Cerent fit the model of what Cisco was looking for . . . a startup company that possessed what it needed to achieve a Number 1 ranking in a new telecom market segment (and do so with an optical transport platform that could quickly penetrate global markets).
 
Cerent’s management team also demonstrated a mindset that did not require, or desire, its own in-house manufacturing (Cerent had already adopted the outsourced factory model), and it was a company that had the technology to evolve well into the next decade [1] so future internal development costs would be low.

​Coase’s Conclusion
One rationale for Cisco’s rapid move to acquire Cerent could have been presaged by economist Ronald Coase, who wrote, “The Nature of the Firm,” an article [2] penned in 1937. Even though vertical integration was all the rage [3] during the interval between the two world wars, Coase sought to define those factors that either favored greater vertical integration or those that constrained it.

Cisco, of course, sought to constrain vertical integration.
 
Coase quotes Mr. Maurice Dobbs in his seminal 1937 piece, “. . . The undertaker busies himself with the division of labor inside each firm and he plans and organises consciously [but] he is related to the much larger economic specialisation, of which he himself is merely one specialised unit.”
 
Cisco, focused on its core competencies and, although it acted as but one cog in the larger telecom machine, it was, just like the undertaker, conscious of the larger eco-system in which it operated. During the 1990s, Cisco and many other companies – such as Dell and the upstart startup Cerent – leveraged late 20th-century technologies that pre-World War II companies did not have.
 
Migrating from Frenzied to Focused
After it abandoned its frenzied acquisition spree of the late 1990s and early 2000s, Cisco opted to focus on innovation (with internal product development), brand and market management (leveraging its industry power), and advancing its IT resource strength.
 
How did Cisco do what companies in a bygone era could not?
 
Cisco exploited three key areas (that had since become available) in which to advance its minimalist headcount environment:

1. Adopt a new management technique,
2. Utilize global transportation networks, and
3. Take advantage of advanced communication technologies, many of which it acquired or developed.

​To wit, managerial techniques were honed with the arrival of “new software tools that shifted power upward within the organization.” For example, Cisco could close its books each quarter in less than 24 hours, an achievement of Larry Carter’s accounting organization.
 
Modern transportation systems with their accompanying logistical support services brought manufacturing resources as far away as Asia to within a day of Cisco’s California home base. Consequently, Cisco contracted out as much of its manufacturing as it could.  
 
And, the rise of the Internet, which Cisco advocated, along with the collapse of telecommunications costs (by simply cannibalizing old telco networks in favor of modern IP–based networks) made it practical to more tightly organize a global workforce. 

And these three productivity improvements, most efficiently exploited by Cisco management through the early 2000s, allowed the company to boost its valuation to stratospheric heights, at least for a short time, while leaving its struggling competitors in its wake.
 
It’s no wonder that Cisco acquired startups such as Cerent as fast as it could in the late 1990s.
 
[1] On December 9, 2010, Carrier Ethernet News reported, “XO Communications CTO Randy Nicklas declared that XO is not planning to buy anymore SONET moving forward as it rolls out a big network upgrade over the next three years. XO isn’t dumping existing SONET equipment in its network, mind you, but it is not continuing to purchase and embed it into its operations.” The story continued, “Most carriers really, really want to kick the SONET habit. Everyone, [including SBC-AT&T’s Chris Rice who tried to stop SONET purchases, in 2006], has vowed to stop buying the technology, but like any addictive drug, it’s hard to go cold turkey overnight.”
[2] The Nature of the Firm, R.H. Coase, Economica, November 1937, pp.386–405.
[3] “Given the existence of a robust base of suppliers and small innovators,” in the 1990s, “Cisco decided to cut loose from the parts of its business that were most difficult to expand organically – namely manufacturing and R&D.” Cisco opted to forego vertical integration, while at the same time, adds Barry Lynn, adopting “a strategy of acquiring a closely related set of products, much like General Motors in the 1910s or General Foods in the 1920s. The result was that managers, over the course of five years and on a global scale, were able to capture for their company the de facto title of General Internet.” It turns out that Cerent fit both of these criteria – no in-house manufacturing nor internal R&D – and so it became a no-brainer to acquire this upstart startup, solely for these reasons. 
Source: End of the Line: The Rise and Coming Fall of the Global Corporation, Barry C. Lynn, 2005, Doubleday, New York, chapter 5.
 

​Cisco, almost from the start, operated as an “arbitrage-oriented firm.”  
 
