Manage Knowledge to Improve

Chapter three

Manage Knowledge to Improve

Strategic Thought

Bad news must travel fast

I have a natural ability to find bad news. If it’s out there, I want to know about it. The people who work for me have realized this and keep me informed.

A lot goes wrong in any organization, even a good one. A product fails. You’re surprised by a customer who suddenly switches to another company. A competitor brings out a product that appeals to a broad new market. Losing market share is the kind of bad news that every organization can understand. Other bad news may be about what’s going on inside the company. Maybe a product is going to be late, or it’s not going to do what you expected it to do, or you haven’t been able to hire enough of the right kinds of people to carry out your plans. An essential quality of a good manager is the desire to seek bad news rather than deny it. An effective manager wants to hear about what’s going wrong before he or she hears about what’s going right. You can’t react appropriately to disappointing news if it doesn’t reach you soon enough.

You concentrate on bad news in order to get started on the solution quickly. As soon as you’re aware of a problem, everybody in your organization must go into action. You can judge a company by how quickly it organizes all of its brain-power to deal with a serious problem. An important measure of a company’s digital nervous system is how quickly people in the company find out about bad news and respond to it. Digital technology speeds business response time in any emergency.

In the old days, every company’s response to bad news was slow. Business leaders often learned about problems only after they became serious, since the only quick way to pass information was to interrupt them with a phone call. Before handling a problem, people had to find information in paper files or go down the hall to find somebody who knew something about the situation. When people had some information, they spoke on the phone or sent papers to one another. Every step in the process took a lot of time. Today the dawn of the Information Age means we can send information fast, but even now most companies don’t gather the key information about customer issues in one place. By contrast, a well-designed digital nervous system operates as an early-warning system.

The Internet was not always Microsoft’s top priority. Its arrival changed our business and became the biggest surprise we’ve ever had to respond to. In fact, in 1995 various experts said that the Internet would put Microsoft out of business. This was bad news on a huge scale. We used our digital nervous system to respond to that crisis.

On August 24, 1995, we introduced Windows 95, the most ambitious software product ever to reach the general customer. This was big news, and hundreds of stories appeared in the months before the release. Windows magazine said, “This year- for better or worse - Microsoft wins the war.” A Time magazine story said that Microsoft was “the center of the computer universe.” The introduction of Windows 95 itself was featured on major TV news broadcasts. Within a couple of months, though, the news stories were all taking the opposite view. The Internet had burst into the public’s consciousness, and people believed that Microsoft hadn’t been invited to the new party. Now stories in the press said we didn’t understand the Internet. Smaller, quicker competitors would put Microsoft out of business.

On December 7, 1995, we held our first Internet Day, where for the first time we publicly displayed the range of technologies we were developing to build Internet support into our products. Within a year of this we had “Internet-enabled” our major products and delivered a number of new ones that concentrated on the Internet. Now we lead in several major Internet areas. No one company will dominate the Internet, but Microsoft has come back to play an important role.

How, our customers and the press often ask me, did we turn the ship around so fast? First, we were never so unaware of the Internet as we might have seemed to outside observers. It wasn’t as if somebody said “Internet” and we didn’t know how to spell it. We had several Internet technologies on our list of things to do. But let’s also be clear. In 1993 we were not concentrating on the Internet. It was a fifth or sixth priority. At this time, we didn’t have an overall Internet plan for the company. We didn’t see that the Internet, a network for scientists and engineers, would grow into the worldwide commercial network it is today. The Internet, had such a limited capacity to carry digital information then that we saw it as a minor stop along the way. The Internet’s sudden growth in popularity changed all the rules. People knew the Internet had faults, but it made vast amounts of information available and enabled easy communication. In 1993 alone Internet use doubled to more than twenty-five million people.

The driving force behind Microsoft’s response to the Internet didn’t come from me or other senior executives. It came from a small number of enthusiastic employees who saw what was happening. Through our electronic systems they were able to get everybody to join them. Their story is an example of our practice, from day one, that smart people anywhere in the company should have the power to change things. It’s an obvious way for Information Age companies to work: all knowledge workers should be part of creating the plan. But we couldn’t make it work without the technology we use. For years everybody at Microsoft has had a PC with e-mail. It’s a famous part of our company culture, and it’s shaped the way we think and act.

