plus IT stability.
20 Brio Technology Inc.
Palo Alto, Calif.
Taking after its Italian name, Brio has been adding energy and brilliance to business intelligence since the release of Data Prism in 1990 and Data Pivot in 1991. An easy-to-use interface and other user-friendly touches gave business analysts a good first impression of Brio, which helped the company maintain a solid market share amid tremendous competition. With Web-based business intelligence getting ready to explode, the company went public in 1998 and is gathering its strength for an allegro con brio surge.
Brio’s Enterprise Server Suite 5.5, which works in both Web and client/ server configurations, has become the product of choice for many large organizations looking to standardize on one vendor for analytical applications. The company’s OnDemand and Broadcast Servers showed Brio’s potential for “push/ pull” Web applications and business-to-business information supply chains.
In August, IBM announced that it would bundle Brio’s products into its Visual Warehouse offering. The first fruit of this combination was eScholar, built by Vision Associates, which gives elementary schools the ability to do reporting and analysis on student demographics, grades, income levels, and more. Could student relationship management and one-to-one education be far behind?
21 Cognos Inc.
Burlington, Mass.
Microsoft’s OLAP Server probably won’t affect civilization as profoundly as its code namesake “Plato” did with his philosophy, but who knows? Panorama Software Systems, the Israeli company that developed Plato for Microsoft, has also created a disciple: Aristotle. Cognos picked up the exclusive rights to Aristotle, a business intelligence analysis tool designed specifically to work with Plato OLAP Server.
Cognos, which already has its Impromptu query reporting tool, Scenario data mining tool, PowerPlay OLAP server, and the exciting Data Merchant “data storefront” Web software, wants Aristotle to stick close to Plato. If Microsoft is as successful with SQL Server 7.0 as it imagines it will be, Aristotle could be more ubiquitous than its namesake is in college philosophy courses. Anyway, who needs philosophy when you have an information system?
In 1998, Cognos updated its popular flagship products for Web browser architectures. Their positive reception, augmented by Aristotle, ensures that Cognos will remain in the forefront of the business intelligence market.
22 AlphaBlox Corp.
Mountain View, Calif.
Often overlooked in the hot buzzword “analytical applications” is that second word: applications. Most query and reporting tools evolved out of a data-centric view, with the query as the root activity from which everything else grows. Increasing the user’s flexibility to customize query and reporting has involved a difficult complexity tradeoff: either fewer users get the tools or more administrators have to spend more time querying and staging the data.
But with the Web becoming the primary user interface, IT better get ready: RAD is going Web time. The query and reporting world is in danger of looking stodgy and inflexible. AlphaBlox, founded in 1996, takes a modern developer’s view of analytical applications. Using Java components and a three-tier Internet architecture, AlphaBlox Enlighten gives developers a set of standard “Blox” that supports an Application Assembly Studio, or template, for specific applications. The company clearly imagines a fast-developing world of snap-together JavaBean Blox that will enrich future analytical applications.
Enlighten offers an exciting concept that has drawn major venture capital backers to the company, as well as major early adopters. Scalability and component integration loom as speed bumps for AlphaBlox, but if it succeeds, query and reporting may never be the same.
23 SilverStream Software Inc.
Burlington, Mass.
Can lightning strike again? Hoping to borrow a little quicksilver from its founders, who supercharged Powersoft, Lotus, and a few others to sky-high client/server success, SilverStream has entered a strong product into the Web application server arena. This hot market is getting crowded with established players that include Oracle, Sun (NetDynamics), and IBM but experienced hands engineer SilverStream as well. It is also a complete Java implementation and comes with a full range of development tools.
SilverStream is designed for database-centric applications, but doesn’t interact with databases in the same way as PowerBuilder or other client/server tools: It does not have to bow that legacy. Instead, it uses a middle-tier application server to store data relationally; it employs multithreading and asynchronous communication to push up performance. Agents are used liberally to manage processes, such as triggers and stored procedures. Finally, SilverStream supports the full range of Web development tools. In a crowded field, SilverStream could be a supernova.
24 Progress Software Corp.
Bedford, Mass.
