(FORTUNE Magazine) – WHEN McKesson Corp. wanted to automate its big drug distribution center in Spokane, it turned to Don Bennett. He's no high-priced computer consultant or programmer. McKesson hired those from EDS, the computer services giant. What Bennett, a receiving clerk with 41 years at McKesson, and his fellow warehouse workers did was to bring their considerable practical experience to bear on developing a small scanner-equipped computer they now use to fill orders. Each device communicates by radio with an IBM minicomputer, tracking the exact location and status of 18,000 different items stored in the building's vast expanse. Says Bennett, 61, of the new system: ''We designed it, more or less. They just put it in.'' That's how McKesson solved a problem that rising numbers of companies in every industry face: getting a work force that's less and less skilled to buy into tools that are more and more high-tech -- computers and complex software. Many workers are suspicious of new technology, even hostile to it. Bennett does not even own a calculator. It's a truism that information technology will reshape virtually every company that survives the 1990s. But it can't reshape yours if you don't find some way to make your workers comfortable with computers. Don't despair. Companies committed to using technology to increase efficiency and speed are finding they can introduce the most sophisticated electronic tools to almost any work force. But it takes a lot of forethought and planning, even a measure of humility among managers. It also takes a lot of money. Estimates of how much companies in the U.S. spend annually educating users of information technology range from $2 billion to $5 billion. And that's just for outside help. Uncounted additional billions go for in-house training. But the payoff is clear when productivity climbs as a result. Motorola calculates that every dollar it spends on training delivers $30 in productivity gains within three years. Almost every white-collar job in America requires some level of familiarity with computers, says Mitchell Fromstein, CEO of Manpower Inc., the largest U.S. temporary-services firm and a leading computer trainer. He estimates that 75% of industrial workers also need at least elementary computer skills. Even if you're running an electronic sweatshop with a revolving-door work force, you've got to understand what enables workers at any skill level to master their computers. Companies that have transformed their work forces with technology have distilled a set of principles that apply equally to workers on an assembly line, clerks in a store, an office full of number crunchers, or salespeople scattered across the globe. Among those hard-won lessons:

-- Think of how to empower your workers, instead of dumping technology on them. To succeed, technology must come to the employees, taking into account their skill levels and their needs -- not like the early days of computer automation, when workers often were expected to learn to use whatever equipment management stuck in front of them, whether it helped them do their jobs better or not. Says Dave Ellett, director of customer training and education at EDS: ''You'll get only minimal impact if your approach is to install new technology and then train the people. That's what companies have been doing for 15 years, but it's also one reason we've had relatively small productivity gains from computing.'' Dale Wartluft, a management consultant at Ernst & Young, adds: ''The most advanced enterprises have realized that they've got to deal with the people side at the same time they deal with the technology.'' Stockbrokers are the kind of knowledge workers who demand attentive care and feeding. Good ones switch firms at the drop of an option. So when Paine Webber decided to install a $75 million trading and information system for its brokers, it studied their needs and wants. The system would upgrade 5,200 brokers in 264 offices across the country to state-of-the-art IBM PS/2 PCs networked together through IBM RS/6000 workstations. The new machines replace dumb Quotron terminals connected to a mainframe, but Quotron remains the primary supplier of financial software and data. The goal: to speed up transactions and give brokers easy access to far more information about their accounts. The system offers them not only power but also control. Brokers are encouraged to pick any financial software they like to analyze data for customers and do their own record keeping. They can store any data they want on a floppy disk. Many securities firms don't give brokers a floppy or even a hard disk, fearing they will record precious client data and spirit it away to another firm. But Paine Webber decided to trust its people, hoping they would in turn trust the system and use it more enthusiastically. Robert Benmosche, Paine Webber's systems operations chief, started by surveying the brokers' attitudes. ''They surprised us,'' he says. ''A third told us that the ten-year-old technology they had on their desks was competitive and met their needs. Another third said they'd like a few more things. The final third said, 'We're nowhere, this place is terrible, get me a mainframe on my desk.' '' Benmosche's response was to create a dazzlingly capable system for the techno-hounds, but build the old system into it. That leaves it up to the individual brokers just how much they use. They can work exclusively on an emulation of the old system or use as many new features as they feel comfortable with. Paine Webber had special keyboards built by IBM with green keys for old Quotron features, white keys for using Microsoft's Windows ! operating system software, and blue keys for the new specialized applications. When it started rolling out the system branch by branch late last year, Paine Webber didn't emphasize the technology itself but rather what it could do for brokers. The firm didn't talk about the multiple sources of data streaming in or the client-server architecture or the operating system. Says Gerard Higgins, director of systems development: ''The broker doesn't know, nor should he care, where all the information is coming from.'' The instructors are Paine Webber back-office workers. Says Higgins: ''Instead of a bunch of trainers who understand only Microsoft Word, we have people from the dividend area, the margin department, operations, and other areas. The brokers are incredibly impressed with what they know.'' At the office in Iselin, New Jersey, where the system has been installed longest, brokers revel in it. Steve Rothman, a 13-year veteran, had never owned a PC before he got the system. Now he has bought one for his home as well. He saved a big account that had threatened to leave unless he could provide more detailed records including unrealized gains and losses, for example, which would have been impossible before. Says Rothman: ''The old stuff we had was just quotes. This is like going from a Model T to a Ferrari.'' By contrast, Rothman's office neighbor, Francis Keegan, a Paine Webber broker for 27 years, calls himself a dinosaur. ''I'm not a natural when it comes to computers,'' he says. ''When I heard this was coming in, I shuddered. But giving me the old Quotron mode has allowed me to swim without sinking.'' Keegan gets increasingly excited as he demonstrates what the system can do. ''There's a zillion ways to go,'' he says. ''It absolutely saves time. At the end of the day you make a couple of extra calls because you can get this stuff off the system easier.''