Such firms, as described by Barry Lynn in his book, End of the Line: The Rise and Coming Fall of the Global Corporation (2005), “are designed to focus much more on using their power over their production systems to wring out wealth immediately, rather than to devote resources to technologies that might create wealth years from now.”
 
The emphasis placed on “production systems” is mine. In an earlier BLOG post, I reviewed how Cisco leveraged outsourcing to bring its manufacturing costs way down. For example, “By 2000, some 90 percent of Cisco’s subassembly work took place outside the company, as Cisco managers kept control only of high-end, developmental production work.”
 
Cisco and other firms of this bent viewed “development of new technologies and new products, rather than an investment in the future, as no more than an unnecessary and probably unwise cost.”
 
A Frenzied Acquisition Force
And it was this inane philosophy that drove Cisco to its acquisition and develop (A&D) mindset instead of the more traditional research and development (R&D) approach used by competitors such as Nortel, Lucent, and Tellabs.
 
During the 1990s, many entrepreneurs in and around Silicon Valley, including Ed Paulson, who, in his book, Inside Cisco (2001), began to refer to Cisco “as the ‘borg,’ [1] owing to the company’s ‘innovation through acquisition strategy.’ Between mid-1993 and mid-2001, Cisco purchased 71 firms in one of the most extravagant buying sprees in high-tech history.”
 
Cisco collected engineering talent and new products with a speed that astounded industry observers and Wall Street denizens alike. The company knew, as Lynn reported, “that success lay in moving very fast to capture control of its marketplace. And in part, the strategy emerged simply because the company could afford to do so.”
 
As the price of the Internet giant’s stock soared, “Cisco’s top managers found themselves atop a vast pool of ‘capital’ for acquisition,” and they wasted no time in using that value to ramp up its buying spree in 1998.

​Cisco’s strategy was simplicity itself.  Customers shared their needs and the company would buy the required technologies to satisfy those needs. Cisco adapted to emerging technologies as close to instantaneous as any company could. One could argue that Cisco was tuned to the market just like a venture capitalist firm. And the company was able to ride the wave of technological fads.
 
For instance, Cisco’s investment in Ethernet switching was a big success.
 
Their subsequent investment in ATM turned out to be a bust in the long-term, but it allowed them a Trojan horse filled with IP/Ethernet-solutions to infiltrate the service provider marketplace.
 
However, the company fared much better in optical investments. Cerent was the standout in optical transport, but the other five of the six company investments in optical products were a bust [2]. However, the good news here is that the engineering talent that remained from the other acquired optical companies helped to rapidly evolve the Cerent 454 (now the Cisco ONS 15454) platform and make Cisco a major optical competitor.
 
Innovation through New Technology Introduction
“Viewed from an innovation system perspective,” writes Lynn, Cisco’s strategy during the 1990s helped “to maintain Cisco as the keystone predator within a venture capital ecosystem. Indeed, Cisco . . . fostered what became an entirely new model for identifying, incubating, and introducing new technologies. Small firms supported by venture capital took the hard initial steps and incurred the initial risks. In exchange, Cisco provided the ‘exit,’ rewarding founders and funders at the small companies with rich chunks of Cisco stock and sometimes lucrative jobs inside the firm.”
 
Upstart startups planted the seeds and tended their saplings, while Cisco harvested the ripest fruit. This process delivered a bumper crop through early 2001, but then, the market leader succumbed to a myriad of events that tell the other half of the story, best illustrated by the graphic below.

​​In response to these challenging times, Cisco shed both employees and product lines. The company surrendered its acquisition-and-develop model and silently telegraphed this shift in strategy as the telecom meltdown reached its height. By 2005, Cisco acknowledged its curtailment of frenzied acquisitions in most of the company’s presentations, in favor of in-house “innovation,” something that it had sporadically done in the past. 

​The question today, “Is this shift in strategy working?” And does this “mature company” approach allow Cisco to achieve growth and higher stock valuations like it once knew?
 
Its shareholders still want to know.
 
[1] In the Star Trek universe, the “Borg” annex the technology and knowledge of other alien species to the Collective by forcibly assimilating them, and in the process transforming individual beings into drones. The Borg's ultimate goal is to achieve perfection.
[2] The other five optical acquisitions included Monterey (1), acquired on the same day as Cerent, in 1999, for its optical cross-connect technology. Pirelli (2), and Qeyton Systems (3) and their WDM-centric products turned out to be a bust in terms of generating revenue. Other optically-oriented resources shoehorned into the optical business unit from other failed acquisitions such as Pipelinks (4) and Fibex (5) had to be redeployed from a management perspective. For example, research scientists out of Monza, Italy from the Pirelli acquisition were able to add value to the acquired Cerent product line, but this occurred years later. For the company as a whole, Cisco doubled its revenue from $19 billion in 2000 to $38 billion in 2013. Cerent’s one billion dollar contribution was more than five percent of Cisco’s 2000 total revenue metric, almost all of which bolstered the company’s service provider segment. Read chapter 9 of The Upstart Startup for more.

