Byrne quickly realized that by applying lean techniques he could run the company at its current sales volume with half the people and half the floor space. Given the financial situation, he had to take immediate action. So, as his first step, he tackled the excess-people problem.
Dealing Up Front with Extra People and Anchor-Draggers
In November 1991, Art Byrne announced that the crew was too large to keep the ship afloat and offered a generous early retirement package to the aging workforce in the plants and to the office staff. Although he believed that only half the workforce was needed, he set the headcount reduction goal at 30 percent, knowing that as soon as he got the product development system working right, sales growth would absorb the remaining excess people.
Almost all of the eligible hourly workers took the retirement offer, but only a small fraction of the office staff accepted. Art and Judy Seyler, his vice president for human resources, therefore conducted a “de-layering.” They classified every job in management as either:
• value creating (defined as the ability of Wiremold to pass the costs of the job along to the customer),
• nonvalue creating (from the standpoint of the customer) but currently necessary to run the business (for example, the environmental expert helping the company meet government regulations, Type One muda ), or
• nonvalue creating and unnecessary (Type Two muda )
They then classified each manager as either:
• able to create value,
• able to create value with some development of skills, or
• unable to create value, even with development (usually due to unwillingness to change their attitudes about the organization of work)
After years of creating lean organizations, Art had concluded that about 10 percent of existing management will not embrace the new system. “Lean thinking is profoundly corrosive of hierarchy and some people just don’t seem to be able to make the adjustment. It’s essential that these anchor-draggers find some other place to work—after all, there’s still plenty of hierarchy out in the world—or the whole campaign will fail.”
The people in the first two categories were therefore matched up with jobs in the first two categories to create a new organization structure (com-pare Figure 7.1 with 7.2) with a new roster of players. Employees not finding useful jobs were given a generous severance and within thirty days of Art’s arrival the new structure and player roster was in place. Only one outsider was recruited, Frank Giannattasio, the new vice president for operations.
F IGURE 7.1: O LD W IREMOLD O RGANIZATION
As Judy Seyler looks back on this event, she notes that it was terribly traumatic in a hierarchical, paternalistic organization in which no one had ever been asked to leave. “Even though the financial cost was very large, particularly given our lack of profits, Art was determined to be generous with people while making it clear that in the future everyone must create value by working together in a different way.”
F IGURE 7.2: N EW W IREMOLD O RGANIZATION
When the headcount reductions were completed, Art Byrne called a meeting of the entire workforce of the parent company and announced that no one would ever lose their job as a result of the improvement activities that would start immediately. “The bad part is over; now we will all learn how to continuously create more value so that we never have bad days again.”
Byrne was in effect giving job guarantees to his union workforce without asking anything in return except that they be open-minded to change. “I’m certain that 99 percent of American companies wouldn’t do this, but taking away the fear of job loss is at the very core of a lean conversion. Think of it logically from a human perspective rather than as some corporate bureaucrat. If I asked you to help me reduce the number of people needed to make a particular product from five down to two, and after you did, I followed up by laying off three people, one of whom was your cousin and another a good friend, what would you say to me when I asked you to help me do the same thing a month later for another product?”
Teaching People How to See
Based on his experience in “leaning” eight separate businesses in the Danaher group, Byrne had concluded that the single most effective action in converting an organization to lean practices is for the CEO to lead the initial improvement activities himself. “This is where most American companies fail right at the outset. CEOs want to delegate improvement activities, partly because they are timid about going out on the shop floor or to the engineering area or to the order-taking and scheduling departments to work hands-on making improvements. As a result, they never really learn anything about change at the level where value is really created. They continue to manage in their old by-the-numbers manner, which kills the improvement activities they thought they started. The fact is that big changes require leaps of faith in which the CEO must say ‘Just do it,’ even when ‘it’ seems contrary to common sense. If the CEO spends time in real operations learning just how bad things really are and begins to see the vast potential for improvements, he or she will make the right decision more often.”
Because no one else in the company understood lean principles, Art Byrne led the initial training sessions himself. Using a manual he had written, he conducted two-day sessions on lean principles for 150 employees followed immediately by three-day kaizen exercises so employees could use the skills they had just learned. (This was very different from Wiremold’s previous improvement activities, conducted as part of TQM, where improvement teams met weekly for an hour or two, mostly to plan improvements to be implemented weeks or months later.)
Byrne then gathered his managers and union head together and took them on “the walk of shame” through every part of the plant and through the engineering and sales departments. “There was muda everywhere and my managers were now able to see it. I told them that we were going to convert every process, including product development and order-taking, into continuous flow and that we were going to learn how to pull. I also told them that I was going to get them the best help in the world, from Iwata and Nakao, who were at the end of their exclusive agreement with Danaher and ready to work for Wiremold.”
