 |
    
|
Filing Information
July 1997
IDC #13848
Volume: 1
Tab: Vendor Strategies
The market backdrop for this product strategy was interesting. Developers in the aerospace and defense industries were typically the first to encounter increasing complexity in the software they were being asked to produce. In the early 1980s, the Department of Defense sponsored the development of Ada as a standard language for use in defense-oriented applications. The intent was to apply a single development platform that would incorporate sophisticated and modern software engineering concepts to these application development challenges.
By the early 1990s, however, object-oriented software development techniques began to take hold as organizations were increasingly required to develop software more quickly and with greater reliability. C++ was particularly well suited for this new paradigm, especially for those developers who were using C (C++ is an object-oriented extension of C; therefore, it provides a logical migration path to object orientation).
With this market state as a backdrop in 1994, Rational completed a merger with Verdix, another company that provided software development tools. Verdix's focus was entirely on Ada development; specifically, compilers, cross-compilers, debuggers, test tools, and configuration management systems. The intent of the merger was to strengthen the position of the combined company in the Ada market. This would be accomplished with a product line that was composed of complementary products from both companies, using the combined strength to dominate the object-oriented development tools marketplace. This latter goal was an important one because the Ada market was maturing and growth was slowing. Of equal significance was both companies' strength in the Unix market, in which the vast majority of both Ada and C++ development was taking place.
Rational's primary market focus over the past few years has been to provide tools that focus on the visual modeling of applications using objects. IDC describes these tools as falling into a category called object-oriented analysis, modeling, design, and construction (OOAMDC). Rose is the name of the company product that addresses the needs of developers who use these OOAMDC tools. Figure 1 shows Rational's revenues for the past five calendar years. Not only has its total revenues increased significantly from year to year, but the percentage of its revenues reflected in OOAMDC tool revenues has steadily increased. From less than 8% of its revenues in 1992, it has grown to more than 34% of its revenues in 1996. As objects and object tools have begun to establish themselves in the marketplace, Rational has not only kept pace, but led the way in the past two years.
IDC has forecast this market to expand at a 1996-2001 compound annual growth rate (CAGR) of about 44%, driven primarily by the accelerated adoption of object technology. This adoption has been driven recently by the IT market's focus on the Internet and the related technological focus on component software. Both of these technologies have, in their own way, brought the previously abstract and very different object concept to a level at which most organizations can grasp and appreciate its benefits. Real world problems can now be addressed more easily and efficiently - a critical step to stimulating adoption by more mainstream IS organizations.
In fact, it is IDC's view that most of the growth in object technology usage in general, and OOAMDC tools in particular, will occur in the previously unexploited business applications development arena. Rational is one of several companies that has recognized this emerging trend and addressed in very significant ways the needs of this market sector.
In the past three years, Rational has made a number of very aggressive moves aimed at improving its position with organizations in this community. These moves have occurred along several fronts, each designed to ensure that a key requirement of IS users is met.
The major standards effort in this area has centered on methodology. OOAMDC tools - and for that matter, the more traditional nonobject-oriented analysis, modeling, and design tools - usually incorporate a formal methodology that encompasses both notation (visual and textual) as well as process. These methodologies vary in both rigor and content, but their goals are the same: to formalize the notation and process used to model, design, and implement software applications. Some tools take a "meta-model" approach to analysis and design, allowing modeling to be accomplished using notations corresponding to two or more methodologies.
Originally, Rational Rose used the Booch methodology, named after its inventor and Rational employee, Grady Booch. The methodology relies on an iterative approach, whereby the domains of analysis and design are explored in such a way that the activities interact and iterate so that there is no clear distinction between analysis and design. This type of methodology is also classified as elaboration.
Figure 2 shows the 1995 market shares of the various methodologies available in the market, based on sales of tools in that use the respective methodologies. At that time, the Object Modeling Technique (OMT) was the most popular methodology. OMT shared many similarities with Booch, including elaboration and some more detailed concepts. OMT was first described in a book published in 1991 by James Rumbaugh, who was with the General Electric Advanced Concepts Center (now part of Lockheed-Martin). In October 1994, Rational hired Rumbaugh and articulated its intent to modify Rose to support both OMT and Booch and to merge the two methodologies into a single, unified approach. The race for a standard had begun, and Rational intended to establish an early lead and stay there. The data in Figure 2 clearly shows that more than 50% of the market as measured by methodology was either Booch or OMT focused. Although the plethora of vendors supporting OMT did not allow Rational's acquisition to translate to immediate sales dominance, it created a perception of technology ownership that would prove to be invaluable.