If nothing else, new Web application servers must be flexible. Web development environments are still evolving, as are Web business objectives. Apptivity, Progress Software’s visual Java development tool and application server offers supreme flexibility, but also stresses enterprise scalability, security, and fault tolerance. With a CORBA-based architecture, Apptivity 3.0 joins server-side Java with traditional application server multithreading and load balancing. Progress may have a hot product on its hands.
Unless Apptivity produces channel conflict, that is. Progress ADE, with Progress 4GL, is a successful, established enterprise application development environment with a loyal customer base. The Progress ADE 8.2 RDBMS sports a 64-bit database engine that exploits very large memory (VLM) technology, putting it on par with Oracle for high-performance database applications.
Apptivity swings to a new beat for a new generation of developers. This stuff looks too easy to the 4GL crowd. Luckily for Apptivity, once things get tough it has a deep pool of experience to draw on.
The Database Derby
Informix (Menlo Park, Calif.) has been paying dearly for the sins of several years of over-optimism throughout the database industry. The company hopes that with a good database product—heck, maybe even more than one—the momentum will swing its way. Informix bought Red Brick Systems, provider of the respected data warehouse-only DBMS and related consulting services, which may be of most interest to Informix. However, if the company finds synergy between the database products, it could have a very powerful offering.
Sybase (Emeryville, Calif.), with its Powersoft tools, Jaguar transaction middleware, gateways and replication software, IQ data warehouse products, and Adaptive Server, has a very diverse portfolio. The question to ask: Does it have the horses to support it all? Right now, the company is aligning its strongest energies to become the definitive vendor for mobile and embedded computing. The patter of Sybase’s small footprints is joined by Centura Software Corp. (formerly Gupta; Redwood Shores, Calif.), which has turned SQLBase into a database designed to be “invisible” to the end user. Both companies must contend with Pervasive Software (Austin, Tex.), which thanks to its Novell origins started with a large installed customer base and has gone on to become dominant for embedded, ultralight databases.
Cloudscape Inc. (Oakland, Calif.) is also making small database footprints, but of the Java variety. Started by ex-Sybase CEO Mark Hoffman and staffed with several other top people formerly of Sybase, Cloudscape hopes to cash in on Java’s cross-platform potential to set it apart from its competitors. Object database vendors hope Java developers might prefer storing objects rather than rows and columns. Poet Software (San Mateo, Calif.), the most successful desktop object database vendor, is also a player in embedded systems. Versant Corp. (Fremont, Calif.), bulked up by partnerships with best-of-breed Web application server, development tool, and middleware providers, introduced its new enterprise component management system.Objectivity Inc. (Mountain View, Calif.) celebrated important customer wins in healthcare and other fields as it enjoyed a growing niche for nontraditional database applications.
Finally, with memory becoming cheap and 64-bit systems on the way, in-memory databases become a serious performance alternative. Several vendors are active in this field, but one new company to watch is TimesTen Performance Software (Mountain View, Calif.). Gary Morgenthaler, Roger Sippl, and other successful software industry “angels” are behind this one. It hopes that its smoking relational engine will attract customers with real-time performance requirements, particularly for e-commerce.
Enterprise Integration
Thin Web-based clients and application servers give middleware a luster it hasn’t previously enjoyed. Few will forget Katrina Garnett’s black dress, which brought fame to CrossWorlds Software Inc. (Burlingame, Calif.) and its products for integrating packaged applications. Component application middleware is the heart of Inprise Corp.’s (Scotts Valley, Calif.) new mission: integrating the enterprise. Shedding its Borland identity, Inprise has Delphi and JBuilder winning the hearts and minds of developers; Application Server is for customers who dream as big as Inprise does. Meanwhile, Bluestone Software (Mount Laurel, N.J.), one of the earliest cross-platform middleware vendors, is going gangbusters with Sapphire/Web.
