Concepts

Current software design concepts for products, processes, projects, programs, portfolios, and research policy overlook a simple but fundamental idea: the goal of software design decision making is to create the maximum value for any given investment of resources. Although the concern is economic, the technical implications are tremendous. We propose to relate software design decision-making concepts, as well as methods and tools to the maximization of economic objectives, with particular attention to strategic approaches to value creation under uncertainty, incomplete knowledge, competition, requirements for cooperation of self-interested stakeholders.

This is a challenging goal. Software design decisions today are made in an economics-independent “Flatland,” where concerns for technical properties dominate. Past work on software economics is relevant but it focuses on cost minimization, rather than value maximization. Yet, as the PITAC report stated, the vast bulk of resources is spent bringing products rapidly to market. Time is often far more important than direct cost today. More generally, value added—not cost, benefits, technical perfection, or any other isolated property—is the arbiter of success. When software contributed mainly to off-line, back-room functions, this situation was not particularly harmful. However, software design decisions are now intimately intertwined with major business and public policy issues in almost all fields of human endeavor. As enterprise success comes to hinge on software design, driving design decisions for value creation is critical. $85 billion was sunk in cancelled software projects in 1998 in the U.S. alone. Value-based decision-making that fostered earlier abandonment of such projects could save tens of billions of dollars annually.

The value maximization axiom is simple to state, but the research challenges are deep. First, enterprise-level value is not a simple idea. Businesses value profit, but also opportunities, as seen in valuations of profitless Internet companies. Philanthropic foundations value solutions to social problems. Universities value creation and dissemination of knowledge. Second, value is not created with certainty in a static world, but has to be sought under uncertainty, incomplete knowledge, competition, change and the need for the cooperation of self-interested stakeholders. Third, sophisticated concepts, models, metrics and analysis techniques have been developed, largely in finance, but also in political science, for creation of value under such conditions. Of particular importance is the emerging appreciation of the need for strategy in design: of products, processes, and multi-project programs; portfolios structured to diversify risk; and research and development investment policies. Software phenomenology is uniquely complex and still not well understood. Models developed for finance and other such domains cannot merely be imported and exploited directly. New syntheses of existing software engineering concepts and methods with the concepts and methods of strategic investment management are needed.

The PI and co-PIs therefore propose to develop and evaluate a spectrum of new concepts, models, and tools for software design based on the idea that design decisions should be traceable to explicit value maximization goals and strategies. Technical concepts to be formulated in these terms will include software architecture, resource budgeting to software verification activities, lifecycle models, tools and environments for management of software design for lifecycle value maximization, and actuarial foundations in software phenomenology for markets in risks of use of third-party components. The industrial evaluation and transition to practice of any advances will be enabled by incorporating them as experimental modification of the MBASE lifecycle models.

Value creation has the potential to become a simplifying and empowering “unified field” concept for software design. The PI has shown how concepts as diverse as information hiding, spiral models, and timing of design decisions can be unified in terms of the value of options. A co-PI established the field of software economics. Success in this work could have an enormous impact on software theory, practice and, importantly, on teaching. Creating value through software development depends on people who understand both software technology and value.