​McLeodUSA was a growing Competitive Local Exchange Provider (CLEC) throughout the 1990s after being founded by Clark McLeod in 1992. The company initially provided fiber optic maintenance services for the Iowa Communications Network. During the early part of Cerent’s engagement with the Cedar Rapids, Iowa-based service provider, Ovation Communications was acquired by McLeodUSA in the fall of 1998, an early sign that the alternative provider market was about to consolidate.
 
Cerent Targets McLeodUSA for Business
On December 5, 1998, Terry Brown, Cerent’s vice-president for sales, marketing, and customer service, and I, met to review McLeodUSA’s need for a “3/1 Digital Cross-connect System” or 3/1 DCS for short.
 
Their Request For Information (RFI), a pre-cursor to a formal bid, intrigued us as McLeod’s ideas contained therein, mirrored our (Cerent’s) philosophy regarding distributing the management of bandwidth around the network. The opportunity to influence their formal Request for Proposal (RFP) and to be considered for delivering $300 million worth of product over a five-year period was enticing. The relatively low circuit count (of DS1s and DS3s) fit the architecture of the Cerent 454 perfectly [1].
 
The Cerent Approach – A 1998 Perspective as it Applied to McLeodUSA
I worked with Jeff Santos and later, Bruce Hollett on Jeff’s behalf, to position Cerent and its groundbreaking product for McLeodUSA’s future growth. Since competitor Positron was already deployed, we positioned our planned integrated ATM capabilities as a technology approach to “cap and kill” Positron [2]. Consequently, we presented Cerent as the “transport vendor of choice” for all of their TDM (DS1 and DS3 circuits, for example) and for their ATM traffic across their expanding backbone network. As Alcatel was also a favorite of McLeodUSA’s engineers, it was important that we demonstrate interoperability with both Positron and Alcatel [3].
 
The McLeodUSA Strategy Defined
I wrote, “The McLeodUSA opportunity enables us to reinforce our position in the marketplace and play to a major strength of the Cerent 454. This strength is our bandwidth management capability in a 3/1 cross-connect application.  Although we will be highlighting the STS and VT management capabilities of the ‘454,’ we will not deviate from our long-term vision of what role we want the ‘454’ to play in customer networks. The Cerent 454 will be positioned as the ‘Transport Manager of choice.’”
 
I added, “It will be presented not only as able to accommodate the ‘asked for’ TDM traffic fill, but also as an ATM aggregator for McLeodUSA’s data traffic . . . We will also take the opportunity to highlight the hubbing capability of the platform, such that multiple UPSR rings can be simultaneously subtended from the [main] ‘cross-connect’ location.  This approach will include interoperability with the new ATM Mux device that Joe Gerke is deciding upon.”
 
In this strategy definition, I included a figure to clearly illustrate the area of the network we targeted with our proposal. Cerent brought significant value to McLeodUSA so much so, that a lab evaluation began in January 1999.

And It’s Even Hotter in the Summer
On June 3, 1999, McLeodUSA announced its intention to buy Access Communications, based in Salt Lake City. The purchase expanded McLeodUSA’s voice and data communications market by 23 percent.
 
A couple of months later, Jeff Santos held a breakfast meeting with McLeodUSA on August 11, 1999, in Cedar Rapids. He met with Mike Boehne, Jeff McKeefry, and Trent Bonstetter. The former two gentleman had already visited Cerent in Petaluma the previous October, when they were with Ovation Communications. Bonstetter hailed from MEANS, another Eric Clelland customer. Jeff characterized their discussion as “extremely friendly and informative.”
 
He wrote to his salesman, Bruce Hollett, and the rest of us supporting Bruce, in growing the ‘454’ business: “Great job to date. These guys are jazzed about a long-term relationship with Cerent. Congrats to you and the support team for a job well done. Let's go drive a truck through this one.”
 
It turned out that Mike Boehne was moving more and more of the vendor selection process into engineering and out of research and development. The impact the acquired Ovation folks were having on McLeodUSA was being felt. Jeff told us, “Steve Graham and Ron Speltz are key folks to work with in Cedar, [with] Scott Jennings and Ron Speltz driving new backbone designs [using] OC-192 for mid-next-year with express channels and ADM sites . . . They still like 4F-BLSR for the backbone, [but they] don’t see the need for matched nodes.”
 
