GRADUATE SCHOOL OF BUSINESS

STANFORD UNIVERSITY
CASE NUMBER: EC-15
FEBRUARY 2000

CISCO SYSTEMS: A NOVEL APPROACH TO

STRUCTURING ENTREPRENEURIAL VENTURES
Mike Volpi, vice president of business development at Cisco Systems, was in his office in
San Jose at Ciscos headquarters on June 27, 1997. He was considering strategic questions that
he had faced many times since joining Ciscos business development group in 1994. Volpis
colleagues had identified a new networking opportunity in optical routers, and Volpi wondered
whether Cisco should develop the product internally or pursue external talent that was more
familiar with the technology and market segment? If the external route was the best strategy to
get the right product to market on time, should Cisco build its own external venture or just
acquire someone outright?

NETWORKING OPPORTUNITY: PIPELINKS

For the previous two years, Cisco had been preaching about the promise of a multiservice network a single network that could transport data, voice, and video. Ciscos service
provider customers agreed that network convergence would ultimately improve cost
effectiveness and allow them to expand their service offerings. However, most service providers
were saddled with huge investments in circuit-based voice networks. This implied a market need
for optical (Sonet/SDH) routers that leveraged the existing infrastructure while enabling a
transition to multi-service networks: the market needed a product that could simultaneously
transport circuit-based traffic and route IP (Internet Protocol) traffic.1
Discussions with representatives from Ciscos Service Provider Line of Business
(SPLOB) indicated that developing optical routers internally was not a viable option. A brief
search had failed to identify attractive acquisition targets-no player had the people, products,
technological innovation, ownership concentration, and location that Cisco wanted. In a

Sonet/SDH (synchronous optical network/synchronous digital hierarchy) is a protocol for data transmission over
fiber optic lines.

Research Associates James McJunkin and Todd Reynders prepared this case under the supervision of Professor Garth Saloner
and Professor A. Michael Spence as the basis for class discussion rather than to illustrate either effective or ineffective handling
of an administrative situation. Margot Sutherland, Executive Director, Center for Electronic Business and Commerce, Stanford
Graduate School of Business managed the development of this case.
Copyright 2000 by the Board of Trustees of the Leland Stanford Junior University. All rights reserved. To order copies or
request permission to reproduce materials, email the Case Writing Office at: cwo@gsb.stanford.edu or write: Case Writing
Office, Graduate School of Business, Stanford University, Stanford, CA 94305-5015. No part of this publication may be
reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means - electronic,
mechanical, photocopying, recording, or otherwise - without the permission of the Stanford Graduate School of Business.
Version: (B 03/20/01

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 2

fortunate coincidence, however, entrepreneur Amit Shah had approached Cisco with an idea for a
Sonet/SDH router that was similar to the product that Cisco envisioned. Shahs company,
Pipelinks, was still in the idea stage so an outright acquisition was not yet appropriate. Volpi
realized that this situation exhibited similarities to one he had faced a year earlier. When Cisco
had created a custom, made-to-order company called Ardent Communications to fill a market
void. Plenty of mistakes had been made in structuring the venture, but much had been done
right. Volpi dug up the Ardent file and contemplated possible strategic and structural
improvements that could be made, in the hope that some incarnation of the spin-in model would
be an effective way to serve the current market need.
BACKGROUND ON CISCO SYSTEMS INC.2
Leonard Bosack and Sandy Lerner, husband and wife computer scientists at Stanford University
who invented a technology to link their disparate computer systems together founded Cisco
Systems in 1984.
They developed the first multi-protocol router a specialized
microcomputer that allowed two or more networks to talk to each other by deciphering,
translating, and funneling data between them. Ciscos technology opened up the potential to link
the worlds disparate computer networks together the way different telephone networks were
linked around the world.
Cisco began by offering high-end routers primarily in the LAN (local-area network) market. The
devices were the traffic cops of cyberspace they directed network traffic to its final destination
via the most efficient, least congested network path. As the global Internet and corporate
Intranets became more important, so too did Cisco. With an early foothold in this rapidly
growing industry, Cisco quickly became the leader in the data networking equipment market
the plumbing of the Internet. By 1997, Cisco made approximately 80% of the large-scale
routers that powered the Internet. Although routers, LAN switches, and wide-area network
(WAN) switches would remain Ciscos core products, the companys product line included other
networking solutions, including Web site management tools, dial-up and other remote access
solutions, Internet appliances, and network management software. Despite the breadth of its
product offerings, Cisco held the number one or two position in most markets in which it
competed. Ciscos Internetwork Operating System (IOS) software was also becoming the de
facto industry standard for delivering network services and enabling networked applications.3
Cisco received its initial funding from the venture capital firm Sequoia capital, who helped to
recruit John Morgridge as CEO in 1988. The company went public in February 1990 with a
$222 million market value and grew into a multinational corporation with over 10,000 employees
in 54 countries. By 1997, revenues had increased over ninety-fold since the IPO, from $69.8
million in fiscal 1990 to $6.4 billion in fiscal 1997. (Exhibit 1) In June 1997, Ciscos market
value totaled $46.3 billion.