-- Listen to your own Don Bennetts. Managers of highly automated operations are virtually unanimous: If you don't involve the users, you'll develop the wrong system. Nobody understands the job like the people who do it. They can tell you how to design the tools that will let them work more efficiently. They will trust new technology more if they had a say in it and knew it was coming, and training will be drastically simpler. When you select trainers, try, as Paine Webber did, to draw from among that same worker pool. There's a substantial element of psychology in successful technology introductions. The secret is making workers feel part of what would otherwise seem a wrenching and foreign set of changes. While the input of Don Bennett and other McKesson workers was certainly critical, they did not actually design the new system. But the company surely won more commitment from its workers because they knew their contributions were indispensable. Employee commitment at McKesson would have been harder to achieve if management hadn't promised that the new technology, introduced in early 1992, would not eliminate jobs. The aim instead was to increase the accuracy of what's called picking. In the past clerks had to flip through long paper printouts with lists of orders, crossing out items as they packed them in boxes for shipment to stores. Many times they didn't get the orders quite right. McKesson calculated that while it cost $8 to do the initial pick, correcting a bungled store order cost ten times as much because bills had to be corrected and new products shipped extra quickly. Guided now by the hand- held computers, worker accuracy has improved 90%. The new machines are built by Symbol Technologies of Bohemia, New York, which specializes in hand- held wireless devices. The computers even figure out the most efficient route through the warehouse for putting together each order. EDS worked closely with McKesson developing early computer prototypes. To get a better sense of exactly what was needed, several managers spent a few days working as clerks. Then EDS brought the prototypes to the distribution center and asked workers to try them out. Says Doug Hoover, the EDS vice president in charge of the project: ''We said, 'You people here are evaluated based on your productivity and service. Here's a tool to do that better. What do you think?' '' Based on discussions with the workers, EDS made more than 50 modifications to the device. For example, initially the laser scanner was triggered by moving the thumb, but that gave some people sore hands. Now workers point at the item and tap a button with their index finger to activate the laser. Much of the training on the new system was the kind EDS's Ellett calls ''Go ask George.'' The first instinct of most baffled computer users is to consult a more knowledgeable friend or colleague. ''You've got to plant the right Georges in the organization,'' says Ellett. ''He may not be a real guru, but at least he's one step ahead of the person asking the question.'' McKesson - asked for volunteers who wanted to learn a little extra, and now encourages other workers to go to them with problems.

-- Understand and communicate your business objectives. People will accept and learn new technologies if they understand their importance. Fancy computers seldom make much difference in productivity if workers don't clearly see how the machines help achieve business goals. The way Texas Instruments went about remaking a factory for radar-seeking missiles underscores the point: It's important to see new technology as only part of a total vision of a changed organization. In the mid-1980s, TI was selling 25 of these so-called HARMs (high-speed antiradiation missiles) a month to the Pentagon for $1 million apiece. But then the government decided it wanted 250 a month, at a much lower price, and said it would shop elsewhere if TI couldn't deliver. So the company undertook to make ten times as many HARMs profitably for $200,000 each. By the late 1980s it had met the goal, using the same factory and without ever stopping production. Nobody was fired. Management put together so-called cross-functional teams from manufacturing, engineering, information systems, and human resources. About one-fifth of the team members were shop-floor workers. Recalls Cindy Johnson, a TI process- engineering expert: ''We didn't say, 'Let's give these floor workers computers, but rather, 'Let's look at the information they need, the materials they need, the incentives they need' -- all the aspects of the business, not just automating.''

