The Silicon Jungle by David H. Rothman

different from the kind used at large military installations. Writing software for micros is like writing short stories instead of novels. You’ve got to be more elegant and get to the point faster. A micro’s memory simply lacks the capacity of a mainframe and brooks less sloppiness than would a larger machine. Many great novelists could never write decent short stories, and many first-class mainframers just couldn’t code cleverly or deviously enough for micros—couldn’t electronically trick them into thinking they were larger machines. “What happened,” said Hillard, “is that we were paying for Brown’s training on how to use FORTRAN on a micro. He spent forty or fifty hours developing specifications for the data base and writing the code. This went on and on. Brown kept running into all these problems, and he could have cleared up some with just a quick call to the manufacturer.” Meanwhile, the money clock was ticking. Several months later, thousands of dollars richer from those $30.02-hours, Brown threw up his hands. “This isn’t going to work in FORTRAN at all,” he said. “Maybe we should do this in BASIC.” “So,” Hillard told me, “all the work he’d done up to that point was out the window.” “And on top of that,” Stewart said, “we were anticipating our new software package and didn’t care to feed the information twice into the computer. So we built up a backlog of well over 10,000 articles waiting for the software that never came.” Brown’s clock ticked away for several thousand dollars more in time, bringing the total billing to $8,630.81. Then it stopped. Stewart Research had simply run out of money. Stewart’s corporate sponsor could send no more. “Their pockets,” he said, “just weren’t deep enough to subsidize both our own computer literacy and that of our consultants.” When I talked to Stewart in mid-1983, his computer was doing word processing and other minor chores. “But,” he said, “it does hardly enough cross-referencing to be a major data bank.” His research firm faced some $30,000 in debts, including much of the $2,500 bill from the consultants with the library-sized building; and he and his wife were the only two employees left in Stewart R & D. He still hoped to get the data base going, however. “We’re going to become computer consultants in self-defense,” said Stewart. “I’ve done nothing but work on this computer for a year.” It would be nice to be able to do the same, but most of us lack the time or talent; the challenge is to find the right computer consultant so you won’t end up feeling you must do the work on your own. Not so coincidentally, Hillard himself actually did go on to become a computer consultant part time. He wasn’t the smartest or the best credentialed, but he had what many of the more established consultants lacked—empathy and humility. Don’t shrug off those traits. The real question, as one consultant pointed out to me, isn’t just whether someone is competent; it’s whether he’s competent enough at the job he’s doing for you. Never forget that when pondering the “Who?” and “What?” Stewart and Hillard obviously aren’t the only people who didn’t ask the right questions in time. A sail maker wasted $18,000 on a computer system recommended by a consultant. The machine didn’t work right with the hard-disk memory, and it lacked proper shielding against the 60-cycle hum given off by power lines. By the time Hillard was on the scene, the consultant had left town. He says the computer was such a disaster that “I suggested to the client that he get a length of rope and use it as an anchor because it was unfixable.” In another case, a truck-rental firm spent $30,000 for $15,000 of computer equipment. And then the consultant didn’t even supply instructions to operate and modify the software. “As far as I know,” said a more ethical expert, “he’s still a consultant. Computers are mystical, and most people don’t know when they’ve got a good system or when they’ve got a bad one.” “The real problem,” Hillard correctly said of consultants, “is that the microcomputer field has only been around since the mid-seventies and things are moving so rapidly. It’s hard for anyone to keep up, consultant or layman.” There is, too, a nasty economic fact. A micro consultant can’t reach full prosperity, in many instances, without having ties with stores or manufacturers. Indeed, most micro consulting nowadays is done in effect by sales reps recommending systems to their customers. How objective can they be? Even conscientious consultants—independent or those with ties—may still favor the computers and software with which they’re most familiar. The man and the machine in effect become one package. Another anticonsultant argument is that small computers may soon be inexpensive and simple enough to make consultants much less necessary. You don’t use a consultant to buy a $1,000 typewriter. And by the time you read this, or a few years afterward, we’ll have good, complete word processors selling for no more—with printers. Yet another anticonsultant argument is that you may have computer-wise friends in your own business or profession, smart people you trust. They may actually anticipate some of your needs better than would a consultant. Moreover, it’s been said that if you know enough to pick the right consultant, then you really don’t need one. I’m not sure. As one consultant pointed out, it’s like choosing doctors. You can be good at choosing one without necessarily knowing brain surgery. I was a borderline case. I still might have bought a Kaypro without Michael Canyes’s approval, since I knew another writer familiar with the glories of the Osborne who appreciated the greater glories of the Kaypro. On the other hand, with my consultant friend’s guidance, I could shop for the lowest price rather than the most technical help from the dealer. And