Netscape Communications Corp. (Mountain View, Calif.), if it can keep its eyes off the courtroom, could develop into an enterprise player. Using technology from its Kiva Software acquisition, Netscape Application Server is attracting fans. With former DBMS columnist David Linthicum as CTO, Software AG (SAGA; Reston, Va.) is in the midst of an aggressive repositioning. SAGA’s agnostic middleware approach will help enterprise customers tie legacy transaction systems to either CORBA or COM-based servers, and then to the Web. Persistence Software Inc. (San Mateo, Calif.) is refitting its object cache architecture to support transaction-oriented EJBs. Object database technicians are behind Novera Software Inc. (Burlington, Mass.), which explains the thorough enterprise database thinking evident in its fast-growing Java application server. TopTier Software Inc. (San Jose, Calif.) introduced its “HyperRelational” integration server that grabbed attention with its hypertext approach.
Information Builders Inc. (New York), working with its huge Focus installed base, has put together a heterogeneous application server and is hoping to fuse Focus with R/3 applications. Mainframe shops looking to join the Web development world found a friend in Attachmate Corp. (Bellevue, Wash.), provider of Web-to-host and other connectivity software.
Performance demands of real-time, event-driven computing have made Tibco Inc. (Palo Alto, Calif.) a magnet for e-commerce systems requiring enterprise application integration. When push becomes publish/subscribe, Talarian Corp.’s (Los Altos, Calif.) SmartSockets logical routing technology helps organizations meet mission-critical goals. Taking a more business-level view, Vitria Technology Inc. (Mountain View, Calif.) has become one of the most interesting companies to emerge in the enterprise application integration space. Using a federated architecture and asynchronous messaging, its BusinessWare manages “business events” across multiple ERPs or other applications.
Application Development
The emergence of Linux was one of the biggest stories of 1998. While Unix has gone uptown and corporate, Linux has kept original spirit alive—enhanced with a modern programming vision. Perhaps out of fondness for Linux—or maybe just to make Microsoft uncomfortable—major vendors are now behind the operating system. While the community fears one vendor taking control, it has to produce some star entities to go anywhere. Red Hat Software Inc. (Research Triangle Park, N.C.), a Linux operating system leader, is developer of the top Linux Web development server.
Getting there first is always a good thing: but in this field, you have to work hard to stay ahead. Cold Fusion from Allaire Corp. (Cambridge, Mass.) was first and remains a leading Web application server. The critical 4.0 release is doing well. Visual Café has been a critical product in Java’s rapid acceptance. Visual Café also revived Symantec Corp. (Cupertino, Calif.) as a technology leader; the company is expanding quickly to become an important player as Java moves into the enterprise.
Rational Software Corp. (Cupertino, Calif.) continues to lead the development world into components. Rational Rose is the de facto standard modeling environment. However, Rational is feeling the heat of competition from Rogue Wave Software Inc. (Boulder, Colo.), which bought Stingray Software this year to emerge as a major software development vendor. Visio Corp. (Seattle), with a collection of tools, is also growing with enterprise luster as visual computing applications become part of the business mainstream.
Glamorous applications depend on less-glamorous technology to perform well. Mercury Interactive Corp. (Sunnyvale, Calif.), the top vendor in testing tools, has made a big move into the ERP space, where testing could not be more critical. From Paris, France, CAST Software (U.S. headquarters are in San Francisco) is turning heads with its break-the-mold “evolutionary management” approach, which analyzes application and database source code to give developers a far more detailed understanding of the impact of software change.
Business Intelligence
In 1998, Web-based data access and analysis dominated all aspects of this field. Pure Java-based SpaceSQL and SpaceOLAP from Infospace Inc. (San Mateo, Calif.) are winning converts, especially among corporations looking at enterprise deployment. Hummingbird Ltd. (Toronto, Ont., Canada), the new home for PaBLO and GQL, developed an application layer called BI/Broker. The layer can handle both thin- (BI/Web) and fat-client applications. Sqribe Technologies Corp. (Menlo Park, Calif.) retooled its desktop OLAP reporting products, aiming for “delivery to the masses.” Meanwhile, Seagate Software (U.S. headquarters are in Scotts Valley, Calif.) introduced an open OLAP architecture designed to give Crystal Info users a seamless “virtual cube” that integrates information from all the popular OLAP servers, including Microsoft’s.