Excerpts taken from Cisco Systems, Inc. Acquisition Integration for Manufacturing, Case # OIT-26, Graduate
School of Business Stanford University and Harvard Business School, revised January 1999.
3
Ciscos IOS Software was the industry leading internetworking software, like Microsoft Windows for networking.
IOS is a platform that delivers network services and enables networked applications. IOS enables interoperability
connections between otherwise disparate hardware, and accommodates network growth, change, and new
applications. It also contains security features, including access control, authentication, firewall, and encryption.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 3

Two respected CEOs have led the company: John Morgridge and John Chambers. Morgridge
shaped the Cisco culture from day one, focusing on customer satisfaction, product quality, and
frugality. He once gave a legendary presentation on frugality to the Cisco sales force, after being
appalled by reports that salespeople were flying first class on business trips. Equipped with
slippers, earplugs, and eye covers, Morgridge displayed how to fly coach and make it seem like
first class. John Chambers, who joined Cisco in 1991 and succeeded Morgridge in January 1995,
was known for his fair but ultra-competitive nature. Chambers, a former IBM and Wang
Laboratories marketing and sales veteran, fostered Ciscos strong customer focus and was
credited with continuing Ciscos striking success in the networking industry.
Corporate Strategy
Throughout the 1990s, organizations of all sizes were beginning to recognize the value of their
information networks and the Internet as a source of business advantage. As a result, more of
Ciscos customers sought end-to-end networking solutions. Building on its expertise in routers,
Cisco strove to deliver a wide range of new products, expand its offerings through internal and
external efforts, enhance customer support, and increase its presence around the world.
The main element of Ciscos strategy during this expansion phase was to maintain a passionate
customer focus and consistently try to exceed customer expectations. To deliver on that goal,
Chambers reorganized Cisco to target three key markets: Enterprise, Service Providers, and
Small/Medium Business. The new organization enabled Cisco to provide market specific, endto-end solutions that included integrated software, hardware, and network management and to
customize its sales, support, and business programs to each market.
One of the keys to the companys success was the Cisco brand, which was recognized as a
leading name in networking. Customers associated the Cisco brand with a secure, reliable, highperformance network. Chambers wanted to enhance and expand the brand, and increased
Ciscos marketing to include television, Internet, and print advertising.
The ongoing deregulation of telecommunications and technology convergence were driving the
trend toward the integration of voice, video, and data networks. Historically, there had been
three separate types of networks: phone networks for transmitting voice, computer networks for
transmitting data, and broadcast networks for transmitting video but advances in digitization
allowed these forms of communication to be translated into binary computer language. This, in
turn, made it possible to transmit voice, data, and video over one network more efficiently and
economically than using three disparate networks. As a result, phone companies were beginning
to transform their archaic voice networks into unified, multi-service networks.
Chambers believed that this transition to the New World of communications would enable
Cisco to capture share in the $250 billion telecom equipment market that huge, well-capitalized
companies such as Lucent Technologies and Northern Telecom had dominated. These
competitors were so large that Chambers instilled a David vs. Goliath mentality within Cisco.
While expanding into these new markets, Cisco also strove to maintain its product leadership in
each of the market segments it already served. The product leadership strategy involved the
innovation of Cisco's engineering teams, complemented by alliances, acquisitions, and minority
investments.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 4