Michael steered me away from the dot-matrix printer, which, with their inferior print quality at the time, would have been bad news for me. I was lucky. I had befriended Michael through an Osborne user group at a time when he himself was looking for advice on writing, so he didn’t even charge me. You may be similarly lucky if you have a skill a computer consultant needs. But don’t count on it. So often, free consultants are worth their charges. Had I not been as fortunate and had I experienced difficulties, I would have been willing to pay for advice. And mind you, my whole system costs less than $3,000. “You don’t want to spend $10,000 on a computer system,” Michael warns, “and find it doesn’t do what you want. You’d be better off spending several hundred dollars or more to learn exactly what you want and get some help installing it. Think of the business you’ll lose if you install your computer and it doesn’t work as planned. It depends. You could also spend a lot of time shopping and talking to people and not pay a consultant. It depends what your time is worth. The average guy has one or two uses for his computer in mind, and he doesn’t want to piddle around with the complexities of the thing. He just wants to cut right down to the core of it and get this job moving. If a consultant spends two or three hours with someone discussing a micro, it may cost perhaps $75 or $100. But the client might master in a day what might normally take a week or even several.” The wrong consultant, however, may actually cost you more—in money and time. When shopping around for the right one, you’ll first want to consider a _what_—what work you need him for. Forget about computers. Does the task make sense commercially? Are you realistic about its costs? A former New England consultant tells of a baseball enthusiast who wanted to recreate some board games on the computer screen. It might have worked. But the man didn’t just want a computerized arcade game here—rather, a replication of ball parks, with their actual proportions. “He had a statistical background,” recalled the ex-consultant, “and had done some statistical consulting for a professional sports organization.” But there was a problem. The man believed that the consultant could perform this miracle on an Apple for less than $5,000. Actually, the program would have cost $50,000 to write—probably more than the man could quickly recoup through sales of the program. The consultant turned down this mission impossible. But the client let his enthusiasm for the project blot out his business sense and in fact squandered $55,000 on another consultant, who, incidentally, botched the job. Now move on to the details of the work you want done. Do as with software; get as clear an idea as you can of the paperwork you’re computerizing. You may want your electronic files organized entirely differently from your paper ones. Whatever the case, plan ahead. You don’t want to—as somebody once put it—automate your confusion. And there’s another reason for knowing your precise needs before calling a consultant. Trying to jack up the bill, some experts can be brilliant at inventing new problems to fit their prepackaged solutions. Or they may downplay your real needs because the answers are not in their tool kits. You’ll also want to decide if the consultant is just to give advice or more. Don’t pay simply for advice, then expect him to haggle with equipment suppliers and install the software. And what about other services besides advice? Take the task of keying paper records into your computer system; some consulting firms will line up temporaries for the job. That way your normal people must worry only about their regular work and learning the new computerized routine. Of course, this =data-entry= work may itself be part of the training, and at least if your people enter the material, they’ll more easily find it later on. As you define the _what_, you also get into a _when_—your deadline for computerization and how long the consultant or the firm is to remain on the job. There is a happy medium. You want him around long enough, full time or part time, to get the system up and running. But don’t count on making him in effect an employee. Hire permanent people if that’s going to happen. Joseph Auer, a $1,500-a-day computer-negotiations expert with blue-chip clients like New York Life, tells what can happen when consultants stay too long. A bank used a consulting firm for years. The outsiders became the main people familiar with the electronic links between the automatic teller machines and the big computer. Then the bank woke up. It realized that the outsiders could cripple its operation. The consultants didn’t blackmail the bank, but it was worried enough not to risk incurring their displeasure. Rather than farming out the work to a facilities management firm as planned, it extended the existing consultants’ contract, with the understanding that the outsiders would make the bank’s own people self-sufficient. Next go on to the _who_ question. A generalist who’s a good, quick learner might be a better deal for you than an overpriced guru who limits himself to one area. It’s like medicine. I’d much rather be cut up by a smart general surgeon with a steady hand than by a mediocre cancer specialist. Of course, for a complex job, you want the equivalent of a top cancer surgeon. At Stewart, anyway, the right specialist might have “programmed” the company to succeed. Intertwined with the _who_ are two _how muches_—the size of your company and the task for which you need the consultant. Are you a small businessman or a Fortune 500 executive? The consulting needs of the two categories may overlap. The general rule, however, is that a large company will feel more comfortable with a large consulting firm (GM, say, with Arthur D. Little) and a small company will be better