We reached most of the important decisions during this crisis in face-to-face talk. But our decisions were all informed by exchanges taking place over e-mail. Working together electronically cannot replace face-to-face meetings. It’s a way to make sure that more work gets done before meetings so that the meetings in person will be more useful. Meeting time is so valuable that you want to be sure you’re dealing with facts and good ideas based on solid analysis. You want to be sure that meetings produce plans, that you don’t just sit around talking philosophy. E-mail helps turn managers from middlemen into “doers.” It encourages people to speak up. It encourages managers to listen. That’s why, when customers ask what’s the first thing they can do to get more value out of their information systems, I always answer, “e-mail.”

How fast a company can respond in an emergency is a good measure of how well it responds to events. People in the organization will feel threatened by bad news, but that’s OK if they feel it as a group. As a leader, I created a sense of crisis about the Internet in 1994 and 1995. Not to leave people unhappy or unable to work, but to excite them to action. The leader needs to create an environment in which people can analyze the situation and develop a good response.

I like good news as much as the next person, but it also makes me suspicious. I wonder what bad news I’m not hearing. When somebody sends me an e-mail about an account we’ve won, I always think, “There are a lot of accounts nobody has sent mail about. Does that mean we’ve lost all of those?”

This reaction may seem unnecessary, but I’ve found there’s a psychological need in people to send good news when bad news is coming. It’s as if they want to lessen the shock. A good e-mail system makes sure that bad news can travel fast, but your people have to be willing to send you the news. You must always be open to bad news, and then you have to act on it. Sometimes I think my most important job as a boss is to listen for bad news. If you don’t act on it, your people will eventually stop bringing bad news to your attention. And that’s the beginning of the end.

In three years every product my company makes will be out of date. The only question is whether we will replace them or someone else will. In the next ten years, if Microsoft remains a leader, we’ll have to cope with at least three major crises. That’s why we’ve always got to do better. I insist that we keep up with events, as well as pursue longer-term projects, and that we use “bad news” to drive us to put new features into our products. One day, somebody will catch us asleep. One day, a new firm will put Microsoft out of business. I just hope it’s fifty years from now, not two or five.

Change bad news to good.

If you treat unpleasant news not as a negative, but as evidence of a need for change, you aren’t defeated by it. You’re learning from it. Learning from failures and constantly improving products is a key to success in all companies. Listening to customers is a big part of that effort. You have to study what customers say about their problems with your products and be constantly aware of what they want.

But it is surprisingly hard to get bad news from customers passed all the way to the product design groups. Most companies don’t have an efficient chain of people and paper between customers and the people who can make major improvements. When the customer data finally reaches the product design group, it often isn’t easy for the team to give it the right priority. All of the delays mean that improvements don’t happen as fast as they should.

I recommend the following approach to bringing customer complaints and wish lists into the process of product and service development:

1. Concentrate on your most unhappy customers.

2. Use technology to gather rich information on their unhappy experiences with your product and to find out what they want you to put into the product.

3. Use technology to pass the news to the right people in a hurry.

If you do these three things, you’ll turn those bad news experiences into an exciting process of improving your product or service. Unhappy customers are always a concern. They’re also your greatest opportunity. Listening and learning rather than being defensive can make customer complaints your best source of significant quality improvements. Adopting the right technology will give you the power to capture and change complaints into better products and services fast.

For example, any employee of Promus Hotels, based in Memphis, Tennessee, can give customers their money back if they are not satisfied. A no-questions-asked money-back guarantee from a service company like Promus means something. The company has a good reason to fix the problem right away. And because every employee can act on the guarantee, employees watch for quality. So how does Promus stop customers using the guarantee for a free stay and complaining when nothing was really wrong? Technology. All complaints are dealt with by the same main computer. It can quickly identify anyone who goes from hotel to hotel claiming poor service and getting money back. Promus then sends a nice letter, regretting that the hotels could not meet the customer’s standards and inviting him to stay in a competitor’s hotels.