Source: International Data Corporation, 1997
The Object-Oriented Software Engineering (OOSE) methodology was invented by Ivar Jacobson, who founded a company in 1987 called Objectory AB to build a tool to support this methodology. OOSE is an elaboration methodology that focuses on "business engineering," an important component of any analysis and design tool that could realistically be adopted by many mainstream IS organizations. The capability to partition the system requirements according to the uses of the system (known as "use cases") was something both of the other methodologies needed to provide for the development of business applications. Thus, Rational acquired Objectory AB in 1995 and then created a triumvirate of methodologists focused on merging their work into a single approach.
The result of their efforts, which is known today as the Unified Modeling Language (UML), has been strongly marketed by Rational; it is difficult to find an elaboration-based OOAMDC tool that does not support UML to some degree. It was Rational's goal to establish UML as a de facto standard and to pursue its acceptance by recognized standards bodies.
The Object Management Group (OMG), a consortium of more than 700 software vendors whose charter is to 'promote the theory and practice of object technology for the development of distributed computing systems,' has established an Object Analysis and Design Task Force. The task force has issued RFIs and RFPs aimed at establishing standards for tools that support object-oriented analysis and design as well as to define semantics and a metamodel standard for this technology. Its goals reflect what its members already know regarding the steps necessary to achieve significant IS adoption levels.
Rational has been extremely active in this process and was one of the original four major response submitters. However, the already strong industry support for UML has made it almost a certainty to be a major part of the standard. Additional support by some strong partners (e.g., Microsoft, Oracle, and Texas Instruments) have given UML an even more potent presence.
Rational has accurately read the need for a methodological standard and has carefully worked to establish high credibility in terms of methodology coverage, tool support, and championing of a standard.
To this end, Rational has taken some steps in the form of acquisitions and partnerships. Indeed, vendors in this market often align themselves with others that can provide complementary functionality because it is a less expensive approach and because there is a significant user base of the complementary product that can be leveraged. The first step, the acquisition of Objectory AB, added cases for modeling business processes and flows, and its integration with Rose and the UML should mean a seamless, highly automated implementation of front-end business process and object modeling.
In November 1996, Rational and SQA Inc. announced their intent to merge the two companies, with all outstanding shares of SQA's common stock to be exchanged for Rational's shares. SQA has been recognized as a leader in the automated testing of Windows client/server applications. Its SQA Suite product is a leading solution for testing client/server applications across the three Windows platforms (Windows 95, Windows NT, and Windows 3.x). In addition to SQA Process, which provides a formal methodology for automated testing, the suite is composed of the following three products:
The suite is built upon a scalable client/server repository based on the Sybase SQL Anywhere database. The repository allows for the integration of data from all SQA users on a network, thus providing a consolidated view of the testing process.
In January 1997, Rational made another significant acquisition of Requisite. Rational has integrated the Version 2.5 release of Requisite's RequisitePro requirements management tool into Rose and SQA Suite. This integration is significant because it gives developers the ability to test and manage applications and their development, from the generation of the requirements through implementation. For example, the RequisitePro/Rose integration means that users can trace end-user and business requirements, specifications, test cases, and other entities that exist within RequisitePro to related use cases and packages within Rational Rose. A visual "link" is displayed to the user within RequisitePro when one or more of its entities change, showing the user which use cases and packages should be investigated. Thus, the effects of a requirement change can be characterized quickly. This process can also work in reverse. That is, changes in use cases or scenarios in Rose can generate requirements entities in RequisitePro. The RequisitePro/SQA Suite integration means that test requirements can be generated from RequisitePro; test procedures are generated from Rose use cases and scenarios. The integration achieved by Rational through these three products forms the basis for the industry's best example of full life-cycle modeling, implementation, and testing.