Attacking Every Value Stream Repeatedly
Soon hundreds of weeklong kaizen activities were under way (and continue to this day), involving practically every employee, as every value stream in Wiremold was repeatedly evaluated for ways to make it flow better and pull more smoothly. Wiremold’s assumption is that every stream can always be improved in pursuit of perfection and that every stream must be improved in pursuit of perfection. Equally important, it is presumed that results can be achieved very quickly, the common expression being that “if you can’t get a major improvement in three days you are doing something wrong.” Once this mentality is reinforced by results—and employees begin to believe management’s guarantee that no job will ever be lost due to improvement activities—improvement can become self-sustaining .
Re-creating the Production Organization to Channel the Value Stream
When Art Byrne de-layered the Wiremold organization (again, see Figure 7.2 ) he did far more than remove tangential jobs and frills that could no longer be afforded. He smashed the departmental barriers to focus every-one’s efforts on the value stream by creating dedicated production teams for each of Wiremold’s six product families. The purchasing, manufacturing, and scheduling (MRP) groups within the Operations Department, the Engineering Department, and the “process villages” (stamping, rolling, molding, painting, assembly, etcetera) in the plants were eliminated, with their personnel reassigned to product teams provided with all of the resources needed to produce a specific product family.
Let’s take Tele-Power[H23008] Poles as an example. (These are the steel or aluminum columns extending from floor to ceiling in open office settings, with power and communications outlets on every side of the column to permit a host of adjacent workstations to plug in. They are offered in an enormous variety of shapes, lengths, plug configurations, and
colors.) Team leader Joe Condeco was given complete responsibility, and profit-and-loss accountability, for Wiremold’s “pole” products from initial launch through their production life. More radically, the team leader, the product planners, the buyers, the factory engineers, the production supervisors, and the production associates were all co-located on the factory floor immediately adjacent to the realigned machinery producing the poles in single-piece-flow cells.
The team was given its own punch presses and rolling mills, as well as assembly equipment, so it could be self-sufficient. Before, the assembly activity was dependent on the Rolling Mill Department for their bases and covers. Despite large stocks on hand, they would often lack the right base or enough covers. When they asked the Rolling Mill Department for more of a missing item, the response would often be, “Sorry, but the master schedule generated by the MRP system calls for us to make other items now. You’ll have to wait until next week or take your problem to a higher level.” Now the Tele-Power[H23008] team has all of the equipment it needs. There can be no excuses.
The new setting was initially a great shock to the “white collars” who had always worked in a remote office and seen themselves in a very different light from “factory workers.” (Wiremold soon implemented a casual dress code based on Art Byrne’s belief that “neckties cut off circulation to the brain and inhibit teamwork.” This was another problem for office workers who somehow felt that their appearance rather than their skills and their contributions made them special.) Reassignment to product teams was also a shock to the process specialists working in the process villages like the Rolling Mill Department who had traditionally hoarded their tricks of the trade. However, everyone soon came to like it. For the first time, they could actually see value flowing!
Introducing a Lean Financial System and “Scoreboard”
To get the production teams to work in accordance with lean principles, Wiremold needed to junk its traditional system of standard cost “absorption” accounting, which allocated costs by labor and machine hours in accordance with mass-production thinking. Production managers knew from experience that they had to “absorb” allocated overhead by spreading it over as many machine and labor hours as possible. This system gave an overwhelming incentive to keep every worker and every machine busy—to “make the numbers”—by producing inventory, even if the inventory consisted of items no one would ever want.
As Orrie Fiume remembers, “Standard cost and variance analysis were declared dead as concepts immediately after Art’s arrival. We looked at Activity Based Costing but knew it wasn’t the answer. Its advocates will tell you it is based on cost drivers, but in reality it’s just a different method of allocating overhead. There is still too much allocation of aggregated costs downward. We were determined to work from the bottom up.”
The key to the new way of thinking was to organize production by product families, then let each product team do its own purchasing and buy all its own tools. A simple system could then be devised to assign real costs to each product line. Today, more than 90 percent of the costs involved in making a Tele-Power[H23008] pole, for example, come from product-specific cost analysis. Only a small fraction of cost is an allocation outside the control of the team, specifically occupancy costs for whatever space the team is using in a plant. And even in this case, the team is charged only for the space it actually uses, so costs can be reduced by using less.
Some elements of the old standard cost accounting system are retained in the computer because the financial statement needs them—for example, the value of in-process inventories. However, these are deemphasized in evaluating the performance of the product teams, which are told to concentrate on the cost of manufacturing instead. Similarly, the financial implications of running down the inventories during the transition period were not shown to the product team leaders for fear they would do the wrong thing. 3
In addition to a simple profit-and-loss calculation, Wiremold’s production teams were given a new “scoreboard” consisting of some simple, quantitative performance indicators:
• productivity of the product team (expressed as sales per employee),
• customer service (expressed as percent of products delivered on time),
• inventory turns, and
• quality (expressed as the number of mistakes made by the team)
The team leaders and their teams can see these indicators at all times because they are prominently posted. In addition, the two primary ways to improve are obvious. First, smooth the flow of products through the system, with no backflows for reworking quality problems, no scrap, and no in-process inventories. Then, make only those products customers actually want, because productivity is measured as end-market sales (not additions to in-process inventories) per employee.
