(FORTUNE Magazine) – No company typifies the new world of M&A better than Cisco Systems. To find a business that has handled acquisitions as well, you'd have to go back to the AT&T of the early part of the century, when Theodore Vail, its legendary CEO, bought hundreds of tiny phone companies and assembled the first nationwide network, giving birth to Ma Bell.

It's fitting that Cisco, the $12.2-billion-a-year giant that provides the hardware and software behind state-of-the-art Internet networks, now wears the mantle of M&A king. To understand how it has built its empire, forget all you know about corporate raiders and their swashbuckling tactics. (That's so Dallas circa 1984.) Disregard the stereotypes of ruthless capitalist villains aiming to gobble up corporate America's finest. (That's so Wall Street.) And ignore everything that's happening on the services side of the telecommunications industry. (It's way too early to tell whether WorldCom can successfully swallow MCI and Sprint.)

Think of Cisco as an acquisition engine, as cleverly designed and highly tuned as the giant routers it builds to handle vast surges of Internet traffic. Like those routers, the acquisition engine runs on Internet time. In the past six years, Cisco has spent $18.8 billion on 42 acquisitions (as of the date this story went to press). Armed with a killer currency (Cisco stock has risen 162% in the past year and is now trading at $72) and a team of well-practiced execs, the company prowls Silicon Valley--and the world--snapping up companies to expand existing product lines or support entirely new initiatives.

The action this August was typical: Cisco absorbed two startups (Monterey Networks and Cerent), closed two other acquisitions, and negotiated two more. Cisco moves at breakneck speed, in some cases sealing a deal after 24 hours of negotiation and closing it in a month or two. "I haven't found anybody in any industry that has a process that is as tuned," says Jon Bayless, a general partner at Sevin Rosen Funds, a Dallas venture capital firm that helped negotiate the Monterey acquisition.

Even by strict quantitative measures, Cisco's strategy has been wildly successful. While absorbing ten companies in its past fiscal year, the company saw sales grow 44% and profits 55%. When Cisco absorbs a company, it makes a no-layoffs pledge; its turnover rate for employees acquired through mergers is a scant 2.1%, vs. an industry average of more than 20%. Its first acquisition, Crescendo, which Cisco bought in 1993 for $100 million, is the foundation of a unit that generates about $4 billion in revenue annually. It is still run by Crescendo's CEO, Mario Mazzola.

Cisco's process does not depend on the idiosyncrasies of the Internet business. In fact, its techniques are a great how-to manual even for managers considering a single acquisition in an old-line industrial field. In a series of interviews with FORTUNE in September and October, Cisco's merger mavens explained it all.

"Cisco's strategy can be boiled down to five things. We look at a company's vision; its short-term success with customers; its long-term strategy; the chemistry of the people with ours; and its geographic proximity." --MIKE VOLPI

The man who runs the acquisition engine is a soft-spoken 33-year-old wunderkind named Mike Volpi, who bears a passing resemblance to actor John Cusack. CEO John Chambers signs off on all the deals but gets involved with only the largest and most intricate. "We're spending $10 billion a year in acquisitions; this is a process that works," he says.

Chambers' trust is well placed: Volpi is arguably Silicon Valley's shrewdest shopper. He needs to be. With a dearth of engineering talent and shrinking product cycles, Cisco realized that it couldn't build everything it needed inside the company. Says Don Listwin, Cisco's No. 2 executive: "Lucent wants the smartest group of people in Bell Labs. But if we're not good at something, we've got Silicon Valley. It's our lab."

The company looks to startups if it decides it's too far behind competitors to take the time to build a product from scratch. Once that decision is made, Volpi's team consults with business units and customers to find out about their technological needs. Customers have a profound influence on Cisco's strategy. Its executives say that US West, the Denver Baby Bell, was a driving factor behind the decision in March 1998 to acquire NetSpeed, a maker of equipment that turns regular phone lines into high-speed digital subscriber line (DSL) data conduits.