TI didn't just put new technology into old structures. The jobs themselves changed. For example, it suddenly became too expensive to have hordes of supervisors roaming the floor, so regular employees now have IBM PCs right next to them to relay information supervisors once collected. Workers can navigate through graphic representations of every stage of the assembly process. TI also devoted a lot of effort to measurement and goal setting. The team used banners and posters to announce targets and gains, and threw a party every time it achieved a new level of production. ''We had something we called the HARM mountain, which showed leapfrog jumps from 30 to 75, 75 to 150, and 150 to 300 a month,'' says Johnson. ''Everyone from the shop floor up to the VP level was constantly monitoring our progress against that mountain. The No. 1 lesson we learned was the value of having a strong business case that people can rally around.'' By including the workers at every stage, the company found putting technology into their hands to be almost a nonissue. It just happened. And it worked. Says Johnson: ''It's threatening to say, 'We're going to give this computer to you to improve how you do your job,' because that places the burden on employees instead of asking them to help meet the overall goals of the business.''

-- Teach your employees by helping them improve their performance. Most important is learning to do the job better, not learning how to operate the computer. Traditional classroom instruction is seldom the best way to go. Often the most useful training comes only when workers need it. The software industry is already moving to ''just in time'' instruction. The best off-the- shelf software today includes built-in help commands that allow users to call for assistance when they need it. They can proceed to various levels of detailed instruction at their own pace. A few companies are already trying that approach with applications software built specifically for their businesses. They want to let workers ask for help only when they're stymied and get what they need instantly. You may not be able to hire the perfect worker for what you can afford to pay, but you can use technology to give employees a brain boost, creating a kind of cyberworker. Dreyfus Funds, for example, is upgrading its computer systems for customer service representatives. To answer questions customers ask on the phone, these reps have been using five mainframes and large bookcases full of binders. Callers need help with matters ranging from how to open an account to why a given fund performed as it did. Ernst & Young's consulting division is helping Dreyfus tie all the data sources together on a single PC-based system that uses artificial intelligence software to help the reps find answers faster. Compaq, Dell, and IBM supply the hardware. Dreyfus wants the technology to compensate for the inevitable limitations of its work force. Says David Shpilberg, who heads financial services consulting at Ernst & Young: ''Ideally you'd like to have a renaissance person who can comprehend the complexities of the problems, has the patience of Job, is an effective communicator, and doesn't want to earn over $30,000 a year. So what do you settle for? A good actor who enjoys interacting with the client and can read scripts.'' The new software will query the representatives about the client and respond with specific suggestions. This way of compensating for the limitations of a work force should become available to more and more companies, even those with rapid turnover in near- minimum-wage jobs. Workers will merely have to read a script, or emulate what they see in a video. At EDS, conventional classroom education -- a lecturer in front of a roomful of silent students -- is passe. Instead, the company increasingly uses simultaneous interactive video training for workers at hundreds of locations around the world. With a combination of videoconferencing equipment and computer networks, students anywhere can see the teacher live, and the teacher can administer pop quizzes to make sure they have understood the material. Special software tots up the responses and lets the instructor know instantly how many students have failed to grasp a point. The cost of delivering that kind of training has plummeted. Until recently it required a dedicated satellite link, but advances in switching and data compression now make it possible on regular telephone lines. It's roughly one- fourth as expensive as it was three years ago.

-- Don't ignore the generation gap. People who grew up playing Nintendo games have a built-in advantage over their elders, even if they've spent their lives doing almost nothing else. At McKesson's Spokane warehouse, it turned out that one of the least productive clerks was a computer game whiz. He blasted his way to the top in output once he got his hands on the new hand-held laser scanner. Conversely, big-time computer klutzes may slip in the pecking order if they can't handle new technology deftly. Ernst & Young's Shpilberg and other experts recommend that companies automating groups of total computerphobes loosen up the workers first by using computer games to make them comfortable interacting with a screen. Too many senior managers still show signs of a precomputer upbringing. Shpilberg says technological illiteracy at the top plagues fully 90% of the companies he deals with. Yet there's no better way to get corporate hoi polloi to use sophisticated computerized tools than to let them see the boss doing it first. ''Planning for the effective use of information technology is still an unnatural act for most managers,'' laments Shpilberg. The first worker who has to be brought up to speed on technology is you.