LOOK FOR THE K MART ON YOUR right. Then, just ahead on the left, is a Denny's. Turn there, and we're behind the parking lot."

Communications Solutions Inc. (CSI) has lived its whole young life in a squatty little building behind a San Jose, Calif., fastfood outlet. It embarrasses chief executive officer Skip Vaccarello, Harvard '71 and Boston University business school '77, to give directions.

On the other hand, squatty or not, a home to call its own has been a comfort to CSI these past few years. The company is with its fourth set of owners, three of them corporate parents. If employees couldn't know who would be buying them next, they at least knew where they'd be working. "Staying here helped reinforce the attitude that we were our own company," says technical sales support John Seal.

Remarkably, that seems to be true. CSI, despite what may be a record for a company of so tender an age -- four owners in eight years -- is close-knit and, management claims, profitable. Sales, which the company says have grown nearly 40% annually over the past five years, hit almost $10 million in 1986. CSI is a leading software supplier in its market, according to Dataquest Inc. director Linda O'Keeffe. "Their shining strength," she says, "has been the quality of their product."

Walk around the place and talk to people; you'll get no hint that CSI is some other company's subsidiary.

Other high-tech companies would surely love to have even twice CSI's employee turnover rate. Just 2 out of 24 software engineers, a notoriously fickle lot, have left voluntarily in the past four years, and one of those asked to return.

Four owners? The place ought to be full of neurotics, suspicious of anyone with a strange face and a three-piece suit. But it isn't, which suggests that Vaccarello and his chief lieutenant, engineering vice-president Gene Buechele, might know something about keeping a company whole while it's being shopped around.

They do. Their first lesson, though, came from the other side of an acquisition deal. They worked for CSI's first acquiror at the time of the acquisition.

Tiny CSI, with about a dozen people in all in 1983, was in two closely related businesses -- consulting and software. It had started life as a consultancy, helping large computer users and manufacturers of computers and peripherals learn how to get their machines to talk intelligibly to IBM mainframes. Eventually, CSI augmented its advice with a newsletter and seminars. That was the information section of the company.

Then, in 1982, CSI began writing software that would permit non-IBM computers to connect to IBM machines. That and other programs for large end-users made up the software section of the business.

In 1983, CSI's two technically oriented cofounders sold out to VisiCorp, the Icarus of the early personal computer software companies. VisiCorp had taken off with VisiCalc, the first popular spreadsheet program for PCs. Shortly after the CSI acquisition, but unrelated to it, VisiCorp began to stall out on delivering follow-up products. Within months of acquiring CSI, VisiCorp started a nosedive into bankruptcy. CSI, however, did not.

The misjudgment of each company by the other was amazing. When VisiCorp bought CSI, according to Buechele, it thought it was getting sophisticated software that it could sell along with other end-user products. For its part, CSI thought it was getting a distribution system and a capital source. Had both assumptions been true, CSI likely would have disappeared within VisiCorp.

But neither company had looked closely enough at the other. CSI's software hadn't progressed nearly to the idiot-proof stage it needed to reach for end-user distribution, says Buechele, and therefore was of little use to VisiCorp, even if VisiCorp had been ready to adopt it. VisiCorp, on the other hand, couldn't help CSI market to non-IBM computer makers, and, anyway, the acquiring company soon proved not to be the stable, cash-rich parent CSI thought it was getting. So, by and large, parent and subsidiary left one another alone -- except in one respect that aided CSI immensely.

CSI acquired its top executives and some of its best programmers from VisiCorp. First Vaccarello, and later Buechele, an early Intel Corp. alumnus, moved over permanently to relieve CSI's founders of corporate and engineering management duties. It's what the founders wanted, and it's the only part of the acquisition that worked as planned. Later, as VisiCorp crashed, other engineers joined the subsidiary.