If your customers can contact you electronically with their problems, you need to be prepared to answer quickly. When a customer mails a letter to a company, he or she doesn’t expect a reply for a week, but the customer knows that the e-mail arrives in a few minutes, if not seconds. It’s usual for businesses to respond to e-mail within a few hours or overnight. A few days is “slow.” If you take weeks to reply, customers will take their business elsewhere. Because e-mail is so much easier to send than paper mail, you’re likely to get many more responses, too. So when you ask for electronic responses, make sure you have the staff and the internal systems in place to handle them quickly.

Listen to your customers and take their bad news as an opportunity to turn your failures into the improvements they want. Companies that invest early in digital nervous systems to capture, analyze, and build on customer opinion will stand out from the competition. You should examine customer complaints more often than company finances. And your digital systems should help you change bad news into improved products and services.

Know your numbers.

Jiffy Lube was the world’s number one servicer of cars, but it was unprofitable when the Pennzoil Company bought it in 1991. By 1997 Jiffy Lube earned $25 million, the highest earnings in its history and an increase of 14 percent over 1996 earnings. Jiffy Lube serviced 21 million cars, an increase of 1.2 million over 1996. Driving this new success was a daily flow of information from each store to headquarters and back. Customer service information from all 1,600 Jiffy Lube outlets is sent to headquarters at Houston every night. Headquarters does immediate analysis on the figures. Beginning as early as 5 A.M., up-to-date data is available to each Jiffy Lube outlet. For example, if people in one neighborhood are not using their Jiffy Lube outlet as much as expected the manager of that outlet will know about it.

“Know your numbers” is a basic business principle. You need to gather your business’ data at every step of the way and in every interaction with your customers - and with your partners, too. Then you need to understand what the data means. The numbers shouldn’t be your only concern but you should objectively understand everything possible about your business. If you’re considering exchanging short-term profits for long-term gains, for example, you need to know as closely as possible the cost of that trade. Companies can use the data they collect to improve the efficiency of their businesses, strengthen their relationships with both customers and partners, extend their businesses in new ways, and develop better service and new products.

There are two ways to use customer data. The first is to gather the numbers that show trends and patterns on which to base analysis, planning, and decisions. The second is to collect detailed information on the individual customer so that you can provide personal service. By creating a digital flow of information from start to finish, businesses are able to create tight links between knowledge management, commerce, and business operations. At Microsoft we showed big improvements as soon as we started using MOET (Microsoft Order Entry Tool) to take customer orders, and MOET is now an advanced website for electronic commerce worldwide.

To use information well, you need to capture it digitally from the start and analyze it in digital form at every stage of your business processes. “Every stage” includes not just what happens within your company walls, but also what happens with both your customers and your suppliers. Knowing your numbers can produce big changes in all of your business relationships and give you a significant competitive advantage.

The business side of any company starts and ends with deep analysis of its numbers. Whatever else you do, if you don’t understand what’s happening in your business factually, and you’re making decisions based on soft data or emotion, you’ll eventually pay a big price. Numbers give you the factual basis for the directions in which you take your products. They tell you objectively what customers like and don’t like. They help you identify your highest priorities so that you can take fast action. Nothing else can take the place of understanding your numbers at a working level. Sometimes my friend Steve Ballmer, Microsoft’s president, surprises the members of a product group by knowing their pricing structures and sales numbers - and the competitors’ - better than the people presenting a plan to him. He has a way of marching into a room and immediately asking the one question the team doesn’t have an answer to. He’s done the work, and he’s thought hard about the issues that come out of the numbers. He sets a high priority on decisions that are based on facts.

The middle managers at a company are the people who should be analyzing the numbers. Other groups can help, but the people who deal with customers and competitive problems need to look at their business in every way possible every day. The analysis should support action and not just more analysis. Analysis should lead you step by step to a decision and to action. You have to think, act, judge, adapt.