With a few targets in mind, Volpi and his team perform an evaluation that's unique in scope. As engineers examine the technology and financiers go over the company's books, Cisco's team also examines the depth of the company's talent, the quality of its management, and its venture funding--all things aimed at making the integration process easier should Cisco buy. Since Cisco will acquire a company as often for its talent as for its technology, it needs to focus on these "softer" issues early.

There's one more test. Volpi's got a knack for identifying a startup at the sweet spot of its development: when it is old enough to have a finished and tested product, yet young enough to be privately held and flexible in its ways. Acquire a business that's too mature, and risk soars. Explains Volpi: "If you buy a company with customers, product flows, and entrenched enterprise resource systems, you have to move very gingerly. Otherwise, you risk customer dissatisfaction. Figuring out how to integrate this type of company could take nine months."

After all this evaluation, Cisco doesn't necessarily leap to purchase. Instead it may pass on a company or may act as a venture firm, making a 10% equity investment to monitor the startup's growth. That's what it did with Monterey last winter. A few months later, after deciding that the Richardson, Texas, startup would give it entree into the $20-billion-a-year optical-internetworking market, it gobbled up the rest. Says Joe Bass, CEO of Monterey: "We had interest from other companies, but they didn't move as fast as Cisco. They were still considering us when the announcement came out that Cisco had bought us."

"The pain if you don't pick the winner is enormous, so we'll bet on the company that has momentum." --DAN SCHEINMAN

Dan Scheinman is Cisco's "culture cop," evaluating possible acquisitions to see how well the companies would fit into Cisco. However, the gregarious, basketball-playing 36-year-old vice president for legal affairs comes off as anything but an enforcer. He's a reporter's dream interview--an articulate spokesman who spouts pithy comments, likes to tell colorful anecdotes, and marks important points with inflection. Here are his rules of thumb for okaying a deal for Cisco:

--Look for a bad deal. Any startup should have one glaring mistake it's learned from. If the company hasn't done a bad deal, it's not daring enough. If it's done too many, it's, well, stupid.

--Role play. Go over the decisions that management made, and see if you'd come to the same conclusion. If so, the company's execs probably think the way you do and are likely to fit in fine.

--Move fast. There's no need to hang up on little issues that don't matter. The reason Cisco was able to beat IBM to Kalpana, its third acquisition, was that the Armonk, N.Y., behemoth was too busy performing groundwater tests designed to check whether Kalpana's headquarters was up to code. Cisco swooped in and cut a deal in a weekend. "I guess in that one we took some groundwater risk--or maybe we didn't. We had the report," he says. Those who negotiate with Cisco appreciate its speediness. "If Cisco wants something, it won't nickel and dime you," says Tom Dyal, now a partner at Redpoint Ventures, a venture capital firm in Menlo Park, Calif. He worked on Cisco's acquisition of StratumOne Communications this summer.

--Observe what's going on in the negotiations. "We've been involved in deals where people start negotiating for themselves," Scheinman says. "There was one deal where the engineering and business teams of a startup didn't show up at the same time, and each negotiated about how to screw the other guys."

--Don't be afraid to pull the plug at the last minute. The night before Cisco was going to buy Chipcom, a maker of Internet switching systems, Scheinman says, he couldn't sleep. "The board asked me if I thought we were doing the right thing, and I said no." Chipcom ended up being acquired by 3Com.

--Finally, forget the fanfare. "We've held only one or two closing dinners," Scheinman says. "It's not the closing we celebrate--it's the integration."

"I'm there from the point of inception till you can't identify the company as an acquisition anymore." --MIMI GIGOUX

Jovial and hard working, Mimi Gigoux came to Cisco with its takeover of Kalpana in October 1994. It was Cisco's third acquisition, and she thought the company did a pretty poor job of integrating its new employees. "They didn't have anyone especially responsible for integration," she recalls during a phone interview from Copenhagen, where she's on a break from her work with CoCom, a cable equipment manufacturer that Cisco recently purchased. "There were these very long and complex meetings with no one tracking progress. It took months before everyone from Kalpana knew what their role was at Cisco and months after that before everyone was moved and located at Cisco."

