Increasingly, global competition requires that manufacturing companies improve their production efficiencies or face extinction. And indeed, by adopting Toyota's Lean System, companies have achieved tremendous gains in productivity and profits by removing the waste, or muda, from their production processes. Many firms have realized the same improvements by applying lean principles to their critical business processes.

However, creating a lean business process is only half the battle against muda. The knowledge workers involved in these processes must also develop lean work habits. A firm that only creates a lean process without creating lean work habits is like a sprinter with a track spike on one foot and an army boot on the other -- and that's a sure way to lose the race to satisfy the customer.

Our work with a major manufacturing firm illustrates the difference between a lean process and lean work habits. The labs were immaculate -- a model of 5S implementation -- but the lab manager's office was a mess: inventory was disorganized and critical information was difficult to retrieve.

Our company has taught lean work habits to a variety of manufacturing firms, including Toyota, over the past several years. These firms have realized increased employee efficiency, improved worker response time and increased customer satisfaction. Most significantly, by reducing the muda endemic in most knowledge workers' behavior, employees have gained nearly 10 hours per week for work that creates customer value.

Lean work habits are critical for knowledge workers, because the multiple value streams flowing through them create a constant tension. Without lean habits to guide their work, the critical flow of information in the value stream clogs up. Think of the information bottlenecks in the form of backlogs on their desks, or the hundreds of unanswered email in their inboxes. Think of the enforced waiting throughout a department when decision-makers read but don't act upon a request. Think unnecessary motion of managers searching for documents amidst the masses of paper piled on their desks. Each of these wastes undermines the gains made by any improvements in the design of a company's business processes.

And these wastes can be catastrophic: in 1999, NASA's Mars Climate Orbiter burned up in orbit due to a miscommunication regarding English and metric units. A task force found that a simple, unanswered email about the correct measurement units led to disaster. The total loss to NASA: $327 million.

In response to these problems, we have developed the following principles that create lean knowledge workers and improve the flow of the value stream:

Principle #1: Screening Muda

Knowledge workers are inundated with information each day. Yet most studies show that 50% of this information has little or no value to the worker.

By contrast, production line employees don't have to deal with this problem. They have limited and clearly defined materials that come down the assembly line to their workstation. They know what to process and how to do it. Imagine the chaos if new raw materials or irrelevant parts suddenly came down the line: they wouldn't be able to do their jobs efficiently.

Therefore, we've taught knowledge workers at ImagePoint, Ernst & Young, Pfizer and other companies how to screen the information that enters their system in order to identify and reduce the muda. They have removed themselves from unnecessary mailing lists, and proactively told co-workers what types of information they need to see and what types are irrelevant to them. They have learned to discard low- or no-value information immediately, rather than let it clutter their inboxes. They have created rules in their email software to direct mail to appropriate folders so that it doesn't sit in their inboxes. Employing this lean work habit, they have reduced the amount of time spent handling email by 22%.

Although it's impossible for them to eradicate the scourge of unnecessary email, memos and magazines, learning to screen the information helps ease the burden and saves time. Even more importantly, reducing the amount of low-value information entering the system improves the "signal to noise ratio", which enables workers to identify the high-value information more quickly and easily.