A bible for benchmarking, by Xerox.

To benchmark effectively, a company needs solid support from the top, but the concept also must be an integral part of the organization, cascading down to every employee.

For Xerox, benchmarking sprang from a competitive crisis. One Sunday morning in the late 1970s, our vice president of manufacturing and engineering read in the paper that a U.S. dealer, supported by its overseas manufacturer, was selling a xerographic device equivalent to a Xerox device at a price equal to our manufacturing cost. To analyze the huge gap between the competitor's market price and our manufacturing cost, we first went the traditional route with a product tear-down/reverse engineering and a visit to our foreign affiliate, Fuji Xerox, which gave us a window into the competition. But the visit merely confirmed that we were 50 percent off the mark. So we decided to try benchmarking, and we haven't looked back. In fact, after our first big successes, senior management required all organizations within Xerox to pursue benchmarking. After 12 years, we are, if anything, intensifying our benchmarking activities.

A Chinese proverb says, "If we don't change our direction, we might end up where we're headed." Benchmarking is a direction-setting exercise, and it is nothing more than a quality tool, just one of many ways to improve and become more productive. However, it has been extremely important for us, so you may want to know why and how it fits in with an overall quality agenda.

Our quality program is called "leadership through quality," and it encompasses three processes: quality, problem-solving and benchmarking. We have trained more than 100,000 employees in the quality process, which shows them our output, as well as who the customers are and what they want. The problem-solving process is a simplified version of the quality process for employees on the line and in the field. It enables them to quickly and simply analyze their own jobs, leverage the personal talents best-suited to their work and improve their performances.

After instituting the first two processes, we realized that they didn't help employees understand where we wanted them to concentrate their improvement activities. That's where benchmarking comes in. It's the process of finding better practices, bringing them back and handing them off, first to quality teams and then to employee involvement teams to see how they can use the first two facets of the quality program to implement those practices.

A TOUGH YARDSTICK

The formal definition of benchmarking is the continuous process of measuring our products, services and practices against those of our toughest competitors or companies renowned as leaders. We have never changed this definition, because it embodies some very tough lessons we learned while introducing benchmarking.

The first lesson concerns the competition. Although you must focus strongly on the competition, if that's the sole objective, playing catch-up is the best you can do. Watching the competition doesn't tell you how to outdistance them. The mix of our benchmarking activities has changed 180 degrees. In the early days, we spent 80 percent of our benchmarking time looking at the competition. Today, we spend 80 percent of that time outside our industry, because we have found innovative ideas from business in other industries.

Other companies have done the same. For example, Remington Arms, a division of the Du Pont Co., makes shotguns, rifles and ammunition. Its customers asked for smoother, shinier ammunition shells. The Du Pont team members wanted to achieve that goal with competitive benchmarking in their own industry, but Du Pont engineers targeted instead the cosmetics industry, which makes smooth and shiny lipstick containers. So Du Pont did some benchmarking with Maybelline and incorporated some practices into its manufacturing process to accommodate customers' request. Anecdotes like that demonstrate why we look beyond our industry for ideas.

In the beginning, we associated benchmarking with numerical measurements. We still do, because we must be concerned about financial ratios like selling, general and administrative expenses as a percent of revenue. But that information doesn't say anything about the changes necessary to improve performance. Most businesses have learned to look for indicators of how competitors are doing things differently. But you can't stop there; you must understand what the practices are to incorporate and measure their effects.

The formal definition is pretty cumbersome, but the operational definition of benchmarking is "finding and implementing the best business practices." It doesn't have to be any more complicated than that.

Taken one step further, the operational definition includes customer satisfaction. This is Xerox's priority, and one way to please customers is showing them how easy it is to do business with us. The primary conduit for that objective is the company-client contact point, which includes taking and filling orders, repairing devices, creating invoices and collecting payments. Those processes must incorporate the best practices to ensure customer satisfaction. With that objective, employees can pursue benchmarking in a straightforward fashion, because they know what they are being asked to do.

The four major types of benchmarking are internal, functional, generic and competitive. Internal benchmarking holds that large organizations have multiples of the same unit set up to do the same thing, such as similar marketing offices, districts, multiple distribution centers and order-taking points. The company can compare practices among internal areas and determine which one is the benchmark, bringing the others up to the same performance level.

Xerox was not always good at that. We associated benchmarking with trips to other companies, something known in benchmarking circles as industrial tourism or "feel good" trips. We did not conduct these trips with the discipline necessary to uncover and thoroughly understand the best practices, but more important, we lacked the necessary perspective on our internal operations to make the trips valuable as external comparisons.

Functional benchmarking is the story of L.L. Bean. In the early 1980s, when most managers were transfixed with competitive benchmarking, the Xerox review team members asked, "Who's the benchmark?" They expected to hear names like Kodak, IBM, Cannon or Minolta. We said no: The benchmark is L.L. Bean. Now, imagine the surprised faces in that audience, because this was the first time we advocated looking outside the industry. The audience said, "Wait a minute -- we're a copying company. How could we possibly compare ourselves to this little outdoor specialty company up in Freeport, Maine?"

