What the hell is a word processor anyway? A word processor is, in the hyperbole of the times, any piece of office equipment more intelligent than a chair. Telephones, dictation equipment, tape recorders, typewriters, smart typewriters, memory typewriters, copiers, and memory copiers are all word processors. No wonder managers of businesses get confused and nervous when someone suggests they should "look into word processing."

But let's assume that what you're most interested in is a machine that will simplify and speed up your office's paperwork -- letters, reports, memos, forms, contracts, records, lists, documents, notes, and so forth and so on. Then what you probably have in mind is one of those devices that consist of a keyboard, a screen, and a printer. When connected to a "memory" and a small computer, appropriately programmed, these elements become a word processor. What the user types on the keyboard is displayed on the screen and stored in the memory. At the stroke of a key, the printer pours forth whatever is stored on the disk faster than the fastest typist. There are "screenless" word processors that have at most a single line of display, but their functions, speed, and expandability are limited. We'll assume that what you want is a full-display system that shows you from 24 to 54 lines of whatever you've typed on it.

If you already have a microcomputer, you can turn it into a word processor by adding a word-processing software package. The micro won't do as many tricks as the so-called "dedicated" word processors, however, and microcomputer dealers don't provide the service and support that the dealers in dedicated word processors do. Still, you can get a good word processor for under $8,000 this way.

Also for $8,000, you can get the simplest dedicated word processors -- the so-called writer systems. They're designed specifically for text processing, but they're limited in speed and memory. The next step up, where you'll find the widest choice of functions and options, is to systems priced between $8,000 and $12,000. If you need really sophisticated functions, such as the ability to handle scientific equations, you may have to go over $12,000. But be cautious at this level: There are a lot of older systems up there that are no longer competitively priced.

That's what a word processor is, but what does it do? Take a letter, please. Very few of them are dictated, typed, and mailed without some revision. If a word is misspelled, misused, or misplaced, the whole thing must be retyped. Even the simplest word processors permit changes to be made quickly and easily. The text can then be printed again -- unattended.

But today's machines can do a lot more than that.They are handy for sending out form letters that don't look like form letters. Someone puts a list of names and addresses and perhaps some personal information such as nicknames into the computer's memory. This, in computer jargon, becomes a "data base." A master letter is prepared with special symbols in the places where the computer is to insert names, addresses, and nicknames. At the touch of a button, the word processor generates hundreds of "personalized" letters ready to be folded, stuffed, stamped, and mailed.

You can also add information to the data base so that useful sub-categories can be created. A stockbroker can send one letter to clients interested in speculating and another to those who buy only blue chips. A physician can send a polite reminder to patients who have neglected their bills for 60 days, and a not-so-polite letter to deadbeats.

But there's a lot more to word processing than managing a data base and sending out form letters. The computer can store and recall whole paragraphs, even pages, of text, and insert them in the body of an original document. This is useful in documents like contracts, in which some elements are always the same, but others have to be changed to fit the particular case. If you discover you've misspelled Zbigniew Brzezinski's name throughout 100 pages of text, there's a command that will find all the Zbignews and change them to Zbigniews.

If all this has made you eager to run out and buy a word processor, maybe you should calm down a bit first. There are more than 150 different word processors/personal computers to choose from, and at least as many companies selling software for these machines. What's more, industry standards are few, so almost nothing that works on one system is compatible with any other. If you decide to trade in the machine four years from now, you may find that your data bases won't be readable by the new equipment, particularly if it's made by another company. The disks produced by some entrylevel word processors are not usable on the more advanced systems of the same make. That's like recording a Beethoven symphony on your home tape recorder only to find that it won't play on your car's tape deck.

Furthermore, the machine you buy today is likely to be outdated in two years by products costing half as much. That doesn't mean you should wait. If the computer can save you time and money now, then by all means go ahead and buy it.

The point is, your word processor is a major first step in automating your office to make it more efficient and productive. So it's a step you should take some time in making. The time we're suggesting here is roughly 90 days -- three months of figuring out what you need, finding out what's available, and making sure you get the most word processing for your money.

Days 1 to 30: What kind of word processing do you need? Mark Nigberg's ad agency has many high-technology clients, so he knew a little about word processors. He figured it would be easy to bring word processing into his office. He went to a computer show, looked at a couple of systems, and picked up some brochures. The next day, he called a manufacturer.

The salesman asked what Nigberg was looking for. "Well," Nigberg said after a moment of hesitation, "we're a 15-person ad agency with $2 million in billings." The next thing Nigberg knew, the salesman was proposing a $50,000 office-automation system.