Gigoux gave a few pointers that so impressed her Cisco bosses that they asked her to help with a few of the following acquisitions. Five years later, she has a staff of 11 solely dedicated to integration. A group stays at a company from the start of the acquisition to the day the deal closes.

Gigoux's process is tailored to each acquisition yet incredibly quick. For each she assembles a customized packet of information that includes descriptions of Cisco's structure and employee benefits, a contact sheet, and an explanation of the strategic importance of the newly acquired company. The day the acquisition is announced, human resources and business development teams travel to the acquired company's headquarters. There they meet in small groups with people from the acquired company to set expectations and answer questions.

The Cisco integration team collaborates with the acquiree's management in "mapping" employees--figuring out where in Cisco they belong, basing decisions on experience. In general, product-engineering and marketing groups remain as independent business units, while sales and manufacturing groups are folded into existing Cisco departments. Meanwhile, back at Cisco, teams get the new employees onto Cisco's computer system, and change payroll and stock-administration programs.

The day after the deal closes, Gigoux's department offers a tailor-made orientation, schooling managers in Cisco's hiring practices, sales people in Cisco's products, and engineers in its development projects. In total, the process takes 30 days. "I joke that I'm going to get companies to the point of do-it-yourself integration, where I just send a videotape of me talking," she says.

Gigoux's being modest; she's much more than a talking head. Outsiders praise her team for being able to digest an entire company without hurting its productivity. While Cisco leaves much of the existing infrastructure in place, it makes sure its new employees know just who their new employer is. "We closed the deal at 11 p.m. on a Wednesday," says Lori Smith, the human resources director at Monterey. "When I walked in Thursday morning, we all had Cisco tags on our doors and a banner on the front of our building. And they had this huge Cisco art thing on the wall in the lobby. I saw someone in here putting bottled water in the fridge to replace our coolers. They really don't mess around."

Internally, Cisco called this merger "boring" since it went so smoothly. All of Monterey's 147 employees remained at the Texas headquarters, and Bass was made a vice president in charge of a new business unit. For Monterey employees, it was anything but boring--they downed shots to celebrate the deal and their newly acquired Cisco options. Of course, it's these options, and the continued run of Cisco's stock, that might prove to be the most valuable integration and retention tool. "It's easier to integrate engineers who are rich and happy than ones looking for a way out," says Paul Sagawa, an analyst at Sanford C. Bernstein.

"I don't believe mergers of equals work." --JOHN CHAMBERS

The best thing about interviewing Cisco CEO John Chambers is that he's as ready to point out his competitors' mistakes as he is his own. So you feel he's being straight with you. "Almost any combination outside of Cisco hasn't lived up to expectations," he says. "There's no better example than the deal between network equipment makers Synoptix and Wellfleet. Our two toughest competitors combined and took themselves out of business."

Turn the conversation to Cisco, and Chambers is equally frank in discussing its acquisition methods. As with any heavily used process, he says, there are bound to be mistakes. For Cisco, they've arisen when the company has deviated from its strategy by forcing a company into Cisco's infrastructure. ("It was my naivete when we acquired software companies and moved them from their own distribution systems into ours. Dumb. Dumb.") Or when Cisco has mistimed its bid. Valley pundits say the 1996 acquisition of Granite Systems, a maker of so-called gigabit ethernet switches, didn't work because Granite's product wasn't as far along as Cisco had believed. On the other hand, people close to one of Cisco's largest acquisitions, that of StrataCom in 1996, say it was difficult because StrataCom was too large and its product too developed. Cisco had trouble integrating elements of its operating system into StrataCom's switches.

This particular problem doesn't bode well, say some analysts. As telecom equipment giants Nortel (revenues: $17.6 billion) and Lucent (revenues: $38 billion) take aim at Cisco's markets, they say, Cisco will need another large acquisition. Chambers, however, is forging ahead with smaller ones, telling attendees at October's Telecom 99 conference in Geneva that he'll make up to 25 this year. "In a merger you can't blend resources and cultures--only one can survive," he says. So Chambers and his team are busy scouring the world to make sure the survivor is Cisco.