Starting with digital numbers doesn’t just prevent unnecessary effort and errors. It also sets in motion the best ways to process the data afterward. Being digital from the start drives efficiency in manufacturing, shipping, billing, and other operational processes. Getting the data digitally is also the only way to make sure that you get information quickly enough to respond to customer needs before your competitor does.

This need for good information to drive employees to quick action is one reason that “paper numbers” bother me as much as paper forms do. A printed sales figure or a printed number on customer trends is fixed. You don’t have the ability to get in and analyze the detail behind the number, or to e-mail it to somebody to talk about it. With a paper number that looks wrong, you have to get hold of somebody and say, “I’m looking at this report, and this number surprises me.” A lot of the time there’s an easy explanation: a customer has put in a big order or has canceled one. There’s nothing you need to do, but you still want to know quickly why the month’s results look strange. If you notice a trend in a paper report, it’s hard to send the paper around and get people to analyze it. Over time you stop paying attention because it’s so hard to investigate. When figures are in digital form, knowledge workers can study them, attach notes to them, look at them in any amount of detail or in any view they want, and pass them around to get help. A number in digital form is the start of thought and action.

Digital technology also enables a company to create a network of partners that serves its customers better. You can create a virtual company in which everyone is joined together by commerce, knowledge management, and operational systems. Your partners are better connected to you, benefit more from your success, and naturally respond to the same patterns of customer behavior that you do.

When your information systems are designed to create a flow of information to and from your customers, the business processes of the entire supply system will become more efficient. Just-in-time delivery can be a reality for any industry.

Digital information flow makes it possible to create an organization without limits, but it takes a new company culture to turn suppliers from “them” into “us”. Traditionally companies have not thought of suppliers as a true part of the overall business process needed to serve customers. Today’s approach is that of a “value network,” a network of partners enabled by digital information flow. Everyone who touches the product must add value, and communications go forward as well as back. Companies in the value network aren’t restricted by heavy chains of process, but can interact and do business with many different sellers as they need to.

If the benefits of knowing your numbers and creating a value network are so persuasive, why don’t more companies do it? Why don’t more companies create a digital information flow to observe trends? Why don’t they follow customer history? The main reason is that too few companies start with digital information. Existing paper systems cause people to assume that data is hard to get and work with. They have to work with piles of papers that they can’t organize or analyze. They can’t find patterns in their data. They can’t turn their paper information into action. Because so few companies are using digital tools internally or with partners today, those firms that act quickly to create a digital nervous system have the opportunity to jump ahead of their competitors.

To make the change to a more efficient virtual company, you need to look first at all of the paper on the desks of knowledge workers and ask: How could digital systems get rid of these piles of papers? As part of this inquiry, think of your business processes as extending far beyond your walls to include the entire network of your partners and customers. You need to develop business processes supported by a fast, reliable flow of information that will enable the customer to drive your responses and the responses of all your sellers as if you were a single unit.

Move people into thinking work.

The inevitable result of better computer systems is a smarter use of people’s time. With intelligent software continuously searching through its sales data, following trends, and noticing what’s selling and what’s not, the British chain store Marks & Spencer can use its 500 to 600 buyers much more efficiently.

Instead of working through fat paper reports from the previous day to try to find out whether sales are going well, the buyers can use their time more efficiently, using what the latest data is telling them. If sales are going well, no human action is needed, but the system checks sales data and notices any items whose sales are higher or lower than expected. Reports on these items are created automatically and they are all that buyers must deal with.

Using software to handle routine data tasks gives you the opportunity to provide the human touch where it really matters. There’s a dramatic difference between getting a note that was clearly written by a person rather than a computer. It’s enormously valuable to have a person working with any customer who is unhappy about something really important, or who has special needs. In a hotel, for instance, smart software can dramatically shorten the check-in and check-out time. Staff can then help customers rather than filling in forms, and guests will enjoy their stays more as a result.

Electronic commerce, though, brings new challenges. In a physical store a sales person can use clues such as the customer’s questions, dress style, and body language to assess his or her interests. However, at a Web store no one sees the customer, and the goal is to let the customer do as much shopping as possible for himself or herself. Web store owners then have some interesting information to find out. Based on customer behavior, how do you construct a model of who the shopper is? It requires smart data analysis.

