Software today is a dominant force in enabling companies to exploit new distribution channels, create new products, and deliver differentiated value-added services to customers. The backbone of industries like banking, airlines, and publishing, it's also an increasingly important value-adding component of consumer products such as television sets, cameras, cars, and mobile phones. In fact, the amount of software code in most consumer products is currently doubling every two to three years. Driven by escalating demand for functionality and rapid development of enabling hardware technology, this pace can only accelerate - and with it, the pressure to develop bigger and more complex systems that can meet the ever-increasing demands for higher quality and stronger performance.
Organizations scrambling to cope with this challenge face two major hurdles. First, there is a need for a better execution of software projects. American Airlines, for example, won a reputation for IT excellence during the 1980s with the launch of its SABRE airline reservation system, yet fumbled a 1992 attempt to build a reservation system for hotel and car rental companies - taking a $165 million charge due to the project's failure. The Internal Revenue Service ponied up $4 billion to develop modern computer systems, only to concede in 1997 that the systems "do not work in the real world."
Such stories are not uncommon. Software projects are notorious for overrunning budget plans, in terms of both cost and timelines. In fact, the average software project overshoots its schedule by half, according to industry experts, who also observe that for every six new large-scale software systems put into operation, two others are canceled.
The second hurdle is the need for organizations to derive greater business value for their key stakeholders from software. Research has shown little or no benefits to organizations from recurrent investments in information technology, even as delivered computing power in many developed economies has more than doubled over the past two decades. This "productivity paradox" points to a fundamental misalignment between the software and business strategies of many firms.
Models for assessing and improving software development processes formed during the past decade typically address issues such as whether an organization has appropriate software project-management procedures in place and that the right metrics are being collected and used for managing software production. Our research sought to develop a model to assess the strategic leverage of software in a broader organizational context. We defined "Software Excellence" as the ability of an organization to create the right organizational context for developing and applying software and to derive appropriate business value for all key stakeholders - shareholders, end-users, software-related employees, and the organization at large.
We then defined 10 aspects that determine excellence and grouped these into two categories: those that focus on how an organization creates the appropriate context for software excellence, termed "enablers," and those that focus on the value accrued from software, which we called "results."
Enablers consist of:
* Leadership. The role of senior managers in creating and driving a culture of software excellence.
* Policy and Strategy. The role of software excellence in the organization's values, vision, strategic direction, and the implementation of its policy and strategy.
* People Management. The management of software-related employees to release their potential to improve the business.
* End-User Management. The nature, extent, and effectiveness of partnerships created between the software-producing unit and the software's end-users.
* Resource Management. The software-producing unit's (SPU's) management, use, and preservation of financial and nonfinancial resources.
* Processes. The identification, management, and continuous improvement of all key software-production processes.
* People Satisfaction. The organization's success in meeting the needs and expectations of software-related employees.
* End-User Satisfaction. The SPU's success in satisfying end users' needs and expectations.
* Impact on Organization. The degree to which the SPU has a beneficial impact, such as by stimulating innovation.
* Business Results. The SPU's success in making the appropriate contribution to financial success or other goals.
In practice, the Software Excellence model indicates ample room for improvement in companies' efforts to create more value for stakeholders from investments in software. We applied the model to 85 European organizations, using the results of a questionnaire-based survey to assign each respondent a score comprised of up to 500 points in each of the two categories - Enablers and Results - for a maximum total score of 1000. The average score across all respondents came to 473 out of a possible 1000, with 64 percent of all respondents scoring below 500.
A strong association between the two categories suggests that investment in the Enabler categories is generally associated with increased value in the Result categories. Banco Commercial Portuguese, for example, adopted effective IT management early on with an articulate vision, a board that participated in hardware and software decisions, and tight integration between IT and business staff. The result was a series of innovations that improved service and drove success. "At other banks, it is the data-processing department that defines the information system," says Jardim Goncalves, chairman of the bank's board, "In BCP, it's the users that decide it."
Furthermore, high-scoring firms tended to score higher for all management practices than their low-scoring counterparts. Results scores also indicate that different Enabler categories play a role in translating investment in software creation and deployment into meeting financial and nonfinancial goals, with these relationships suggesting the following best practices:
1. Management leadership is critical. The senior management of high-scoring organizations were more intimately involved in setting the basic strategy and mission of the SPU and took more active steps to instill a culture of software excellence. Yet only about half of respondents reported that the strategy and mission of the SPU is set personally by senior management, with a large number stating that progress is hindered by a lack of commitment from senior management and failure to take appropriate follow-up actions.
2. Create multiple links with end-users for effective partnerships. High scorers tended to have personnel from all levels and functions actively involved in partnerships with end-users and were far more likely to have formal processes for obtaining regular feedback from end-users, as well as systematic reviews and updates of the scope and coverage of the partnerships. Yet, most respondents did not see partnerships created between software-related personnel and their end-users as effective and reported that procedures for obtaining regular feedback on relationship were weak.
3. Focus on keeping software-related personnel motivated and happy. Most firms neglect to regularly measure factors, such as staff turnover, that influence or predict personnel satisfaction, and few have adequately linked career development plans for software-related personnel to the company's business plans. High-scoring organizations fared only slightly better than average on this front, tending to give slightly greater emphasis to personnel empowerment and to more actively seek out the perceptions of software-related personnel.
4. Manage software production processes more rigorously. While project-management procedures are common among most respondents, few reported systematic processes for evaluating and managing project-related risks. Use of metrics for managing software processes is poor overall, with little emphasis given to the systematic reuse of software components. High-scoring companies stood out by having rigorous processes by which project risks are evaluated and taking steps to manage these risks continuously. They also collect and analyze a more extensive set of software-related metrics to improve project performance.
Soumitra Dutta is associate professor of information systems at INSEAD in Fontainebleau, France; Luk Van Wassenhove is professor of operations management at INSEAD.
|Printer friendly Cite/link Email Feedback|
|Publication:||Chief Executive (U.S.)|
|Date:||Jan 1, 1998|
|Previous Article:||Working on the chain gang.|
|Next Article:||The P&G of prisons.|