Rational's most recent move came in April 1997 when it announced its intention to merge with Pure Atria. The latter company was formed from a merger in June 1996 of Pure Software, a leading vendor of automated software quality products, and Atria Software, arguably the market's most well-known vendor of software configuration management tools. The merger resulted in a company that could offer a comprehensive set of tools for managing the software development process and ensuring high-quality results. These tools fall into the following two major product categories, as defined by Pure Atria and roughly conforming to the businesses represented by each of the original organizations:
In one sense, components represent the original intent of objects: to make software development easier, faster, and more reliable, and to take advantage of the concept of the reuse of elements that are "tested in the field" (a potentially major reliability factor). Components can speed development because needed functionality is developed only once and then reused. If an organization is intelligent about component design up front, reuse can approach completeness, with little or no modification required.
Managing the development of large, complex applications using components requires tools for visualization and for management of both the assets (components) generated and available and of the development process itself. Rational, which was already the leading vendor of visual modeling tools for object-oriented development, has correctly determined that future success will hinge on its ability to extend its Rose product line to support the design of components, the management of component assets, and the assembly of applications from components. Thus, not only has Rational added great breadth to its life-cycle support through the aforementioned merger, acquisition, and licensing activities, but it has brought the acquired technologies to bear on this very new and important market focus.
A visual representation of Rational's approach is shown in Figure 3. The core product is Rational Rose, the company's visual modeling tool, from which all other Rational products (those acquired and those developed internally) are or will be integrated to provide all of the capabilities in the areas of requirements management, automated software quality, process automation, and software configuration management. The current Rose Version 4.0 is equipped with full UML support (including use cases) and a user interface attractive to the Windows user.
Rational approaches component-based development from the standpoint of a traditional OOAD vendor, which is to extend an already robust visual object modeling capability to include support for component development and application assembly from components1. Other vendors that do not have an OOAD tool history fall into a recently created market that IDC calls "component construction and assembly." These vendors will often partner with other vendors that provide UML-based visual modeling support in order to serve those customers who require a standards-based visual modeling capability. With extensive experience in the visual modeling of objects, Rational is one vendor that can provide it all.
IDC strongly believes that Rational soon needs to articulate and demonstrate the roles each of its products will play in its market strategy, and that Rational will in fact be in a position to do so. It may be that some products disappear. Others may offer a solution on one set of platforms, with a similar product offering a solution on others, or similar products can be part of different "levels" of the solution (e.g., "learning editions" versus "professional editions"). Whatever the result, Rational has consistently demonstrated a strong ability to adapt and eventually lead. In IDC's view, this situation should not be an exception.
Rational's relationship with Microsoft must be viewed as a very important component to its long-term success. As a traditional OOAD tool vendor, Rational's past is rooted in the technical application development market and on Unix platforms. The Microsoft relationship has given Rational a strong foothold for its Rose product with the large and growing installed base of Microsoft developers - those using the visual development tools Microsoft has packaged into Visual Studio and who develop on Windows platforms. Combined with the joint marketing and product distribution partnerships between the two companies, Rational is in a position to move its dominance of the object-oriented and component-based software development markets to the Microsoft world.
IDC believes that Rational has both the product backbone and marketing prowess to continue its market leadership and very strong revenue growth through at least the next few years. The market for tools that aid the development of applications using components will soon become very competitive, with competitors coming from the ranks of traditional OOAD and CASE competitors, vendors of 4GL and RAD tools, vendors of IDEs for specific environments, and even system vendors. Rational has a clear advantage in terms of experience and mindshare.
Rational's challenge will be to continue to move quickly to adapt to changes in focus and technology. One cannot assess Rational's prospects for the future without considering its traditional technical base. Although the company claims that more than half of its new business is based on product sales and services to mainstream commercial users, Rational has a "legacy" of customers who do not fit into this category. The challenge will be for Rational to enhance its products to meet the needs of both types of developer and to maintain the resources required to do so. With smart management, Rational can leverage its traditional installed base and, using the profits that result from servicing this legacy base, continue to finance its aggressive move into the growing commercial market.
Copyright 1997 International Data Corporation. Reproduction is forbidden unless authorized.
For additional copies please contact Cheryl Toffel, 508-935-4389 or E-mail [email protected]
| Questions or comments regarding this service? [email protected]
Copyright © 1998 by Rational Software Corporation. All rights reserved. |