Digital tools for analysis are changing the nature of work. Knowledge workers can concentrate on unusual events rather than on the routine. Of course, people don’t like allowing machines to take their decisions for them. But when a database gets big enough and complex enough, the computer can do the initial searching and sorting far better than a human being. We’re simply not able to recognize patterns in large amounts of data. And the available data - in databases, file systems, message systems, and websites - is growing all the time. The only way we can get the full value of all this data is to use computer tools to find the useful information.

HarperCollins, the publishing company, uses a PC-based system to follow book sales so that it can print just enough books to meet demand. That way it won’t be caught with large stocks of unsold books in stores, which publishers have to take back. After only a year in operation the new system has helped HarperCollins reduce returns of unsold copies of its most popular books from over 30 percent to about 10 percent. Each percentage point represents millions of dollars in savings.

Using software to find useful patterns in large amounts of data is called data mining. It can help to predict whether customers are likely to buy an item because of their age, sex, hometown, and other characteristics. It can also identify customers with similar shopping behavior, and customers with specific tastes, in order to provide improved individual service. An Australian health-care company used data mining to follow buying patterns and discovered a $10 million fraud.

The most common use of data mining is for database marketing, in which companies analyze data to discover customer tastes and then make offers to specific sets of people. For example, American Airlines uses information about the twenty-six million members of its frequent flier program - such as the car rental companies, hotels, and restaurants they use - to develop targeted marketing efforts that have saved more than $100 million in costs.

Data mining is part of customer relationship management (CRM), in which information technology helps companies manage customer relationships individually instead of all together. With the patterns revealed by data mining, you can present your products to a customer in a way that’s most likely to increase your value to the customer, and his or her value to you. Cost savings come from having more exact models of customers and so the amount of mailings is reduced. A direct marketing campaign for something like credit cards, for example, would normally get a return of about 2 percent. Mellon Bank USA in 1997 had a target of 200,000 new customers, which would normally mean a mailing to 10 million people. Instead, the bank used data mining to produce models of the most likely new customers. The bank was then able to meet its target of 200,000 new customers with a mailing of 2 million, not 10 million. In addition to reducing costs, each new customer was three times more profitable than usual because data mining had targeted the customers whose needs best matched Mellon Bank’s services.

This personalization will deeply affect all forms of advertising. As digital TV becomes more popular and electronic books become the preferred way to read magazines and newspapers, almost all advertising will move away from mass advertising to personalized advertising. The commercials that appear on screen will differ according to who is looking at them; different neighborhoods or even different homes in the same area could see different commercials. Big companies can become more efficient with their advertising, and small companies can consider TV and magazine advertisements for the first time. Today many types of advertisement are too expensive for any except true mass-market products. Soon even the neighborhood grocer might be able to afford commercials for people living near the store.

Targeted advertisements should make customers happy. They’re more likely to see advertisements that are relevant to them. Some people may be worried that advertisers have too much information about them, but software will make it possible for people to reveal only the information they want to reveal. For example, most people wouldn’t mind if advertisers were given viewing patterns. Most regular readers of specialized magazines - whether the topic is sports, food, science, gardening, or cars - look at the advertisements as carefully as the articles. If you watch TV in the same way, mainly for one or two interests, you probably wouldn’t object to seeing commercials that concentrate on those interests.

Today most data mining systems are quite expensive, ranging from $25,000 to $150,000 for a small or medium-size business to millions of dollars for a big customer like Wal-Mart. One insurance company spent $10 million for a data mining system five years ago. The boss said he knew he could get the same results for a lot less money with today’s technology, but the benefits had been worth the investment.

That remark indicates the value of data mining, but these high prices reflect the old days of complex software, in which only the largest organizations, using a lot of staff or hiring specialists, could make good use of data. With the growth of competition in our information-based economy, customer data has become an increasingly important asset. Every company and everyknowledge worker must try to get the most out of the company’s data assets. These new users can’t afford big database budgets or specialized database experts.