off with a small one. A small businessman may enjoy more personal attention from a small consultant. He knows the people he’s dealing with, and it’s harder for the firm to pass the work on to a green employee—while charging the full price. The smallness of the firm, of course, makes it that much more important to check out references. Of course, sometimes the small guy may fit the large company’s bill; Michael Canyes, for instance, has tutored WordStar classes for the Federal Trade Commission—a good buy for the taxpayers, considering his skill and low overhead. “The smaller consulting firms,” he wisely says, “are likely to save a large business money when the consultants are working within their limitations.” Canyes, however, at the time I wrote this chapter, wasn’t the consultant to install a $5 million micro network. He might be one day, but back then he just didn’t have the resources to do the job justice. A Big Eight accounting firm would have been better than Michael; in fact, an executive with a large company mightn’t need an outside consultant of any kind—large or small. His or her employer may boast a large microcomputer center and a helpful data-processing department. But you never know. Sometimes, Data Processing is either ignorant of micros or discourages their use. Ask some test questions. Has anyone in Data Processing ever installed a micro for business purposes? Do many in the department have micros at home? Are they reading the micro magazines and newspapers—_InfoWorld_, say, rather than just magazines about the big machines? If Data Processing isn’t comfortable with micros, you’d do well turning to the outside. One of the qualifications for the outsider, of course, might be his ability to work harmoniously with Data Processing if you’re hoping to hook into the mainframe. Then again, your micro might be a stand-alone, not exchanging information with other computers, so that you needn’t worry what Data Processing thinks. Even if you’re using a Data-Processing staffer or other internal person, you still may want to find out his track record. Don’t be a corporate chauvinist. Every company has its share of gobblers. Turning to an outside consultant? Then, through computer stores, you might try to track down a local users group for a brand of hardware or software. Go to a meeting. Find out whom the members regard as an ace consultant. In fact, try to go to several user groups—remember the preference that Apple owners, say, may have in favor of an Apple solution or that Kaypro owners may have in favor of a Kaypro one. Mere word of mouth, though, is still no substitute for a thorough interview with the consultant and reference checks. Certainly, too, you can’t use a consultant just because he or she shows up in the directory of a consultants group. Charles E. Harris, a lawyer who, with Joseph Auer, wrote _Computer Contract Negotiations_, says, “I know of no organizational membership that proves anything other than membership and the fact that the man’s paid his dues.” “The person to watch out for in particular is the guy who’s doing this part time and he works on a mainframe or something like that,” warns Mark Robinson, a California consultant. He’s been in the micro business for years and handles dealer relations for Lexisoft, a software company—a micro one, mind you. “Anybody who hasn’t made this his business full time for a significant period of time,” he said of micro consulting, “is a great risk.” Consultants can relax. Robinson said eighteen months could be “significant.” I myself wouldn’t shrug off the part timers from Data-Processing departments if they have a long list of pleased micro clients. But as a generality Robinson’s observations seem valid. Not only is micro software written differently from the mainframe kind; there’s also a different mentality. Cloistered away in the computer rooms, some mainframers may strive for technical perfection at the expense of economy. “They may be right about things,” Robinson says, “but that’s not the point. The point may be more like what Mr. Osborne prates about all the time—mere adequacy. I’m thinking of posting an aphorism on the wall from _The Soul of a New Machine_: not everything worth doing is worth doing well.” The same wisdom might apply to consultants. You don’t need a $1,500-a-day man to help you choose a good micro for word processing—not unless you’re about to buy a few hundred. Opinions vary about pay. Some observers of consultants say you get what you pay for. “There’s very little relationship between pay and value except for one thing,” Robinson says. “People who don’t charge enough eventually go out of business, or that business. If they don’t charge enough, they’re not likely to have been around long enough to have gotten experience for you.” Then again, as Robinson points out, experience in the microcomputer business rapidly decays in value, weakening the relationship between pay and competence. Back in the late 1970s Bernard McGowan, a Massachusetts ophthalmologist, bought an Apple and hired an expert at all of $5 an hour. His consultant was a scruffy young man in an old jacket. And this guru had himself owned an Apple only a day. But it didn’t matter. His name was Mitch Kapor, and he was good, apparently, for he went on to found Lotus Development Corporation, the company that developed the best-selling 1-2-3 integrated program. Today McGowan is still using $5-an-hour men at times. He told me, in fact, that one consultant, a teenager digesting software manuals for him, doubled as a baby-sitter. No, this isn’t an argument for merciless exploitation of child labor à la nineteenth-century England. For a small businessman with a simple task, however, and no pressing deadlines, a teenager might do well. Why must every sixteen-year-old be flipping hamburgers around after school at McDonald’s? But teenage