People tend to think of quality as a final result or destination. It is not; it is a journey that never ends. As you measure and manage quality, you will learn more about it. Then each improvement step will provide the knowledge, experience, and data needed for the next step. Focus on continuous improvement and help your team to truly believe and follow the principles of quality management. Since each person's needs are different, you must recognize where each of your developers is on this journey and help everyone take the next step. The steps in the quality journey are as follows.
1. Test and fix. At first, the focus of almost all software groups is on getting products to work. The developers' objective is to get the product into test as quickly as possible, and then to test and fix (and test and fix and test and fix and . . .) until it works sufficiently well to ship to the users. At this stage, the only way that the members know to improve quality is to spend more time and money on testing. Your challenge is to move the team as quickly as possible to steps 2 through 8 of the quality journey.
2. Inspect. The next step is when the developers and managers start removing defects before test. This is usually done with various kinds of walkthroughs and inspections. The typical challenge in this step is to get the developers to do all of the required inspections and to do them properly.
3. Partial measurement. As inspection programs mature, some groups begin to measure and use inspection data both to improve the inspection process and to focus the inspections on the most defective product elements. The challenge is to get adequate data and to use these data to improve the products.
4. Quality ownership. As they participate in team inspections, developers may become more sensitive to the mistakes that they make and start reviewing their personal work in advance to eliminate as many of these problems as they can. Once developers reach this point, the quality of their products will quickly improve.
5. Personal measurement. To know how to improve the quality of their personal work, developers need objective data. The required data concern the defects they personally inject and remove, the sizes of their products, and the time they spend. The challenge is to get them to gather and use these data. As they examine the data on the defects that escaped inspections, testing, and the final user, product quality will again increase sharply.
6. Design. Once developers have learned to manage their coding defects, they can focus on design defects. This requires precise and well-defined design practices and sound design verification methods. The challenge is to use sound design methods for all programs—large and small— and to use sound design verification methods in all design inspections and personal design reviews.
7. Defect prevention. While using sound design and measurement methods will reduce defect injection rates by about two times, effective defect-prevention programs follow a structured procedure to identify process problems and make the changes needed to eliminate even more defects. The challenge is to get the defect-prevention program initiated and then to sustain and broaden it to cover the full product life cycle.
8. User-based measurement. Ultimately, the quality program should be driven by user-based quality measures. The principal challenge here is to understand the quality characteristics that are most important to the users and to measure these characteristics in a way that is meaningful both to you and to the users.
An Alcoa executive once invited me to visit one of their plants that manufactured sheet aluminum. He said it produced the highest quality aluminum in the world. In talking to the engineers, I was surprised to find that their quality measurements weren't about making aluminum—they concerned making cans. For example, the thicker the aluminum, the more cans cost. However, the thinner the aluminum, the more likely it was that defects in the aluminum sheet would cause an expensive and time-consuming "punch-through" that would interrupt production. Alcoa was the market leader because their quality was so good, cans made with their aluminum were thinner.
The reason that the quality journey is never ending should be clear from this example. As long as technology advances and as long as it attracts newer and different kinds of users, we will face new quality needs. The principal message from this eightfold quality journey is that it must be traveled in steps. Until the developers have made reasonable progress with step 5, they will not have the data to support steps 6, 7, or 8. So, while you should take the long view, keep your team focused on the next step in this quality journey.
Was this article helpful?