To keep everyone marching at the same pace, Wiremold equips the scoreboard with a set of expectations as well. Specifically, team leaders and their teams are expected to:
• reduce defects, as shown in the quality indicator, by 50 percent every year;
• improve productivity, expressed as sales per employee in constant dollars, by 20 percent every year;
• deliver 100 percent of products exactly on time;
• increase inventory turns to a minimum of twenty per year; and
• increase profit sharing to 20 percent of straight wages (as explained in a moment)
“Variance analysis” is still performed but not based on variances from standard costs. Instead, when the trend line starts to diverge from performance targets, the team collectively searches for the root cause of the variance rather than maneuvering to “make the numbers,” as in the old days.
Running Down Inventories
Because Wiremold is privately held and the board of directors understood what was happening, the special financial problem of inventory reduction in a lean transition was not a major concern. For a publicly traded company, however, rapidly running down inventories can be a real problem, one worth a brief digression. As firms move from batch-and-queue to flow systems, enormous amounts of cash are suddenly made available from freed-up inventories. (This offers the firm a special strategic opportunity, as we will see in a moment.) The problem is that the removal of these inventories increases production costs, as shown on the financial statement, and can easily wipe out profits.
Let’s take a simple example. Firms typically calculate their costs of production and profits in the following manner, as shown in the left-hand column of Table 7.1 .
T ABLE 7.1: C ONSEQUENCES OF I NVENTORY R EDUCTIONS FOR P ROFITABILITY
Now, suppose that the new “lean” management takes in-process inventory down dramatically from $576,000 to $100,000 while holding everything else constant (except, of course, material purchasing, because products are being made largely from the inventories already on hand). Running the numbers again, as shown in the right-hand column, it’s apparent that the new management, while trying to “do the right thing,” has moved the company from a $153,000 profit to a $36,000 loss (even as cash flow has zoomed).
This phenomenon can be very bad for publicly traded companies unless the management actively explains the situation to stockholders in advance. The only alternative to education is a slash-and-burn campaign of head-count and cost reduction (in the direct labor and indirect manufacturing cost accounts) to restore short-term profits. This can, however, easily set back the introduction of lean thinking or even make it impossible if the traumatized workforce refuses to cooperate with lean initiatives .
Creating a “Lean” Function
To help the product teams continuously improve, Art Byrne created a new function, the JIT Promotion Office (JPO). The old Quality Department, some of the training activities formerly conducted by the Personnel Department, and several high-potential associates from different areas of the firm were grouped under the JPO. With the JPO, the task of working through the entirety of Wiremold, value stream by value stream, could be speeded up.
The
product team leader and the JIT Promotion Office jointly evaluate the value stream for the product to determine what types of kaikaku and kaizen activities should be performed and when. A team leader from the product team and a facilitator from the JPO are then assigned to each improvement team (which might be a subset of the product team, the whole team, or some portion of the team plus outside experts with needed skills). Because the team leader will go back to her or his job on the product team once the kaizen is finished, the facilitator from the JPO shoulders the critical responsibility for seeing to the completion of the follow-up work invariably resulting from a weeklong improvement effort.
In addition to planning and facilitating improvement exercises, the JPO teaches every employee the principles of lean thinking (identifying the value stream, flow, pull, and the endless pursuit of perfection) plus lean techniques (standard work, takt time, visual control, pull scheduling, and single-piece flow in particular) and periodically reteaches them. As Frank Giannattasio notes, “This is an enormous but critical challenge. Your middle management, in particular, feels threatened by the lean transition and the removal of all those safety nets. When in doubt, they will take you right back to making batches and building inventories unless you reinforce the message through continued teaching, coupled with continuous hands-on improvement exercises.”
Offering Ironclad Job Guarantees in Return for Flexibility
As we noted earlier, Art Byrne knew that if the value stream for every product was going to be unkinked continuously, people would continually be left by the side of the stream. Resistance to continuous improvement would be chronic unless he guaranteed that workers would not be out on the street, even if their specific job was eliminated. He also knew that the existing work rules in Wiremold’s union contract—restricting stampers to stamping, painters to painting, molders to molding, and so on—would make it impossible to introduce flow and to continuously improve every activity. Finally, he knew that his workforce would have a very hard time differentiating layoffs due to weak demand from layoffs due to kaizen. Therefore, once the initial retirement offer was accepted, Art went immediately to his union and offered job guarantees for the remaining workers in return for their cooperation in working in a new way.
Lean Thinking Page 16