As acquirors in this case, Vaccarello and Buechele learned from the experience that they have used subsequently as acquirees. The first, and less surprising, is: plan ahead; know what you want; know what you're getting. More on that in a minute.

The second, and less obvious, is: synergy and autonomy probably don't mix. A buzzword in acquisition talk, synergy is supposed to mean that the acquiree and acquiror together can do something that neither could do alone. In practical terms, Vaccarello and Buechele learned that synergy means the acquiror is interested mostly in the acquiree's product. Sooner or later, the acquired company runs the risk of becoming an in-house supplier to its new parent. Autonomy, on the other hand, comes only to companies that are bought and valued for their earnings, or earning potential, alone.

Only the fact that its end-user program wasn't ready to go to market kept CSI from being absorbed into VisiCorp as just another division. Buechele knew that, because it was the end-user program that he'd wanted when VisiCorp was buying CSI. So, during the next two acquisitions, when Vaccarello and Buechele were on the acquiree side of the deals, they knew precisely what they didn't want in a parent: synergy.

VisiCorp's growing need for cash and CSI's growing fear of failure by association soon suggested that a sale might be good for both companies. By this time, Vaccarello, at 39 the younger of the two, says he and Buechele had figured out what they did want in a parent: a substantial, financially stable company that was interested in CSI strictly as an investment. "We sat down," says Buechele, "and made a list of what the company wanted, what individuals wanted, and who the people were that we would and would not do business with. I had seen the founders fail to do that with VisiCorp, and it had given me an advantage because I knew exactly what I wanted from that deal."

Vaccarello and Buechele may have known what they wanted, but in subsequent negotiations they would be neither the buyer nor the seller, just the property being bought and sold. So, you might wonder, who were they to be making choices and dictating terms?

Ironically, their company's very fragility gave them the bargaining strength they needed to be part of the negotiations and to influence their outcome. CSI's only assets were people -- who could walk out the door at any time. "That's the company," says senior communications engineer Gary Jaszewski, "the collective knowledge of everybody who works here -- not the software." So VisiCorp and potential buyers understood that with rough handling, CSI could easily break apart, leaving little of value. That realization got CSI listened to when VisiCorp set out to find a buyer, and in the subsequent negotiations with Minne-apolis-based Control Data Corp., CSI got a seat at the table. "VisiCorp didn't force us to look at companies that we didn't want to look at," Vaccarello says.

VisiCorp began shopping CSI around among potential buyers in 1984. The lesson CSI took away from this experience, and the three-way bargaining that followed, proved useful later on when Control Data sold the company to Altos Computer Systems. It was: know whose side you're on.

Control Data, which had pioneered in making supercomputers, the most powerful and costly computers on the market, was broadening its interests. It wanted, among other things, to assemble a portfolio of complementary software companies, one of which would be CSI.

Understandably, Vaccarello was skeptical of the prospective new parents, first Control Data and later Altos. "We went through a painstaking process of trying to figure out what the real motives of the acquiring companies were," he says. "We talked to people at different levels in Control Data and Altos and to companies they had acquired."

Once the motives were identified, however, and negotiations got underway, Vaccarello figured out whose side he needed to be on. Although it might appear to be a conflict of interest, he began to identify with Control Data, the acquiror. "It became apparent," he says, "that CSI's interests were closer to Control Data's than to VisiCorp's. We had to live with these people. We didn't want them paying top price."

In the negotiation process, specific issues pushed Vaccarello into the buyer's camp even before he'd struck his tent in the seller's. It was an asset sale, so CSI's bad receivables, for instance, became a point on which people could disagree. "The fight was over how much time do we allow before our bad receivables became a Control Data problem instead of a VisiCorp problem," recalls Vaccarello. He sided with Control Data to stretch the period. Another touchy item was the bonus CSI people were to receive if they stayed with the company for a year after the acquisition. "The size of that bonus became a point of contention," says Vaccarello, "because it would come out of VisiCorp's pocket." In issues like that, old loyalties gave way quickly.