At the time, McLeodUSA remained interested in DWDM functionality on our OC-48 configurations and sough even better OC-48 optical performance. Jeff summarized the state of affairs, “They are deploying ‘454s’ today in city rings . . . Release schedule integrity is one of their TOP priorities.” 
 
In a sign that Cerent was growing too fast in mid-1999, Trent Bonstetter made a couple of comments about the fact that he ran into some ‘newbies’ on Cerent’s Customer Technical Assistance Center (CTAC) phone line. Jeff reported, “Trent knew more about the equipment than they did.”
 
Fun in the Fall
On October 4, 1999, Jack DeGrace, a Cerent engineer supporting McLeodUSA in Iowa wrote, “Overall, very positive feedback on the capabilities and ease of use of the ‘454.’”
 
What a year!
 
[1] In a distributed-bandwidth-management (or “cross-connect”) configuration, performance monitoring and built-in testing capabilities of the DS1s, in particular, was a crucial feature, that the Cerent 454 did not possess in its initial Release 1 and Release 1.1 offerings. However, the fact that the Cerent 454 could address 80 percent of the small (distributed) cross-connect market more cost-effectively than Tellabs and its vaunted Titan product came as a major threat to that telecom supplier.
[2] At Supercomm 1998, held in Atlanta during June of that year, Positron Fiber Systems (PFS) Corporation of Montreal announced that its Osiris line of SONET multiplexers could be field-upgraded from its existing OC-3 or OC-12 rates to higher-rate OC-48. Positron claimed that all that was needed was a new OC-48 plug-in card to replace either the existing OC-3 or OC-12 one. Sounding like a Cerent 454, yet to be formally announced one year later, Positron claimed that service providers could directly access Ethernet and FastEthernet LANs as well as DS1 applications, such as PBXs, without having to use external bridges, routers, or lower speed SONET boxes. Having a direct DS1 to OC-48 connection was becoming a big deal. Positron started the ball rolling, but it would be Cerent that would soon eclipse PFS in this space.
[3] By the time Cisco completed its acquisition of Cerent, the upstart startup had not yet formally documented interoperability with Positron. However, Cerent had already completed interoperability with the more popular Alcatel products: The 1603SM with its OC-3, DS1, and DS3 interfaces, performed in Cerent’s Petaluma interoperability lab that Ron Ostrowski and I set up; the 1612 at the OC-12 rate in the field during the Golden Valley Electric (Alaska) installation; and the 1648 at the OC-48 rate in the field during the Grande River Communication (Texas) system line up and test.
[4] According to Acre Broadband, as of August 2017, the company presented Joe Gerke’s credentials, a portion which follows: “In 1993, Mr. Gerke joined McLeodUSA . . . he held a variety of roles in Information Systems, Research & Development, Network Services and Network Operations.  He played a key role in developing the strategic proposal for the Iowa Communication Network expansion, which resulted in a $45 million network build.  He was also a key team member that designed, constructed and operated of McLeodUSA ATS, one of the first ‘fiber to the curb’ networks in the nation throughout Cedar Rapids, Iowa, a network that is still operational today as ImOn Communications.” As I see it, Gerke evaluated ATM-edge technology that was soon superseded by packet-based technology to accommodate data across all telecom networks. Gerke left McLeodUSA in 2001.
[5] Cerent’s initial product release was not enough to satisfy McLeod USA. A subsequent release, due out in the summer of 1999, would be needed to support UPSR rings, VT-bandwidth management in support of the DS1 management capability, an EC-1 functionality to support SONET electrical interfaces (STS-1), and transmux functionality. Cerent also had to better define its DS1 test access capability, a feature also demanded by Williams and others. Already agreed upon was the ability of the ‘454’ to provide continuous DS1 performance monitoring and information on a planned “Cerent Management System” or CMS.
[6] On October 27, 1998, Ovation Communications visited Cerent’s Petaluma facility. Eric Clelland hosted the customer. The three attendees included Mike Boehne, Vice-president Technology Services, with responsibility for all network deployment; Greg Hutterer, Director of Product (Services) Development, responsible for new service development such as xDSL, ATM, LAN Service (TLS), and Frame Relay; and Jeff McKeefry, Director Network Engineering, responsible for deciding which equipment would be “standardized” for deployment in building the multiple city (Chicago, Milwaukee, Detroit, Minneapolis, Indianapolis, and Cleveland) interconnections. Although Ovation was using Fujitsu and Lucent equipment, Jeff was interested in looking beyond the established vendors when Eric made his initial contact with Ovation. Besides the inter-city opportunities, local ring opportunities were emerging, as well as the 3/3 and 3/1 cross-connect applications, similar to those at McLeodUSA.
[7] Keith Hanna took over the direct sales interface from Bruce Hollett in 1999 after Cisco acquired Cerent. Keith notes, “Shortly after Cisco acquired Cerent, I was hired on and ended up managing the McLeodUSA account as they ramped up during the fever pitch of the 1999-2000 market expansion. I remember that my regular travel routine was Chicago to Cedar Rapids to Minneapolis to support and serve McLeodUSA. It was quite a ride until the bubble burst in late 2000, early 2001. It was fun while it lasted.”