computer geniuses are special cases. Don’t count on a $5-an-hour consultant to write up a complex accounting program that you need done within three months to avoid bankruptcy. Moreover, if you’re a business of any size, you might end up spending hundreds or even thousands of dollars just to audition consultants. That’s right; you may pay them. Canyes will talk to people by phone if he thinks they’re good prospects, but the moment he gets in the car, the meter starts running. That’s basically in line with what Adam Green, a software training expert, told me. Green has conducted seminars for big-name clients like Price Waterhouse and Xerox, and earlier he worked as a programmer and salesman, so he’s been around. “A good consultant in my view would come out [in many cases] and spend a day or two looking at the operation and then give back a written proposal saying how they could help the client,” Green said. “And then they should charge the client about $250 for that.” “If someone comes out and spends a day or two with you and writes something up for free, obviously they don’t have anything better to do. They’re not working. They’re amateurs. They don’t consider their time worth anything.” Paying a consultant to write a proposal, you’ll learn (1) if he can communicate in comprehensible English, (2) if he can think logically, and (3) if he understands your business and its needs. But maybe that’s a step or so ahead of where you now are in the game. Having found a promising candidate, make certain you feel comfortable with him or her. Trust your instincts. Beware of snake-oil artists who, through greed or incompetence, may actually increase your time and expense. People like Bob Hillard and Michael Canyes come across as teacherlike in their eagerness to pass on their wisdom to the appreciative. But not all consultants, in and out of computers, seem that way. A few might not even give you a written report of their findings. Many intimidate you through jargon, even through dress. Quite correctly, some young business school hotshots have been depicted as punks in pin-striped suits. “You have to know what your client wears and dress one level above him,” said an ex-Bain and Company man. “If he wears a sport coat, you wear a suit. If you’re meeting him in a midwestern cornfield and he’s wearing a T-shirt, you wear a button-down. You might actually drive to his office in beige pants, a jacket, and a tie. But you can look in through the window when you get out of your car and then take off as much as you have to.”[31] The consultant wasn’t in the computer trade, but he might as well have been. One industry magazine began a “So You Want to be a Consultant” article with a warning against polyester. Footnote 31: The ex-Bain man’s quote is from _Harper’s_, November 1982. “Well,” I asked Mark Robinson, “what about dress? Can you trust a consultant in a three-piece suit more than one who looks like a refugee from a university computer center?” “It’s really irrelevant,” he said, “I’d look more at things like Can the guy focus on the relevant features of my business? Does he seem to understand what my business is about? This is one of the things that probably isn’t even related to the number of years in micros or computers or what not.” Robinson suggests asking yourself, “Does he quickly grasp the nature of my relationships with my suppliers, my clients, my promotion methods? Is he focusing on the technical solutions rather than my needs first? Does he even consider a question like whether I need a computer at all? Does he try to look at the highest-priority needs that I have in terms of the effect on my bottom line? Does he really have a grasp on why I might have a computer—to move into new things and serve my customers better, to replace existing functions, and so on? Does he know general things like that? If he doesn’t have that sort of perspective, if he isn’t even able to back off as to whether I should have a computer, I may be getting a slightly different type of salesman, someone who is committed to a computer solution, for instance, and isn’t really looking at it from my viewpoint. He is not totally professional.” You might still use such a person, but you’d better consider his limitations and biases. I asked about “pure” consultants versus those selling software or hardware. “Most consultants,” said Robinson, “have found that small businessmen’s budgets for pure consulting and their appreciation of it aren’t high enough so they make enough money. So they’ve got to sell software and hardware. And also purity means that a consultant is spread too thin. He’s not able to specialize and know enough about a particular package to be competent in it. So either he has a superficial view of these packages, or else he’s a partisan.” While working as a full-time consultant, Robinson hoped to maintain his purity. “I found myself dealing with a lot of invoices of fairly small amounts,” he said, “and most of my support came from contract programming.” What with the line so thin between salesman and consultant, you should politely but persistently ask about business connections that could sway their judgment. Obviously, too, you’ll want a consultant’s résumé if he or she has one. He should. One way or another you’re after the normal personnel-office answers—his schooling (anything that could help him understand your business?), his previous jobs (one computer consultant ran a nightclub once—perhaps not such a bad background for an expert dealing with small businesses), maybe even if he has an arrest record (a perfectly suitable question if he’s setting up an accounting system or doing security consulting). Be polite but aggressive. By not shying away from the basics, you’ll be reminding him who’s in charge—the client. Don’t forget the _why_. Why is he in “consulting”? It isn’t a euphemism for