"For the acquiree," adds Buechele, "it's easy to do nothing. . . . The tendency of both managements [VisiCorp and later Control Data] was to tell CSI, 'You don't have to know about all of this. We'll tell you when you need to know.' You can see why [they'd like to have it that way]. The hardest negotiations in the world are three-way negotiations."

Being part of the high-level wheeling and dealing when you're being bought and sold, as Vaccarello points out, is crucial. But it's not the only thing going on. There's still a company to run and employees to be dealt with. Not only do they need continued direction, they need reassurance. What the wheelers and dealers are doing affects them profoundly, only unlike the negotiators, they don't have the luxury of control. "It's the difference between driving and riding in a car on a twisting mountain road," salesman Rick Swan says. "It isn't the driver but the passenger who gets sick."

But Swan goes on to suggest that Vaccarello took it easy on the curves and, even when he couldn't tell employees exactly where they were going, kept telling them things were going fine.

First, he and Buechele split up. Vaccarello handled the negotiations; Buechele stepped over to run the company and see to the day-to-day comfort of the troops.

Vaccarello would hold meetings to tell everyone in the company what was going on. When he couldn't mention names, he didn't. But names were less important than assurances -- that the company would stay independent, that salaries and benefits wouldn't be cut, that people could go on doing the same kind of work.

Since he had always shared full financial results -- orders, sales, profits -- with employees anyway, it didn't set new precedents for Vaccarello to drive home CSI's continued profitability and growth. "At meetings," says CSI's newsletter writer Howard Bernstein, "one of the last things Skip does is to ask, 'What do you want to know? Got any questions?' Any apprehension was nipped in the bud."

"Skip did a great job of isolating people from details that we didn't need to know," says Swan. "He wouldn't talk about which company, but he'd talk about the type of parent we were looking for, one that wanted a good return on investment but not our technology. He'd keep reminding people of this. We humans need some comforting from time to time."

"He told me enough to keep me from spreading false rumors," says engineer Gary Jaszewski. And as John Seal points out, after the first acquisition, Vaccarello had a lot more credibility with the troops on the second. "Management had assured us," he says, "that we wouldn't be affected, and the more acquisitions there are, the more they can say, 'See, we told you."

Buechele was the day-to-day morale builder. "I spent a lot of money during that period," he says, "taking people to lunch . . . reminding them of the good things."

And demonstrating continuity. VisiCorp might be going down the drain, but CSI was profitable. So Buechele kept buying the equipment his people needed. When they ran short of space, negotiations didn't stop Buechele from renting more space in their homey little building. While VisiCorp handed out pink slips, CSI handed out profit-sharing checks. We just kept on growing," says Buechele, "and engineers equate growth with opportunity."

Between the two of them, Vaccarello and Buechele tried to be sure that their precious human assets found no good reason to jump out while the ride was still on.

Can an acquired company really remain independent of its acquiror? Not likely -- at least not if the acquiror is intent on seizing control. Yet CSI is on its fourth owner, and people who have been there since the VisiCorp days testify to the continuity of style, purpose, and culture inside the company. How has CSI managed to maintain that level of independence?

Vaccarello and Buechele have had the good sense to keep the company focused on serving its well-defined market with what customers and others claim are quality products, well delivered. "As far as I'm concerned," says Frank Dzubeck of Communications Network Architects Inc., a Washington, D.C.-based consulting firm that recommends CSI software to its clients, "they're at the top of the list."

Thus, no acquiror has been tempted to move quickly to redirect the company's efforts or replace its management. No one has seen CSI as a makeover opportunity. It's more like a company with a strong outside board. "They put a plan together for our approval," says Dave Zacarias, president of current parent Altos, "and as long as they are meeting those objectives, then we'll stay out of it. We really get involved only at the board level."

