A market is emerging for component-management tools to support business-reuse initiatives. Companies investing in the reuse of application components, models, specifications, documentation, or other application-development assets need an environment to manage these assets. Process is key to the success of reuse, and tools play a major role in helping the process work. Component management is critical to the maturity of development processes and achieving high levels of reuse.
However, growth of this market is slow because of economic conditions and the complexity of these products. Vendor offerings are expensive, unproven, and require consulting services to supplement them. IntellectMarket, one of the five early players, already is out of business, leaving four competitors--Aonix, ComponentSource, Flashline and LogicLibrary--to do the missionary work of justifying the steep investment. No vendor has more than a dozen paying U.S. customers. Most are still in the early testing stage and are unwilling to provide public references.
Despite these barriers, component-management environments and reuse processes are drawing interest. It's no longer a question of why reuse, but when to do it? The biggest barrier is justification: How can a company's IT shop justify a $100,000 to $300,000 expense for a back-end toolset that promises to lower costs and help improve the quality of future development projects? These products won't succeed without strong justification. However, if the vendors can successfully draw a set of metrics from early customers' experiences, this will become a viable market next year.
Are these products too risky? Not necessarily. Some risk is acceptable as long as the core investment is in the maturity of the processes around the tool. Much of the investment then will be preserved even if a company has to later abandon the chosen tool.
The needs of businesses have changed. Not only is there the desire to share internally created components and assets, but many teams hope to be able to buy components and share them internally.
Another factor is the increased diversity of application-development assets. Since Web-based applications are composed of code and content, text and graphics, various types of data, custom code, and third-party components, it's clear that the component-management environment must be able to define and manage true diversity. Metadata varies tremendously among these assets, as does the user community trying to access and share them.
Reuse shouldn't be approached as a process or, worse, a project in its own right. Reuse involves collaboration, focus on quality, and emphasis on well-built systems. IT groups that invest appropriately in development processes and underlying component architectures, tools, and organization should achieve reuse as a natural result. Therefore, introducing reuse implies modifying existing processes to include a reuse mentality.
Evaluating component-management tools is difficult. Return on investment and benefit models are weak and the tools are costly, certainly not an ideal combination. Without strong models to justify the benefits and value of a component-management tool, IT managers will have a hard time selling other executives on this level of investment. Therefore, IT departments should build justification metrics and models before, during, and after reuse initiatives. A model such as total economic impact could apply, as it focuses on benefits and risks, as well as total costs. Similarly, IT managers must push vendors to provide other customers' metrics and processes, and build a library of best practices.
Pay particular attention to the tool's ability to search and find components based on a wide range of criteria and metadata. This will be the most important feature for technical and nontechnical user success. Many tools can store a wide variety of application assets. Only those that can truly assist in understanding the company's asset base and determining which components might satisfy new requirements will succeed as reuse tools.
Demand customer references, incremental implementation, and a payment plan. These six-figure investments clearly can pay off in the long term, but these aren't shrink-wrapped environments. Evaluate related consulting services and partnerships along with tool capability. And accept that reuse implies changes in processes. Each company must develop ownership, development, and maintenance models when a reuse initiative starts. These models will drive the use of component-management tools and support reuse throughout the applications' life cycles.
Robert K. Weiler is the chairman, president, and CEO of Giga Information Group, a global technology advisory firm. You can reach him at [email protected]. Liz Barnett, a Giga VP and research leader, also contributed to this column.