NEWS AT SEI
This article was originally published in News at SEI on: September 1, 2003
This is the third in a series Mcolumn in March discussed some general principles of programming, with particular emphasis on the changing and ill-defined nature of software requirements. The second column in June addressed those software principles that relate to the nature of our products. These principles concern the fact that our products are intangible, can last essentially forever, and are increasingly important to our businesses and to society in general.
In this column, I discuss the principles that relate to software engineering projects. Many of these principles are common to engineering projects of all kinds, but software projects present some issues that make our work uniquely challenging and rewarding. In discussing project principles, it is important to start with the fundamental purpose or objective of most software projects. This is to develop or enhance a product to meet a business need. In fact, this defines the following important principle about almost any software project.
The principal objective of almost all software projects is to meet a business need.
This means that the schedule, cost, and quality of the work is of paramount importance. Therefore, in addressing the principles that govern software projects, I will talk about schedule, cost, and quality. Of course, by quality, I refer to the ability of the product to reliably produce the user’s desired results. While there are many other important project considerations, they all relate directly or indirectly to schedule, cost, and quality performance. In closing, I will comment on the benefits of successful projects and the characteristics of project success from both a business and an engineering perspective.
Project Schedule Performance
Project schedule performance has three interesting characteristics. First, with few exceptions, if you don’t meet the committed schedule or a revised schedule that everyone knows about and has previously agreed to accept, your project will not be judged fully successful. In other words, schedule performance comes first. For example, in assessing the best projects, any that are late, even by only a few days or weeks, never make it to the top of the list.
Second, and particularly for projects that last for more than a few weeks, the important stakeholders need to know where you stand and if you are likely to finish on time. The end users need to make installation and conversion plans, the testers need to schedule their resources and facilities, and management needs to know if there will be any business problems or if they will have to step in to accelerate or redirect the work.
Keeping managers and customers properly informed requires accurate and timely status reports. This is the most common area where software projects run into difficulty. Since software engineers rarely know precisely where they stand against their schedules, they cannot make a convincing report on their status or accurately forecast when they will finish. From a management perspective, this proves that they do not know how to manage their work. This leads to distrust, management meddling, and often even to project redirection or cancellation. In fact, I have seen management interference destroy projects that otherwise would have been reasonably successful, all because of poor status-tracking and reporting.
Third, schedule performance by itself is not personally rewarding. While it is essential to meet other people’s criteria for project success, the satisfaction that comes from meeting a schedule is ephemeral. After a rather brief period, management’s reaction will become “So what was the big deal?” You just did what you said you would do. Consider schedule performance like a down payment: it is essential to get into the game and it will improve your personal reputation with management, but, by itself, it will not produce lasting rewards or personal satisfaction.
Project Cost Performance
Most engineers feel that if they meet the schedule, any cost overruns will be small and not worth worrying about. But that is becoming less and less true. As many software projects last longer and become more expensive, both cost and schedule management are increasingly important. To see why, suppose that you were building a new house and that the builder assured you that he would finish on schedule. Then, just before the final closing, he told you that your changes had cost more than expected and that he had to pay some overtime to finish on the promised date. The bill is therefore $50,000 more than you had previously agreed. This would likely cause a serious problem. The mortgage commitment probably would not cover the added costs and you probably don’t have that kind of money lying around. While meeting the schedule was nice, the cost overrun could easily be a deal breaker.
Cost is equally important for software work. However, the time to address cost problems is when you first detect them, not at the end when no one has any flexibility. If your customer wants a change, if you have technical problems, or if your original estimates were way off, you should figure out what the job is now likely to cost and get agreement before you proceed with the work. While this will involve lots of nitty debates during the project, it will avoid the big cost surprises at the end. When you are at it, also make sure that you put all of these cost negotiations and agreements in writing.
Of course, the problem that cost management presents for most software projects is that we rarely know enough about the costs of our work to estimate the impact of small changes. This again is a fairly unique problem for software, but it is a problem we must learn to address if we are ever going to effectively manage the costs of our work.
Project Quality Performance
While cost and schedule performance are important, our product must work. Also, cost and schedule performance are not, in the long run, satisfying from an engineering point of view. We like to build great products. If you finished development on schedule and within planned costs but the product was a disaster, the brief glory you got from delivering on time would quickly fade. While being known for meeting schedules is important, being known as the developer of a poor-quality product is an engineering kiss of death.
The world is changing and the importance of delivering quality products will only increase. After they have lived through a few software disasters, many software developers properly conclude that it is better to deliver good products late than to produce poor products on time. This strategy, however, will continue to have business problems. The engineers who get ahead in the future are almost certain to be those who deliver quality products on time and for their committed costs. If no one in your organization can consistently do this, there are lots of organizations all over the world that would love to take your place. Many of these organizations are already demonstrating the ability to do superior work on schedule and they are growing very quickly. So, if you want to keep your job, it would be a good idea to learn how to meet both the business and technical needs for your products.
Project Success Criteria
The real satisfaction we get from our work is the thrill of working on a great team, creatively using the latest technology, and producing superior products that truly meet our users’ needs. As engineers, we are creators and we want our creations to be accepted, used, and appreciated. This requires quality products. In short, the truly successful and rewarding projects of the future will be those that produce quality products on schedule and for their committed costs. These are the only projects that will consistently satisfy the engineers, their managers, and the users.
The criteria that engineers and managers use for judging project success historically have been very different. As I will discuss in the next column, it is important that we learn to consistently meet both management’s success criteria as well as our own. The key point to remember, however, is that management decides who to hire and how much to pay for our work. This means that we must meet management’s criteria if we want to get ahead. In the next column, I will conclude this series on programming principles with a discussion of the perceptions that engineers and managers have about project success and what this means for each of us.
In writing papers and columns, I make a practice of asking associates to review early drafts. For this column, I particularly appreciate the helpful comments and suggestions of Dan Burton, Don Firesmith, Marsha Pomeroy-Huff, Julia Mullaney, Bill Peterson, and Alan Willett.
In closing, an invitation to readers. . .
In these columns, I discuss software issues and the impact of quality and process on engineers and their organizations. However, I am most interested in addressing the issues that you feel are important. So, please drop me a note with your comments, questions, or suggestions. I will read your notes and consider them when planning future columns.
Thanks for your attention and please stay tuned in.
Watts S. Humphrey
About the Author
Watts S. Humphrey founded the Software Process Program at the SEI. He is a fellow of the institute and is a research scientist on its staff. From 1959 to 1986, he was associated with IBM Corporation, where he was director of programming quality and process. His publications include many technical papers and several books. His most recent books are Introduction to the Team Software Process (2000) and Winning With Software: An Executive Strategy (2002). He holds five U.S. patents. He is a member of the Association for Computing Machinery, a fellow of the Institute for Electrical and Electronics Engineers, and a past member of the Malcolm Baldrige National Quality Award Board of Examiners. He holds a BS in physics from the University of Chicago, an MS in physics from the Illinois Institute of Technology, and an MBA from the University of Chicago.
The views expressed in this article are the author's only and do not represent directly or imply any official position or view of the Software Engineering Institute or Carnegie Mellon University. This article is intended to stimulate further discussion about this topic.