A smart acquiree will discover what the parent wants, then deliver it. "The first thing to do," says Buechele, once you've been acquired, "is figure out how you're going to be measured. Then, optimize your performance toward those measures. For instance, VisiCorp had its own problems, and they didn't want to hear about ours. They were just looking for the highest cash balance. . . . So, we only bothered people at the lowest possible level at VisiCorp, and we maximized our cash.

"At Control Data, it was different. They had procedure books three inches thick. The secret of survival there was to show a detailed balance sheet in their rigorous format. They didn't care if you were showing a profit so long as you met the reporting format. . . . We spent a lot of time learning who wanted to know what when."

Small things count, too. The immediate threats to CSI's independence don't come at board meetings. They're more insidious. The big organization doesn't eat you whole, it nibbles at you. "Everybody in the organization wants to help," says Buechele, "and they'll help you to death."

For instance, soon after Control Data bought it, CSI wanted to rent still more space in its building behind Denny's. We'll help you negotiate the lease, said the Control Data facilities department. No thanks, said CSI. We can negotiate our own lease. But you can't sign it, said the Control Data bureaucrat. Only three people in the company can sign leases. CSI appealed to a Control Data VP, who called the functionary off. CSI signed its own lease.

When CSI installed a new Coke machine, Control Data engineers insisted it had to meet their installation specs. Furthermore, said the computer company's personnel department, you can't give Cokes and coffee away. That's not consistent with our policy. CSI insisted it would install the machine as it liked, where it liked, and charge what it liked. Cokes are still free.

"It sounds silly," says Buechele, "but those were big problems. . . . It's mostly service organizations within the parent corporation that are trying to carve out their fiefdoms, so they want to get their fingers into your business. . . ." The corporate public-relations people urged CSI to use corporate PR. The accounting department suggested CSI use the corporate accounting system. CSI resisted both. The secret, says Buechele, "is not letting the personnel department fill out that first form for you, because when it comes to the second form, they'll say, 'Well, we did the last one. Why not this one?' This one," Buechele says, "may be your own pay raise."

When Control Data issued employee memos, Vaccarello screened them first. Did CSI employees need to see them? If so, he sent them out under a CSI cover letter. When Control Data distributed the CEO's annual videotaped message to employees, Vaccarello got it but never showed it to CSI employees. "We rarely mentioned Control Data at meetings," he says. "On the letter-head, in very small print, it said, 'A Control Data company.' The only people interacting with Control Data were the controller and I, and I deliberately kept it that way."

In the end, though, luck helps you stay independent, too. It was lucky for CSI that Control Data ran into profitability problems when it did. "We lost $567 million in 1985," says a company spokesman, "because we were trying to do too many things at once." Selling CSI, he says, was part of Control Data's "restructuring."

But it also saved the CSI name. Despite everything that CSI was doing right, Control Data had its own long-term agenda. Ultimately, a parent company willing to risk changing a successful acquiree will have its way. In the future, the Control Data spokesman says, "CSI would have become part of our operations . . . writing software for our internal business."

Zacarias claims that won't happen at Altos, as long as CSI continues to perform well. "We don't believe we understand that piece of the business well enough to bring CSI in-house." Altos, he says, would rather keep CSI's people motivated and happy by leaving them alone.

It's been just a few months, but what Zacarias says seems to be the case. "Altos," testifies salesman Roger Fisher, "is just a shareholder. You don't see any Altos managers around here, do you?"

"We still have the small-company atmosphere," insists engineer Jaszewski. "I know Skip and Gene, and I can talk to them about company policy any day I want."

Seal remembers a joke that made the rounds at CSI during the last acquisition. "We brought down VisiCorp," it went, "and Control Data didn't thrive under our policies. Maybe, if we get acquired by IBM, we can cut them down to size, too."


In May, CSI left Denny's parking lot for a larger building at a classier address. But it's still CSI's own, CEO Vaccarello says. The company had planned the move long before the Altos acquisition.