​“Carl and I were chatting on top of the stairs at Petaluma one day,” Dave Cesca told me, regarding some time he spent with Carl Russo, Cerent’s CEO, on a rare visit Dave made to Cerent’s Petaluma headquarters. 
 
“We were talking about the There’s Something About Mary movie [released in 1998],” Dave said. “A couple of investors were coming up the stairs. Carl and I sort of look alike, and they asked, “Who is Carl Russo . . .?” 
 
“We pointed at each other at the same time, and we both said, ‘Him.’”
 
Indeed, there was something about this relaxed, but intense pair of Cisco executives. Although they were joking around with Cerent’s arriving (potential) investors that day, both were serious about growing the business – Carl as CEO and Dave as one of Cerent’s sales leaders.

​Less than a year later, both Carl and Dave became part of Cisco, as the Internet giant completed its acquisition of the upstart startup, Cerent. And there was something about Cisco.
 
Cisco Could Do No Wrong . . .
Kevin Tolly, in December 1999, wrote, “Cisco has its own allure and is not governed by the rules that govern other vendors . . .” He took a lot of flak for apparently elevating Cisco above the rest of the equipment manufacturing pack – Nortel, Lucent, Alcatel, and many others.
 
Tolly had to quickly clarify his position: “What I mean is, Cisco's customers often let the company ‘get away’ with things that other vendors would never be allowed to get away with. For example, if the Accelar did not perform at wire-speed at Layer 2 and 3, for all ports, with quality of service disabled or enabled -- Nortel Networks would have a tough time trying to sell it in the marketplace. Cisco's customers, however, don't seem to mind a bit that the Catalyst cannot perform anywhere near wire-speed.”
 
Looking back, by 1999, Cisco still had not cracked the service provider market in a big way; at least not with the network infrastructure organizations that decided where the billions of dollars of transport and switching equipment were deployed.
 
Unlike the Enterprise customer segment, the Service Providers (in particular, telcos), would not let Cisco get away with products that were not RUS-listed, nor Telcordia compliant, let alone not meeting the basic five ‘9s’ of reliability. And for a product not to meet baseline specifications, such as Cisco’s Catalyst’s poor performance, it would never reach the field. Orders for it would not be allowed to be placed.

​Tolly argued his case when it came to Cisco’s growing customer base in the late 1990s, “I try to deal with facts, and the fact is that customers and prospects treat Cisco differently than they treat its competitors. Right or wrong, that is what I observe.”
 
Indeed, Cisco continued to grow faster than all of its competitors, the company kept acquiring startups to bolster its engineering staff and expand its market reach [1], and the San Jose-based giant could do no wrong, especially in the eyes of Wall Street and the industry analysts.

​Tolly found one example of a happy Cisco customer, a retailer, “contemplating a major migration to Nortel.”
 
The reason for selecting one of Cisco's competitors?
 
Tolly adds, “The Nortel gear was perfectly matched to its need, and with it, the company calculated WAN link savings over the Cisco products that was measured in millions of dollars annually.”
 
Do you agree with Tolly’s assertion, at least before the dot.com bust and telecom meltdown that . . . “There's something about Cisco?”
 
What about today?
 
[1] Cisco acquired Cerent for $6.9 billion in the summer of 1999 and finalized the deal in November 1999, just before Tolly penned these words.
[2] The Cerent acquisition not only gave Cisco engineering talent and hundreds of new service provider customers, but Cerent’s service provider orientation helped to transform Cisco into a company that built higher quality products and networking platforms that suited the service providers’ stringent product and network performance requirements.
[3] Guest column: There’s something about Cisco, Kevin Tolly, Computerworld. December 1, 1999.
[4] There's Something About Cisco, Andy Serwer with reporter associates Irene Gashurov and Angela Key, Fortune, May 15, 2000.