unemployment, presumably. Also, why is he working with micros rather than mainframes? Attitude matters as much as brains and skill. Ideally, your consultant will be comfortable with the flora and fauna outside the antiseptic computer rooms. While he’s auditioning, see if that’s true of him. How do you tell? Introduce him to your people, not just your executives. Encourage them to ask questions. If the consultant condescends toward a secretary or clerk now, he may very well do so toward you later. And if he rubs your people the wrong way, they may resent him in the future and take it out on the computerization project. Well, there’s a caveat here. Maybe the specs for the job are so clear-cut that your consultant can work in a little corner away from the rest of the world. With _personal_ computing, that’s not going to happen as much as with mainframes. In winding up the interview, ask about the future. Will the consultant be around in the future to do work or answer questions? He won’t be? You still might use him. But police him very carefully to make certain he’s following standardized procedures that another consultant can pick up on. You don’t want your business to go down the tubes when your software man goes off to find his karma atop Mt. Fuji. Satisfied? Then move on to references—which, if you’re a small, budget-minded business, you might want checked before a full-fledged audition. Remember Stewart’s ordeal. Ask for the names of people whom the consultant served in tasks similar to the proposed one. Request ten names. The consultant needn’t have done—in each case—jobs like the proposed one. Moreover, you often don’t have to call more than five or six references. But with ten names, at least you’ll know you’re far from the first client. Don’t, incidentally, phone just the references at the top of the list; outfox the consultant in your most constructive way. You’ll want the references to tell you the usual—what the consultant achieved within the time agreed. Would they use him or her again? What kind of hardware, what software, did the consultant recommend? The same kind for clients with different needs? Be on guard if he seems to have a one-shoe-fits-all philosophy. Also, ask about the references themselves. See if they have any connections with the consultant, social or business. How many consultants should you try out before signing the contract? That’s up to your needs and your budget. There’s no law saying you have to go through all the screening steps here in every case. Why pay $250 to a consultant to audition when that very likely will be the cost of the whole job (assuming you’re lucky enough to find one willing to take on a project with such a low budget)? For all tasks, though, even small ones, you’d do well to bother with at least one formality—a contract with your micro consultant. ■ ■ ■ A Sample Consultant’s Contract (Featuring Some Plain English From a Washington Lawyer) No, you’re not going to pay a lawyer $50 to draw up a contract for a consultant doing a $60 modification of WordStar. But you can help protect yourself with your own version of the sample contract below. It’s nothing more than a letter in simple English. William Wewer, however, the Washington lawyer who drafted it, says it probably would be just as binding legally as a document filled with “whereases.” _Dear_ ——: _This will confirm our agreement that you will perform the following_ [software?] _services_: [Precisely list the services in numbered paragraphs.] _You have represented to me that you can complete this project by_ [date]. _You will charge me $—— per hour for your consulting time. Do not exceed $—— without my written consent._ _You agree that you are performing these services as an independent consultant and are not my employee._ _If you are willing to work under the terms of this letter agreement, please sign the enclosed copy and return it by_ [date]. Under your signature, use this language: _I agree to and accept the terms of this agreement to_ [quick description of service provided]. ———- [consultant’s name], ———- [date of signing] Remember, this sample is just a starter. See Backup VIII, “Consultant Contracts: Some Who-How Questions”, for a list of some of the points a contract might address. One of the more important ones is ownership and control of software. If you want to own the newly developed programs and keep them away from your competitors, here’s what Wewer might insert after the paragraph saying the consultant isn’t your employee: I am paying you to do this work for me alone and require that you not retain any copies of it or give any copies to anyone else or tell anyone else how to recreate what you have done for me. We agree that the software you create for me will be my trade secret. “‘Trade secret’ is a key phrase to get in there,” Wewer says. “Then it’s within common law.” Again, your informal contract won’t give you iron-clad protection—that could depend on the mood of a judge—but it’s better than nothing. “It’s the handshake deals,” Wewer says, “that always fall apart.” ■ ■ ■ “The real problem,” Adam Green says, “is that no one ever has to sign a contract. People get concerned. They say, ‘Boy, lawyers cost a lot,’ and I know the feeling.” Then Green brings up an exception to that opinion, lawyers. “I know a lot of consultants who will never work for lawyers because as far as lawyers are concerned the job is never complete. They understand what it means to be complete.” Their contracts reflect this; yours should, too. Have the consultant do the first draft of the contract, perhaps; but don’t shy away from extensive changes if need be. “Hell,” tell him, “that’s what your word processor’s for.” Back out of the deal if he isn’t reasonable now. He might not be later on. When hammering out a consulting contract, you should fret over three variables emphasized by Joe Auer: