Introducing a quality process calls for a change in behavior by every person on the payroll, so the need for training is absolute. It isn't fair to expect a particular level of performance if training hasn't been provided. That's true in family life, in sports and in business.

Also remember that quality-specific training should never replace technical training. It should be an addition to the training budget, both in terms of money and time. This column focuses on quality-specific training at the outset of a quality process; technical training is assumed.

When picking a consultant to help establish a training program, the single most important thing for an organization to do is retain control of the process. This may introduce some tension. The organization is, after all, concerned with its future while the consulting firm is naturally concerned primarily about its own future. What is best for the latter (e.g., big sale, prolonged engagement) may not necessarily be best for the former.

When interviewing consulting firms, a company should beware of statements along the line of, "This will take a long time" and "We'll need to train everyone in the company." Statements like these made by the large quality consulting firms in the late 1980s and early 1990s are a major reason why so many business executives shy away from the word "quality" these days.

Two simple rules can guide a company through the jumble of ideas and options presented by competing training consultants: An organization need not train everyone upfront and, once the initial round of training is complete, the organization should buy the "customized" program. If the consulting firm won't sell the internal-use-only rights to the program, the company should find another vendor.

On not having to train everyone. Suppose, for a moment, that an organization's quality process has just started and a brand-new quality team is meeting for the first time. Imagine further that all team members have received one or two overview presentations from senior managers, explaining--in broad strokes--what this quality process is, why the company is investing time and energy in it, and the hopes and goals for it.

For this meeting, only one person must know how to run it: the team leader. Also, only the team leader needs to know problem-solving procedures and how to use basic SPC tools. The same holds true for the other major components of an initial training program, such as sound grounding in the theoretical foundation for quality and a working knowledge of the mechanics of the company's quality process.

This will probably add up to about 10 percent to 15 percent of the employee base (assuming 100-percent employee involvement, and how can any organization justify anything less?) plus the senior executive team. Of course, training only the team leaders requires an organization to assume that employees know their own jobs and can take part in productive conversations. In the course of the coming months, team leaders will pass on much of what they have learned at the team leader course.

Reducing the approach from an everybody-into-the-pool, assembly-line training program to one focused specifically on the people most vital to the process's success not only brings some logic to the effort, it saves huge amounts of money and time. Several years ago, a large telephone company on the East Coast bought the "train everybody" approach from a major quality consulting firm. After three years and $10 million, they finally had everybody trained, but, by then, the process was riddled with weaknesses and, in fact, collapsed within months. Efforts that could have been used to better define the process had all been spent reaching the "train everybody" goal.

Once the outside consultants have taught the initial course to the first round of team leaders, the course must be brought in-house. The appropriate cost? Whatever seems like fair value to both parties.

The reason for taking control of the course matter is simple: If the course is left in the consultants' hands, next year's team leaders will receive the same course, at the same or higher cost. If, however, the organization takes control of the course, it can be truly customized during the following months by adding stories about organizational successes and challenges, clarifying those sections that nobody understood or used, and adding changes recommended by team leaders. When the next round of team leaders takes the course, they'll be receiving better material than the first ones did.

The consulting firm is, of course, welcome to come back and attempt to sell the organization more courses. Quality training should be an ever-expanding field, as employees become more aware of possibilities and as the competition raises the bar. And, if the team leader course proves to have lasting value, the firm's additional courses will be looked at carefully.

Is it all worthwhile? In her excellent book, The Corporate Guide to the Malcolm Baldrige National Quality Award, Marion Steeples reported that in 1991, Motorola spent $150 million on training, with 40 percent of that devoted to quality-specific training. An independent audit verified that the return on investment for that $60 million was 30-to-1.

Training is expensive. It will be the most expensive single cash outlay in the course of implementing and maintaining a quality process. It can, however, be controlled and, if justification for the expenditure is needed, a 30-to-1 ROI is a pretty good place to start the discussion.