A CEO's view: lead by example, and respect your people.

Paul H. O'Neill Chairman of the Board and CEO Alcoa

How does a company achieve an international competitive edge? At Alcoa, we think the key lies in quality. And quality means systematically taking apart the ideas of strategic analysis, strategic planning, and operational planning so that each is done separately.

As part of our quality process, we want to benchmark everything we do against the best practice in the world. So our strategic analysis begins with collecting and organizing data about what the best practice is. We know, for example, that the number of days an aluminum company can go before relining a smelting pot varies from 1,200 to a world's-best 3,200. Thus far, we've collected a body of data that tells us where everyone in the industry stands, what each company's experience is, where Alcoa stands in comparison, and what the fundamental economics of success or failure in each part of the business appear to be.

We put this information in the context of supply and demand, and add to it the degrees of freedom that Alcoa and its competitors have to expand into areas with good economic merit. So, on the front end of our quality planning is strategic analysis, taking into account not only our material's operating characteristics but our customers' alternatives. When we look at the aluminum can, for instance, we consider the customer's alternatives in glass, plastic, and steel, so we don't miscalculate the rate of productivity improvement Alcoa requires. In this particular case, we want to keep steel out and pursue the concept of replacing glass and plastic, taking into account the environmental and energy aspects, so that in the next step - strategic planning - we can begin to define the alternatives.

From strategic planning flows operational planning, which links where we're spending our money with where we're allocating our human resources to give us the best economic positioning. This process is complicated. When we first began, we had a difficult learning period because people weren't used to thinking about their work in such an orderly way. But now we're beginning to see the benefits of tying together the rational analysis process with the day-to-day operating plan. And every step of the way we need to emphasize quality in five key areas: safety, management, return to shareholders, compensation, and repairs.

1. Safety. Surprising to some in the financial community, I'm convinced a quality program begins with safety. It is not a given that people will be injured in the work-place. Achieving this standard of excellence isn't easy, but it is possible if you really understand your process and you operate under control all the time.

To me, safety is the primary indicator of how well a company can lead globally. Alcoa, for instance, has the best safety record in the aluminum industry. Despite that record, we continue to stress the importance of safety, and we've managed to further reduce our serious-injury rate by 25 percent in the past three years. I think we can reduce it even more. It's not easy, though, when employees exhibit so much habitual behavior and, in my perception, a kind of Western macho-ness that enjoys "riding in the fear of danger." You have to convince your people that that kind of behavior is not in their own interest.

We have a practice, called the Core Program, that is a systematic safety plan. Under it, we send groups of employees to visit companies like DuPont, Dow, and Hercules to find out how these different firms work. I encourage you to do the same, to send people in a lagging part of your organization to watch how somebody else is doing the same job they're doing but with a marked degree of excellence.

In our organization, we've found that if you're inside your own box and never have a chance to look at how somebody else is working, you may believe you're excellent. The irony is that you indeed may be excellent by your past standard. You may be improving. But in reality your work is not so impressive when compared to what is both theoretically and practically possible.

Alcoa also systematically analyzes the causes of accidents and then just as systematically takes away those causes. For instance, we spent four million dollars making engineering changes to our facilities to protect people against falls. We've also adopted a policy, on the environment that's a companion to our safety policy. it says, among other things, that we won't wait for the government to tell Lis what we should do. if we know that we're in danger of jeopardizing our people or our communities, we'll act without government regulation or intervention. Furthermore, as a matter of policy, we will correct problems that interfere with safety or the environment before we take care of any other capital spending.

Alcoa's safety program has been in place for three and a half years. Some people say a company can't accomplish anything significant in fewer than 10 or 15 years. I don't believe that's true, but the process certainly is neverending.

2. Management. At Alcoa, managing for quality means several things. First is reducing the variability in processes. To do this, we have to understand how much variability there is. We have to be able to measure it on a regular basis. We have to wage a concerted, organized attack against variability. And, as we achieve one level of perfection, we need to challenge ourselves to go on to the next level.

It's not something a company can accomplish with cheerleading. You must take a significant step up, at least to the level of providing education and training, another area of managing for quality.

One of the things we know about our population at Alcoa is that many people cannot read or compute. At some firms, this absence of skill is invisible, so you may think you don't have these kinds of problems. But, if you could measure the skill levels of all your employees, I believe you'd discover many people can't read even safety signs. Some people don't understand a sign on a wire fence that says: "Do not pass this area or you will be electrocuted." So, of course, you have people who look at a statistical process control chart and don't know which way is up.

To be a world-leading company in the next century, you can't leave these problems unattended. You have to get people in positions where they can make mental, not just physical, contributions to the work.

To compete now and in the future, you also need modern equipment and facilities, still another area of managing for quality. Yet I don't think modern equipment and facilities will be a dividing line for success in the century ahead, because equipment is going to be readily available to people of means everywhere in the world. Technology may give you an immediate advantage over your competition, but it won't last very long.

What, then, is the basis for competition in the next century? Quality and a motivated, energized, well-trained - constantly training - workforce. R&D is important, too, though, like technology, R&D won't provide lasting leads. The skill with which companies are organized and assets deployed will determine economic success.

Let me give you an example of how this philosophy works at Alcoa: two employees at our Rockdale, Texas, plant wanted to improve operations by decreasing the variation in the height of the anodes, or terminals that carry the current, during the smelting process. (In smelting pots where aluminum is made, carbon anodes are held partially submerged in molten cryolite, and their precise elevation is critical to efficiency.)

The industry-accepted variation in height is one inch. But, by developing new gauges and tracking the results with statistical process controls, the two employees were able to reduce Alcoa's variation by 80 percent and thereby achieve the best pot control levels in Rockdale's 35-year history. They also contributed $80,000 of economic value to the Rockdale plant with that one simple idea.

Now, $80,000 may not be a huge amount at the corporate level, but I'm convinced that it's that kind of initiative that will make the difference between companies that are successful in the next century and those that aren't. It won't be through the financial "engineering" that has dominated the last 20 or 30 years. It will be through setting objectives, like Alcoa's: our goal is to earn, on average, a 15-percent return on shareholder's equity. Before 1987, Alcoa's return on equity was averaging about 10 percent. We set as our goal to earn not less than 10 percent per year, even in a serious downturn, and to achieve growth by improving the way we use our assets.

3. Return to shareholders. For the concept of quality to work in a company, the concentration can't be on mechanical processes alone. That's why Alcoa established a new policy for our shareholders that shows them we believe we're all in this together.

Since we commit to earning at least a 10-percent return on equity, we should always be able to pay our shareholders a base dividend of 40 cents per share every quarter. When we do better than 10 percent, we should automatically pay 30 percent of everything we make over $6 per share, which is basically a 10-percent return to our shareholders.

We announced this policy a couple of years ago. When we declared the final dividend for 1990, which included an additional 18 cents based on our earnings in excess of $6, we discovered one wrinkle. In the fourth quarter, we took a $275 million write-off, which made it look as if we didn't really make $6.79 - we made a little more than $3. We proposed to our directors, and they agreed, that we should ignore the write-off in calculating dividends to demonstrate to our shareholders that Alcoa didn't take the write-off to avoid paying a dividend.

This kind of quality consideration solidifies a firm's relationship with its shareholders, making the point that we're going to do what we think is right and not play around with the numbers.

4. Compensation. Another example of how we close the circle with quality at Alcoa is the profit-sharing plan we instituted in cooperation with members of our three major unions - the steel workers, aluminum workers, and auto workers. We've been providing profit-sharing to our 23,000 employees in the U.S. aluminum-based operations since 1988. Our basic compensation rate for employees who are in the profit-sharing plan is 137 percent of the U.S. industrial average.

But profit-sharing is not a substitute for fair compensation to those already at the upper end of the compensation scale. We told our hourly and non-incentive compensation people that, when we do well, everyone should be able to tell the difference, not just the people in the proxy statement and the officers of the corporation. We insist of our unions that they agree to institute profit-sharing on top of regular compensation. We still need to tighten the linkage between the profit-sharing population and the incentive-compensation population and to figure out how to effectively apply the same policy in places like Brazil, Australia, Western Europe, and Africa, but we're working on the problems.

5. Repairs. When I first came to Alcoa, I pressed the financial staff to reduce the amount of time it takes to close our books - not because I need the data a few days sooner, but because I think it's such an obvious area in which to apply the ideas of quality.

The amount of time big corporations typically take to close their books is between six and 15 days, with most companies clustered in the eight-to-ten day range. If we use the quality principles to examine where the time goes, we find that almost all the time is spent on repair work, getting the numbers corrected. But, if we get it right the first time, in every activity throughout the organization, we free up highly skilled people to work instead on creating value.

Unfortunately, when you begin taking these issues apart, you have to attack your information system, examine your worldwide definitions of charts of accounts, implement a superior training program, and, of course, stop moving employees around every 18 months.

Every spoke in the wheel

How does Alcoa's management team communicate these quality concepts to its employees so they enlist themselves enthusiastically in a full-fledged quality effort?

We began the training process in Alcoa with the management committee and it hasn't stopped. Based on my experience with another employer, I'm convinced a company cannot possibly get this done if it skips any levels of management.

The reason is simple. if someone is sitting on top of a pyramid, and he or she either is a booking agent or just doesn't agree with what you're trying to do, the people under that person are not going to do what you want. When we began this effort at Alcoa, I said we've got to start with the management committee and everybody needs to understand not just the words but how to use the ideas behind statistical process control and benchmarking and organized problem-solving - all concepts we're teaching to every employee.

You have to be dedicated to the idea that people matter. They're not substitutes for mechanical instruments. You have to lead by example, and you have to get out of the office and out there on the floor, where you don't just observe but you ask questions. You have to show that you know what your employees mean when they tell you something they've discovered about their process. It's hard to do, especially with people who have been with a company 30 or 35 years and believe they know everything there is to know about the industry. Trying to change the culture of an old-line organization begins at the top and cascades down through every level of management.

Now, how does a firm calculate the corporate cost of this drive for quality? The idea is to calculate the distance between where you are and where you could be if you did everything right the first time. At Alcoa, I think our cost is 20 percent of the cost of goods sold, which means about a billion and a half dollars annually.

We haven't formally calculated the figure because there's a divided school of thought on the subject. Some say we have to provide a number to motivate the organization and help everyone understand how big the target is. Others say that people really view the figure as a threat by management, which some believe set unobtainable productivity goals. We've consciously elected not to make such a detailed calculation, although measuring how each of our processes compares to the world's benchmark is in effect a step toward figuring a cost of quality.

Alcoa also has chosen not to adopt the philosophies of any of the quality gurus. We've crafted our own set of visions, values, and milestones of achievement. We've created our own training material. In fact, I think the proprietary training material we devised on benchmarking, identifying a critical process, controlling a critical process, and continuously improving the process is better than anything available. These materials offer the link between the abstract ideas and how they're put into practice at the local level.

When we were creating our problem-solving process, I worried that it was too routine, that it would turn off people because it requires a very demanding, rational thought process. It insists you go through the steps of idea creation, realization, and solution in a systematic way. But employees seem to like it better than any other aspect of our quality program, because it creates for them a common language, which allows their conversations to transcend reporting relationships. It helps people talk to each other, and that's really powerful.