"It was like being docused with cold water," Nigberg says. He was certain the system the salesman was pushing was much too elaborate and expensive, but, he says, "I had no idea how to evaluate a word processor in terms of our needs."

Managers of small companies often think what they need in word processing will become obvious as they look at more equipment, so they skip the important first step -- assessing their office's needs. But they end up paralyzed when they find out how complicated, varied, and expensive word-processing systems can be.

Often, too, they find that they don't really know how their own offices work. They haven't taken the time to analyze the information flow, to measure the volume of paperwork, or to figure out how much their paperwork will increase as the company grows. That's why it's important to spend about 30 days figuring out what kind of word processing your office needs.

Paperwork that can be handled by a word processor falls into eight categories. The first step should be to determine how many of these kinds of paperwork are being handled by your office staff:

1. INDIVIDUAL CORRESPONDENCE: How many one-of-a-kind letters does your secretarial staff turn out?

2. REPETITIVE CORRESPONDENCE: Do you send out batches of letters to groups of people such as sales reps, customers, or stockholders, in which the only changes are the addresses and perhaps a few numbers?

3. INTERNAL MEMOS: How frequently do you communicate with your staff by memoranda?

4. REPORTS: How many of these do you produce? Are they edited and revised extensively?

5. PRINTED FORMS: Do you use purchase orders, travel vouchers, etc. ?

6. LEGAL DOCUMENTS: How many standard contracts or government forms does your company handle that require cutting and pasting to put in "boilerplate" paragraphs?

7. STATISTICAL MATERIAL: Do you produce spread sheets or other reports with tabular lists of numbers?

8. RECORDS: Do you keep mailing lists or other files that need to be changed or updated periodically?

Once you've identified the kinds of paperwork your office handles, find out how much of it is produced, how frequently, and how much it's likely to increase if your company grows according to projections.

As you begin to define your need for a word processor, base your assessment on what your company should be doing, not on what it's doing now. What information do you need that you aren't getting now? Would more detailed sales information, for example, help you plan better? Would your cash flow improve if you could get invoices out faster?

Look at your office as a system, noting especially the interaction between your clerical and your professional staffs. "Often management is unaware of what's happening at the support level, and the truth is, they don't want to know," says Thomas Billadeau, president of Automated Office Systems Corp., a Boston-based office automation consulting firm.

Remember that you can't computerize a bad system. If the work and information flow in your office is chaotic, then a word processor won't make much difference. Get those systems straightened out first, then start thinking about word processing again.

Once you've determined your needs, you should put together a committee of two to five people that will make the final selection. Include people who will actually be using the machine, as well as managers. Take the committee out to a few dealers to see several word processors in operation. At this point, it doesn't really matter which systems you go to see. All you want is an understanding of what a word processor can do. And make it clear to the people demonstrating the equipment that you're not buying anything yet.

Except to be dazzled and intimidated. Word-processing software can do all sorts of tricks and salespeople's mouths are full of jargon. Marke the salesperson explain things clearly, and don't drop your air of "We're just looking, thank you."

By the time you've seen what a few word processors can do, you should be able to put together a set of criteria for your office's system. Mark Nigberg, for example, went back and did a careful assessment of his ad agency's word-processing needs, then developed a checklist that would help him make the final decision:

* The system had to improve his secretary's typing efficiency, of course. But that alone was not enough to justify its cost, since when he analyzed the volume of typing she did, Nigberg discovered that it took only 14% of her time. The rest was spent on filing, on logistical tasks like travel arrangements, on maintaining mailing lists, sorting mail, and answering the phone.

* The system had to provide administrative support by handling filing tasks, storing documents so they could be retrieved easily, and doing electronic proofreading -- checking all work produced on the system for spelling errors and typos.

* It had to provide management support by handling budgets keeping records, and monitoring Nigberg's calendar.

* It had to fit into his secretary's work area. The terminals of some major systems were too large for his secretary's desk. They required a separate work station that would have to be set up in another part of the office, away from the telephones she had to answer.

* The price had to be reasonable. Nigberg wanted to stay under $10,000.

* Expanding the system had to be affordable. Some systems required too big an investment each time a new function was added.

Now that you've determined your needs and seen what a few word-processing systems are capable of, you should estimate whether investing in a word-processing system will have a sufficient payback. "You're in business to make a profit," says Paul Truax, president of Truax & Associates, a Wilmington, Del., consulting firm, "and the only reason to consider a word processor is if it will enhance that objective."