SURPRISING SIMILARITIES

But all companies take their customers' orders and, provided we maintain some common characteristics, we can compare ourselves to companies in other industries. L.L. Bean's products, like ours, don't come in nice little standard packages, as you might guess about a company that sells ax handles, red flannel shirts, boots and canoes. And the company picked its orders manually, as we did, but the similarities stopped there: L.L. Bean picked its orders three times faster.

How did we know L.L. Bean was three times faster? We researched several companies as potential benchmarking candidates, scanning topical publications on material handling and engineering and attending a conference at which some Bean executives spoke. A magazine article gave us enough information to estimate the company's productivity rate. Eventually, we called them and simply asked, and after confirming their rate, we decided to compare processes.

Today, generic benchmarking is one of our most important focal points. We have been trying to determine if we can dedicate precious resources to continuous improvement and in which areas. After benchmarking with several other prominent firms and conducting some studies, we recognized that we would get the greatest return by improving basic business processes. We identified 67 processes for one unit and assembled a plan to revamp all of those steps.

Let's take an example of each of the four types, beginning with internal benchmarking. Quickly answering customer inquiries and complaints makes customers happy, so we set up a customer problem resolution process as the benchmark, derived from our sister affiliate, Xerox Canada. We did not have to leave the Xerox family to find benchmarks in this case.

Competitive benchmarking for us meant streamlining without compromising service. Before benchmarking, Xerox had four layers -- four places where we stored and handled material, and we were 30 percent off the mark. Our competitor had two layers, at manufacturing and the service person. The cost and asset levels affecting the profits and losses or the balance sheet were almost directly proportionate to the number of layers. Since then, we have changed our structure to look more like our competitor's, without any loss of service quality.

L.L. Bean is a good example of functional benchmarking. The first rule of running an efficient warehouse is ensuring that the fast-moving merchandise is closest to the main aisle, a principle called velocity sequencing. The employee can push a cart with customer order boxes and pick the necessary items. In fact, L.L. Bean uses the number of feet that a picker travels daily as a productivity indicator.

But L.L. Bean did something else. The company recognized that orders came in randomly and were unlikely to coincide with the locations of the items, so it changed the order fulfillment system. Within a set time period, employees sorted the orders and put like items together, enabling the picker to make one trip and pick all the red flannel shirts necessary for that batch of orders. Xerox did not have a comparable system, which accounted for the productivity differences. For functional benchmarking, the practices are the most important consideration.

Generic process benchmarking for us is the scheme under which we roll up the 67 processes. In developing this plan, we found that we needed some accountability. One individual is responsible for each major area, so process owners are designated for 10 areas and 67 processes. These individuals document their processes and ensure that they are benchmarked.

Process owners have another responsibility, because many processes are cross-functional. For instance, an order-taking process creates an invoice, so decisions must be made about resource levels. The designated individuals are responsible for making trade-offs in the company's best interest. Our scheme gives us accountability and management.

For Xerox, benchmarking is a 10-step process, as shown in the table below, but the first three steps are the most important: identifying what to benchmark, who to benchmark and where to get the data and information. In the early days, we did not mandate which companies to benchmark; we simply told our people to do it. The result was problem-based benchmarking, which we still use. However, this is not necessarily the most goal-oriented approach, so we refocused our methodology on process-based benchmarking, which will be the foundation for most of our continuous improvement efforts during the next five to eight years.

A PLAN OF ATTACK

With 67 business processes, we needed a plan of attack. Since customer satisfaction is number one, we began by targeting the processes that affected customers most. That brought us to customer engagement, the second category. The other nine areas did not want to spend years waiting their turn, so we began a demonstration project for them.

Whom to benchmark is the second step. To develop a candidate list, we might begin by looking at a list of companies cited for excellence in their business practices -- the type of ranking that various trade and business journals might compile. We would then conduct secondary research to pinpoint three to six companies for a specific benchmarking activity, which allows us to be certain that we have found the benchmark.

To obtain benchmarking information, you can tap more than 5,000 electronic data bases. Pay attention to conferences, professional associations and anything published on the subject, including software. Ask your software provider which customers use its software packages well, because these companies may be the ones to benchmark. Other options include original research, focus groups, commissioned studies and site visits to companies.

Step 10, recalibrating, is also important. Typically, we recalibrate once annually on the manufacturing side. On the business or marketing side, recalibrating every three years is adequate. In both cases, recalibrating is necessary because the benchmark, like an Olympic record, will change.

The best way to accomplish positive change is to believe in a need for improvement, determine what to improve and give people a vision of the goal. Benchmarking achieves all those objectives by bridging the gap between internal and external practices. However, it is not an isolated quick fix, but a continuous practice that must harmonize with other company initiatives.