OLAP vendor Gentia Software (Wakefield, Mass.) attracted business-level attention by repositioning itself as a supplier of Balanced Scorecard analytical applications. Sagent Technology Inc. (Mountain View, Calif.) rode its Data Mart Solution 3.0 to “turnkey” success in 1998. With a component user interface, the product is the engine behind Sagent’s “democratizing corporate data access” strategy. Tyrannical executives, beware. Sagent competitor Broadbase Inc. (Menlo Park, Calif.) unveiled an analytic application designed specifically for CRM. WhiteLight Systems (Palo Alto, Calif.) grabbed the spotlight with a hybrid OLAP approach that offered large database installations new flexibility. QueryObject Systems Corp. (Uniondale, N.Y.), formerly CrossZ, won fans with its distinctive, mathematically formulated data mart that enables flexible queries and on-the-fly dimensions.
“Data mining” as an earth-moving category has not been as successful as the specific technologies that fall under its classification. Led by Robert Grossman, Magnify Inc. (Chicago) has taken pattern matching to new levels for predictive modeling. Data mining is a critical part of the CRM movement. CRM is bringing data access and management companies into close partnership with Exchange Applications (Boston). Exchange markets Valex, a suite of front-office marketing automation tools that increasingly sit on top of huge warehouses. Builders of scalable warehouses running on parallel databases are turning to Torrent Systems Inc. (Cambridge, Mass.). Torrent’s biggest news was Intel’s equity investment; the two will collaborate on optimizing Orchestrate for Intel Pentium clusters and parallel processors.
Data Warehouse Management
Prism Solutions Inc. (Sunnyvale, Calif.), the data warehouse pioneer, brought server-side Java to its Prism Executive Suite. Users can write extraction and transformation applications and deploy them in any Java-enabled environment. Prism is using TimesTen’s in-memory database to speed up execution. One of Prism’s founders, Bill Inmon, has Pine Cone Systems Inc. (Englewood, Colo.) going in the right direction. The company’s Usage Tracker helps managers recognize usage and activity patterns—or more to the point, who is bringing the warehouse to its knees. Meanwhile, Vality Technology Inc. (Boston) continued to beat the drum for better data quality; it is a favorite partner of larger data warehouse vendors.
Acta Technology Inc. (Palo Alto, Calif.), 1998 startup, headed straight for the dynamic ERP market with metadata-driven tools specialized for extraction, transformation, and loading (ETL) of R/3 data. D2K Inc. (San Jose, Calif.) announced a close partnership with PeopleSoft and made its Java browser-based ETL tool, Tapestry, work with other ERP products. Ardent Software Inc. (Westboro, Mass.), which also markets the O2 database, had success with DataStage’s visual workflow paradigm for ETL. DataMirror Corp. (Markham, Ont., Canada) offered a strong transformation and replication solution for warehouses. Finally, Savant Corp. (Bethesda, Md.) won new customers and partners for its Q Diagnostic Center real-time performance-monitoring tool, which could be a good friend for performance-sensitive warehouses.
Y2K: The Winners
The Y2K problem has global corporations scurrying for tools and services that will help them beat the clock. Two enterprise vendors have answered the call, and will continue to do well after the millennium business shuts down. Compuware Corp. (Farmington Hills, Mich.) has a huge, experienced service organization to go with a successful tools division. Viasoft Inc. (Phoenix), long ago a proponent of reverse engineering, couldn’t get anyone to listen then; now it has both feet in the door with its Y2K development tools.
The Final Three
Back when the DB2 third-party market was screaming, three vendors stood tall. They have now headed in different directions, but all are enterprise fixtures. Candle Corp. (Santa Monica, Calif.) has found a strong market for messaging application tools to support IBM’s MQSeries and Microsoft’s MSMQ. As asynchronous messaging takes off, Candle will be in a good position. BMC Software Inc. (Houston) has a growing suite of automated tools for databases and ERP applications based on Patrol and its forward-thinking enterprise integration architecture. And Platinum Technology Inc. (Oakbrook Terrace, Ill.)…well, Platinum is everywhere! Its ProVision integrated management suite wires a set of best-of-breed tools into a coherent whole. Isn’t that the enterprise dream?