"The system should pay for itself in a year and a half to two years," says Patricia Seybold, editor of The Seybold Report on Office Systems. "Assume it will be obsolete in that time. That doesn't mean that you won't continue to use it, but by then a better system will have come along."

There are several ways to figure the quantifiable improvements that a word processor can make in your office. Determine, for example, how much you currently spend on outside clerical services, temporary help during rush periods, and overtime. If you're under pressure to add more office staff, figure the cost in additional salaries. And if your managers are currently called on to produce periodic reports, estimate the number of hours involved in producing them and figure out what percentage of the manager's salary pays for the time spent on those reports.

"It's naive to cost-justify word processing only on the secretarial time saved," says Billadeau. Most salary expenses go to professionals and management, so any word-processing system should also be evaluated in terms of its effect on their time. If your managers currently spend 64 man-hours a month preparing reports -- drafting, dictating, proofreading, editing, and correcting -- you should be able to estimate how much of that time can be cut by more efficient office systems.

If the payback on a word-processing system seems likely to come as quickly as Seybold suggests, before the changes in technology make the system obsolete, then you're ready to start shopping.

Days 31 to 60: How do you avoid being sold something you don't need? While you were figuring out what kind of word processor you need, you collected a lot of literature on various systems. All of them claim to be "easy to use" or to "improve your office productivity," but those buzzwords are useless in comparing one product with another.

The product literature has its uses, however. It can give you a sense of the marketplace. And with your list of criteria, you should now be able to choose a dozen vendors you want to visit. There's a good chance that you may want to go back to some of the vendors you saw when you were "just looking" last month. You'll probably add several major manufacturers and some of the newer and less well known systems you have come across in your reading.

Set up demonstrations with the vendors on the list, making sure that all of the members of your committee are available for each demonstration. If you leave someone out now, you may run into trouble when you come to the final decision next month. You'll probably also eliminate four of your dozen vendors at this point. Two won't have products in your price range, another won't have a dealer in your city, and the fourth one won't return your calls.

At the demonstrations, the salesperson usually handles the sales pitch, while the marketing services representative operates the machine. It's critical, however, that you or your secretary sit down at the keyboard and have the rep turn the machine off and then on. Make notes on how long it takes for the machine to get working again. Then have the rep take you through some basic functions that you know you need. Make notes on how difficult or easy they are.

Making notes is important. After a while, all the systems will begin to blur together, so your notebook will be invaluable. Write down how readable the characters on the screen are, how well the keyboard is laid out, how much disk storage there is, and, in general, how easy the system is to use. And always note the problems you encountered during the demo.

In the second half of the meeting, find out about the manufacturer. How large is the company? How many systems has it installed locally? How many of these are in businesses like yours? When can the company deliver? Will the vendor give you names of some users who have offices similar to yours?

As you go from vendor to vendor, you should begin to take advantage of their competition with one another. Mark Nigberg, for example, was very impressed by the automatic spelling verification of one of the first systems he saw. But when he asked a competitor why his system didn't have such a feature, the second salesman pointed out serious defects in the first company's machine. Nigberg returned to the original vendor and learned that there were indeed problems with spelling verification in the system.

You can probably eliminate three more systems in this first round of demonstrations. One will be too expensive. Another will be uninterested in companies as small as yours. And at a third demonstration, you will decide you just don't like the way the system feels.

You now have a list of users of the five remaining systems. Visit them and see their systems in operation. Talk to the secretaries who use them and get information about the quality of the dealer's service and support. Remember that every minute the system is not working up to capacity is costing you money, so pay attention to the reliability of the system and of the dealer's service. The dealer's support also includes training on the system, so make sure you have a sense of how quickly the dealer responds when you have a question or something goes wrong.

Nigberg left one showroom demonstration with the impression that his secretary could learn to use the machine in 15 minutes. Users of the system convinced him it would take three weeks. "The users bring you down to earth," he says.

You should also check out the financial stability of the vendors you are considering. "When I'm dealing with a company under $100 million," says Nigberg, "I don't have any qualms about calling the president and asking, 'Hey, will you be here in five years?' You've got to get the flavor of their financial situation." And with the giant corporations, you may find that they're in a hurry to get rid of unprofitable office-automation divisions. Find out what stock analysts are saying about the corporation's word-processing division.

Your second round of showroom demonstrations should be a contest. Put together a sample package of the paper-work you expect your word processor to handle. Include some of the most difficult typing jobs you have, such as statistical typing or footnoted reports. Take the identical package to all of the vendors remaining on your list and have one of their trained operators reproduce the sample on their system, while you time the operators.