Fortunately, as data mining systems become available for PCs, you’ll see prices drop dramatically and the use of data mining explode in companies and departments of all sizes. Soon every company will be able to do the kind of analysis that used to be reserved for the richest organizations.

The greatest value of data mining will be to help companies decide the right products to build and the right way to price them. Companies will be able to look at a variety of packaging options and prices to see which are most attractive to customers and profitable to themselves. These possibilities are of special interest to companies that sell information products. Unlike a car or a chair, products such as insurance, financial services, and books cost much more to develop than to produce and have a value fixed more by the customer than by the physical cost of goods. The secret of success with information products is understanding the characteristics and buying habits of your most likely customers.

The power of data mining will help companies to work out how to acquire new customers, whom to market to, how to price their products, and how to attract individual customers. Human creativity and skill are needed to use this information to create new packaging and pricing ideas, to see new products in the patterns of the computer results, and to develop exciting new offers. The better the tool, the more creative they can be.

Managers need to invest in advanced tools that help people work. You should budget 3 to 4 percent of the salary of your knowledge workers to make sure they have the best tools, which free people to use their energy for creative responses to the patterns and trends identified by the computer. Using information to develop new products and services, and to work together more closely with partners and customers, will always remain a uniquely human task.

Raise your corporate IQ.

Like reengineering before it, “knowledge management” has been given so many meanings that it sometimes seems to mean whatever someone wants it to mean. News articles on the topic appear regularly. Consulting companies and websites specialize in knowledge management, and a “knowledge management” magazine started in 1998. If reporters talk to a database company, they find that knowledge management is the newest thing in databases. If reporters talk to a network company, they find that knowledge management means the future of networking. But as a general concept - to gather and organize information, pass the information on to the people who need it, and constantly improve the information through analysis and teamwork - knowledge management is useful.

So let’s be clear on a couple of things first. Knowledge management, as I use the phrase here, is not a software product or a software category. Knowledge management doesn’t even start with technology. It starts with business goals and processes and an understanding of the need to share information. Knowledge management is nothing more than managing information flow, getting the right information to the people who need it so that they can act on it quickly. And knowledge management is a means, not an end.

The aim is to increase corporate IQ. In today’s fast-moving markets, a company needs high corporate IQ to succeed. By corporate IQ I don’t just mean having a lot of smart people at your company - though it helps to start with smart people.

Corporate IQ is a measure of how easily your company can share information broadly and of how well people within your organization can build on each other’s ideas. Corporate IQ involves sharing both history and current knowledge. The workers in a company with a high corporate IQ work together efficiently so that all of the key people on a project are well-informed and full of energy. The final goal is to have a team develop the best ideas from throughout an organization, and then act with the same purpose and concentration that a single, enthusiastic person would bring to a situation. Digital information flow can help groups to function like individuals.

A company’s high-level executives need to believe in knowledge sharing, or even a major effort in sharing will fail. Leaders must also show that they themselves are not locked away in a palace, isolated from everyone else, but are willing to engage with employees. Jacques (Jac) Nasser, operations president at Ford, sends an e-mail every Friday afternoon to 89,000 Ford employees all around the world, sharing the week’s news - the good and the bad - with everybody. No one checks the e-mail. He talks straight to the employees. He also reads the hundreds of responses he gets each month and assigns a member of his team to reply to any that need an answer.

I don’t send out weekly reports, but I do e-mail employees around the world on major topics. Like Jac Nasser, I read all the e-mail that employees send me, and I pass items on to people for action. I find the e-mail that people send me an amazingly good way to stay aware of the attitudes and issues affecting the many people who work at Microsoft.

When business leaders have created an atmosphere that encourages teamwork and knowledge sharing, they need to set up specific knowledge-sharing projects across the organization and make knowledge sharing a key part of the work itself - not an extra that can safely be ignored. Then leaders need to make sure that the people who share knowledge are rewarded. The old saying “Knowledge is power” sometimes makes people keep knowledge to themselves. They believe that this makes them more valuable to the company. Power comes not from knowledge kept, but from knowledge shared. A company’s values and reward system should reflect that idea.