To make benchmarking an integral part of your organization, emphasize it in your business and operating plans and in operations reviews. If you don't ask for it, people won't do it. Employees must be aware of the concept and their roles and responsibilities in conducting benchmarking. On-demand training for teams is also important.

A FEW ENDNOTES

A competency center, of which I'm the head, is crucial. I do not benchmark, but I help the teams and sustain the benchmarking concept. If the organization must make an important decision, I remind the appropriate individuals to make certain we have benchmarked the activity or process.

When we are planning to benchmark, we commission a team. Also, we follow a process, and so should you, whether it's our 10-step process or another company's model. Documenting the process is another must.

With the help of benchmarking, we have cut unit manufacturing costs in half since the early 1980s, and parts acceptance is nearly 100 percent. In-process inventory has been reduced by two-thirds. Service labor costs have dropped, and in the distribution organization; where I spent 15 years of my career, we upped the productivity rate substantially.

The most recent development in benchmarking circles is the International Benchmarking Clearinghouse. As an adjunct of the American Productivity and Quality Center, Xerox has helped set up a single organization in the United States as a central repository for benchmarking products and services. We use it to help qualify and teach people before they benchmark with us.

We want effective benchmarking to become prevalent in the organization and for employees to recognize its importance. Our current chairman, Paul Allaire, says the primary objective of benchmarking is understanding exemplary business practices. Four things count in benchmarking: the process on which you focus, the organizations you visit, the best practices you find and the changes you institute. Target-setting is secondary.

FINANCE GETS ON THE BENCHMARKING EXPRESS

Benchmarking is a two-way street to John Newcomer, manager of finance business planning at Xerox's U.S. customer operations division. "There's always something to learn, even if you think you're at the top of the class," he says. The finance department evidently agrees with him, because it's benchmarking at a steady pace. In 1992, it conducted 22 studies at 24 companies, not counting informal benchmarking activities, Newcomer reports.

Although the finance group benchmarks with a variety of companies, including Kodak, AT&T, IBM, Abbott Laboratories, and Hewlett Packard, it favors those of roughly similar size and complexity. The company also participates in competitive studies, in which a company surveys other companies on a particular process or activity, analyzes the data it receives and shares the results with the respondents, he says.

Benchmarking efforts have focused on taxes, financing businesses, financial accounting and, more recently, financial systems. These are all process-oriented areas to which benchmarking is easier to apply than in other areas of finance, such as traditional financial planning and investment areas, Newcomer points out.

Internal benchmarking is a key information conduit. Because of the overall benchmarking drive, "There is an ongoing transference of ideas between divisions in Xerox," although the divisions do not formally disseminate information, he notes. In keeping with that tradition, the customer operations division recently benchmarked Xerox of Canada to investigate the mechanics behind its quick monthly close. The answer lay partly in the Canadian affiliate's early cutoff date for journal and data entries. To explore other methods of reducing cycle time for the monthly close, finance also will benchmark three external companies, according to Newcomer.

Benchmarking with Digital Equipment Corp. helped the finance department address a different question: How much of its financial resources were necessary to drive the business? By gauging its practices against Digital's, finance could more clearly see opportunities to reallocate resources. Also, the company was interested in Digital's financial architecture, especially the financial systems developed during the last 10 years. Through its discussions with Digital, Xerox learned to recognize possible pitfalls and the need for "realistic time frame and resource commitment expectations" in developing a superior financial management system, Newcomer says.

In general, finance benchmarking "is a process that goes up and down the whole organization," he explains. The CFO of the U.S. customer operations division is very involved in the process, he reports, and a typical benchmarking team also includes senior analysts, first-line and sometimes second-line managers. The usual number of people on a team ranges from three to eight people.

What's next on the finance agenda? Newcomer says his department "will continue looking hard at cycle times for our various processes and the associated concept of more productivity." If finance can reduce its cycle times, the department eventually will be able to spend less time on creating information and more on analyzing it. And, since benchmarking is a nonstop activity at Xerox, "we will continue to examine ourselves and to emphasize the overall quality initiative that began 10 years ago."

NAVIGATING THE PROCESS

Ask five companies and you'll probably get five different benchmarking methods, but just for starters, here's the process Xerox uses:

PLANNING

1. Identify what is to be benchmarked.

2. Identify comparative companies.

3. Determine data collection method and collect data.

ANALYSIS

4. Determine current performance levels.

5. Project future performance levels.

6. Communicate benchmark findings and gain acceptance.

INTEGRATION

7. Establish functional goals.

8. Develop action plans.

ACTION

9. Implement specific actions and monitor progress.

10. Recalibrate benchmarks.

Mr. Camp is manager of benchmarking competency quality and customer satisfaction at Xerox Corp. in Rochester, N.Y.