Every system remaining on your list at the end of this round should be capable of doing everything you will demand of it. You may, of course, find that you have to revise your original criteria somewhat. Nigberg, for example, decided that any system meeting four of his six major criteria would remain in the running, since no system met all of them.

Don't leave anything to doubt at this point. If you except the word processor to communicate with a computer you already have in-house, make sure it will. The salesperson can demonstrate this capability by hooking up the word processor in the showroom to your computer. You'll need a modem and external communications software to do so, but it's very important that you make this test. Be sure you see with your own eyes that the word processor will actually do everything you expect it to do.

By now, you should be down to three finalists. You are ready to ask them for bids and sample contracts, which should include rental, lease, and purchase options. Make sure they know specifically what you want in the system -- for example, a certain printer, a minimum amount of disk storage, and training for so many operators. And it doesn't hurt to let them know that they're competing with other vendors.

Days 61 to 90: How do you decide which one you want? You've come a long way in two months. Now you're casually using terms like "bi-directional printing," "global search," and "systems diskettes." You're even giving others advice on word processing.

By now you've received proposals and sample contracts. Go over them carefully with each vendor and make sure you know exactly what they contain. Sometimes, because competition is fierce, manufacturers will omit a critical item to keep the price down. (One of the most commonly omitted "extras" is a sheet feeder for the printer.) As you go over the proposals, make sure you're comparing apples with apples, and that the lowest-priced system actually contains as many of the essentials as the higher bids.

Now there are five factors that will influence your final decision:

* Your company's requirements.

* The cost of each system.

* The vendor's reputation for service and support.

* Your feelings about each system and the people selling it.

* The feelings of your secretary (or the primary users) and the others involved in the decision.

Don't discount the last two -- the feelings you and the other members of the search committee have about the systems. The decision should be based on reason and fact, but you should also feel comfortable with the system. At least, having gone through the steps we've outlined, you have some logic to balance against the emotions. The decision will be easy if everyone on your committee wants the same system. But suppose you want one system because it's easily expandable, while your secretary wants another because it's the easiest to use. What do you do?

"The rule of thumb is to let the user cast the decising vote," says Seybold, "unless there are overriding concerns about long-term growth. In this case, that would tip the scales toward the more expandable system." You'll have to decide how much risk you take if you override your secretary's vote.

Another decision may also face you: the choice of buying, leasing, or renting. If this is your first try at word processing, you may want to rent for a year to make sure the system is the one you want. This is practical only if the dollars make sense, and new depreciation rules make it less attractive than ever. There is a large resale market for well-known systems by large manufacturers, which could influence your decision in the direction of buying, but selling used computers has always been risky.

Once you make the choice, don't let your guard down when you sign the contract.Because the market is so competitive, you have more leverage than you realize. Because they understand contracts, lawyers who buy word processors are notorious for writing clauses in the agreement that say, for example, if their system is not repaired within 48 hours, they will be reimbursed by the vendor at an hourly rate. If you buy any new, unproven system, you should make sure the contract includes safeguards, guaranteeing that the word processor will perform all the functions the vendor says it will.

Service contracts are one of the most misunderstood elements of a word-processor sale. Vendors don't price these contracts, which amount to a kind of insurance policy, to lose money. If you have done your homework and bought a system with a reputation for being reliable, you may not want to sign a service contract, under which you will be asked to put up about $1,000 a year to cover repairs and maintenance. It's probably best not to take the service agreement until you see some pattern of repairs developing, so check to see if the service agreement can be started after the system is delivered.

As soon as you've signed the contract and placed the order, the primary user should begin training on the machine, and at least one backup user should have been trained by the time the system arrives. Make sure, too, that the office is physically ready for the machine. Electrical connections are critical here. If you put the system on the same circuit as your copier, you may find that every time the copier is used it creates a power drain on the word processor, which can create all kinds of problems. A dedicated electrical line is the best way to go, but voltage regulators and isolators are cheaper and sometimes just as effective.

Once the machine arrives, don't expect an instant surge in productivity. Mark Nigberg's secretary spent four hours typing a two-page letter. And while she was wrestling with the keyboard, all the other people in the office put down their work to marvel at the high-speed printer. Two weeks after the system was installed, Nigberg called the manufacturer and told him he thought he had made a mistake. The manufacturer said to give it time.

He did, any by the second month, the system was so effective that Nigberg had two more installed. Suddenly, the company's clerical workload dropped so remarkably that Nigberg found he had to reduce the size of his staff. The system was well on its way toward paying for itself.