Knowledge management can help any business in four major areas: planning, customer service, training, and teamwork on projects. If you haven’t done any work on knowledge management in your company yet, consider picking one or two areas in which to begin knowledge management projects. You can use the success of your projects in those areas to encourage knowledge management projects in your other business areas. Within a few years all leading companies will have achieved high levels of digitally aided knowledge sharing.

Knowledge sharing can help brand planning across countries. Coca-Cola set up worldwide communication with its own e-mail system in the 1980s but it saw IT as an expense to be controlled rather than a tool to help better business. When this changed, all systems and functions were made compatible worldwide and the biggest user of technology was marketing, not finance. The marketing tool Inform (Information for Marketing) looks at data about consumer preferences in every country. From sales data, it can tell you what types of people in a small town in South Africa drink Sprite every day and how much they drank last March.

Knowledge management can be used to make response to customers faster. Yamanouchi Pharmaceuticals, at $3.9 billion, is the third largest drugs company in Japan. They and Microsoft have made Web-based information systems an important part of improving quality and speed of answers to customers’ technical questions. Product support at Yamanouchi can immediately answer about half the questions that come in from doctors. To find answers to more difficult questions they use Yamanouchi’s Web-based PRoduct INformation CEnter Supporting System or PRINCESS. This uses electronic searches by product and by word. Answers are put into the system for use next time. In 1998, PRINCESS was made available to the sales team. The next step could be to make it available to doctors themselves.

Online training is another area where knowledge management can make things faster and better. US Web trains people to use technology in business. It has developed SiteCast for interactive seminars. Online training has been really popular at Microsoft. In 1998 twice as many people were training online as in class.

Digital information flow can help in product development, too. Nabisco has some of the most popular foods in the world. About a third of its new products are worldwide successes, a third do OK, and another third underperform. These results are better than average, but with competition increasing Nabisco decided to use information technology more in the product development process. Nabisco didn’t need a new product development process. The technology was needed to help the product development team communicate with each other better and faster. A system called Journey now organizes the information that used to be anywhere on any desk in the office. All of the people involved in the development of any new product need only to click on the product to find all the information about it. For example, Nabisco developed a new product and something in the product was changed. The people writing the description of the product on the label saw the change online and changed the label immediately.

Knowledge management is a fancy term for a simple idea. You’re managing data, documents, and people’s efforts. Your aim should be to improve the way people work together, share ideas, sometimes argue, and build on one another’s ideas-and then act together for a common purpose. The boss’s role in raising a company’s corporate IQ is to create an atmosphere that promotes knowledge sharing and teamwork, to make a priority of those areas in which knowledge sharing is most valuable, to provide the digital tools that make knowledge sharing possible, and to reward people for contributing to a full flow of information.

Big wins require big risks.

To be a market leader, you have to have big goals. You can’t just look at the past or the current state of the market. You also have to look at where it’s likely to go, and where it might go under certain circumstances, and then direct your company according to your best predictions. To win big, sometimes you have to take big risks. Big bets mean big failures as well as successes. Today, looking back, it’s easy to believe that Microsoft’s current success was inevitable. But at the time we made our big bets - including starting the company as the first personal computer software firm - most people thought we would fail.

Many leading companies hesitated to move to new technologies for fear of ruining the success of their existing technologies. They learned a hard lesson. If you don’t take risks early, you’ll decline in the market later. If you bet big, though, only a few of these risks have to succeed to provide for your future. Microsoft’s current goals include improving the PC’s performance beyond all existing systems, developing computers that “see, listen, and learn,” and creating software to power the new personal companion computers. These plans are Microsoft’s response to the future, in which all machines will use digital technology and need to work with one another. Whether or not these plans succeed, one fact is clear: we have to take these risks in order to have a long-term future.

Risk taking is natural in a new industry. The computer industry is about as far into its development as cars were in the 1910s and planes were in the 1930s. Those industries went through revolutionary, and often messy, technical and business change before they became stable, and the same thing is happening in the computer industry.

Develop processes that give people power.

A business has processes that are similar to the basic biological processes that keep us alive. One of these processes is the function that defines the company’s reason for existing - its manufacturing process, for example. This function has to be as efficient and reliable as the beating of a heart.

A second kind of automatic process in business is administrative - the process of receiving payments, paying bills and salaries, for example. The administrative processes are as essential to a business as breathing. If the basic operational processes of your business fail, your company fails.

Because the basic operational processes are so important - and so expensive - most big companies began to invest heavily in automating them years ago. But too often the automated processes were isolated from one another. Overall efficiency wasn’t nearly as good as it could be. Until recently, for example, in the manufacture of some aircraft parts only 10 percent of the original metal was actually used in an aircraft. The manufacturing process had been improved at many individual stages along the way rather than as a whole. There was an enormous amount of waste.

I have talked about business operations such as financial and other administrative systems in other sections. In this section I concentrate on production processes. An automated production process is necessary, but not enough on its own to make a company competitive today. A good digital nervous system can help you develop your employees into knowledge workers, changing your company’s basic production processes into a competitive advantage.

First, you need to use information technology to understand the inner workings of the process itself better in order to make it both more efficient and more responsive to changing circumstances. Entergy Corporation of New Orleans has increased the efficiency of its power plants with a new process-control system. This allows plant operators to adjust plant efficiency and analyze performance trends as they occur. Operators can actually look inside the plant systems in order to understand exactly how the machines are functioning and to work out whether a minor change or a repair might save an expensive repair later.

An intelligent PC-based system makes sure that the highest-priority items are repaired first. The process-control system actually shows operators the cost of reduced efficiency if the temperature is ten degrees lower than it should be, for example. By attaching a dollar figure to operational decisions, Entergy is turning its operators into business people, giving them the information they need to run their units efficiently, and making them a lot more responsible for their decisions. And because production costs for Entergy’s plants are available to senior staff electronically on a minute-by-minute basis, the company can improve its profits by moving power production constantly to those units that are delivering the most cost-efficient energy.

Then you need to be able to get data from your production process to inform other business systems. The Stepan Company, which produces cleaning chemicals, has developed a process-control system that has raised the plant’s output by 300 percent and saved the company millions of dollars through more efficient use of its equipment. But the efficiency gains are not as valuable to Stepan as the way that its process-control system can deal with changing customer demands.

Finally, and most important, you need to feed the data from your production process to your production workers so that they can improve the quality of the product itself. If you provide the right technology to help production workers do fast analysis, they’ll turn data into information that will help you improve design and reduce faults. Developing a digital nervous system allows you to give more power to as many of your workers as possible.

In the new organization the worker is no longer a part of a machine but is an intelligent part of the overall process. Some metal workers now have to know mathematics to work out angles from computerized designs. Water-treatment companies train production-line workers how to do computerized measurement and math. New digital copiers require the service personnel to have an understanding of computers and the Internet, not just skill with a screwdriver.

Human beings remain essential in operational processes that have to constantly improve and adapt to changing circumstances. An efficient production line needs people-well-informed, responsible people. As we organize tasks into processes, we give workers more responsibility. Computers will take over some jobs, but they will mean that many other jobs are no longer boring.

When people can concentrate on whole processes, their work is more interesting and challenging. Simple tasks will disappear, or be automated or built into a bigger process. Work that involves repeating the same task over and over again is exactly what computers and other machines are best at - and what human workers are poorly suited to and almost always hate. Managing a process instead of performing a task makes someone a knowledge worker, and good digital information flow enables knowledge workers to play their unique roles.

Most companies have been willing to give information tools to their high-paid office staff whose job is information work. Entergy and Stepan are proving that building systems around information flow and giving information tools to line workers can also provide enormous value. Entergy is changing all of its key business processes and pushing information and decision-making down to the operational level. Stepan is using information to manage its plants as a whole to adapt to changing customer needs. Both of these companies are applying knowledge management to business operations to analyze and improve production, quality, and failure rates. Digital tools bring more intelligence to their business operations.

Give your workers better jobs with better tools, and you’ll discover that your employees will become more responsible and bring more intelligence to their work. In the digital age you need to make knowledge workers out of every employee that you can.