Building Shareholder Value through Acquisitions, Investments and Alliances

As the networking space became more competitive, and as minimizing time to market became
more important, Chambers realized that Cisco could not keep up with the changing market needs
solely through internal development. Acquisitions and alliances to gain access to world-class
technologies and people became a defining component of Ciscos strategy. This strategy was
relatively unique: most high-tech companies considered looking to the outside for technological
help a sign of weakness. Chambers commented on the acquisitions and alliances strategy:
They are a requirement, given how rapidly customer expectations change. The
companies who emerge as industry leaders will be those who understand how to
partner and those who understand how to acquire. Customers today are not just
looking for pinpoint products, but end-to-end solutions. A horizontal business
model always beats a vertical business model. So youve got to be able to provide
that horizontal capability in your product line, either through your own R&D, or
through acquisitions.4
Although Chambers and Ed Kozel, Ciscos chief technology officer, were a key driving force
behind Ciscos business development strategy, many in the industry regarded Mike Volpi as the
man responsible for shaping Ciscos legendary business development practice.5 When Volpi
joined Cisco in 1994 after graduating from the Stanford Graduate School of Business, Cisco had
completed only one acquisition, Crescendo Communications. Two more acquisitions closed
soon after Volpi arrived, but he was involved in all subsequent acquisitions.
Before pursuing a new market opportunity, Volpis group assessed the buy vs. build strategies.
If Cisco did not have the technological capability, engineering capacity, or time to develop the
product internally, the business development group would often opt to acquire or partner with an
external player. Although the acquisitions made headlines, licensing, partnering, and investing
were equally important to Ciscos strategy. Cisco was an active minority investor, which gave it
insight into new technologies without having to deploy internal development resources. Volpi
used a simple chart to assess companies (Figure 1).
Figure 1: Range of Ciscos Business Development Activities
Acquire
Invest
Joint R&D/ Marketing/Sales

Degree of
Strategic
Value

OEM/ License

Level of Commitment

The Art of the Deal, Business 2.0, October 1999.

Volpi initially reported to Charles Giancarlo, who joined Cisco through the Kalpana acquisition in 1994 and served
as VP of business development until March 1997, when Volpi assumed that title.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 5

The public equity markets were the principal exit strategy for hot high-tech start-ups, but a Cisco
acquisition appealed to many networking companies. Cisco was the most effective tech company
at identifying, acquiring, and successfully integrating companies into its culture. By June 1997,
after the Ardent deal closed, Cisco had acquired 19 companies for an aggregate total of roughly
$7 billion (Exhibit 3). Why did Cisco do this better than the competition? We made every
mistake in the book, Volpi stated, but we learned from these mistakes, and they have helped us
in subsequent transactions.
Instead of acquiring large, established, public companies, Cisco typically acquired small
private companies, for $200 million or less.6 The smaller acquisitions made integration easier
large, established companies with strong corporate cultures were more difficult to integrate.
Chambers also asserted that Cisco did not acquire to gain short-term market share, but to find
technology and talent for the future:
When we acquire a company, we arent simply acquiring its current products,
were acquiring the next generation of products through its people. If you pay
between $500,000 and $3 million per employee, and all you are doing is buying
the current research and current market share, youre making a terrible
investment. In the average acquisition, 40 to 80 percent of the top management
and key engineers are gone in two years. By those metrics, most acquisitions
fail.7
Charles Giancarlo, Ciscos vice president of business development from 1994 to 1997, reiterated
the importance of acquiring and retaining key people:
When you are buying a company its obviously not for todays products. That
means keeping the people in place who can create that growth. We wont do a
deal if a company has golden parachutes accelerated vesting for employees
that kicks in once a company is sold. The minute you buy the company, they all
get rich. We prefer golden handcuffs, which are applied with two-year
noncompete agreements with key executives and technical personnel at the target
companies, and the provision of Cisco stock options that vest over time.8

LOOKING BACK TO 1996: THE MARKET OPPORTUNITY FOR A NEW ACCESS PRODUCT
In 1996, the evolution of network infrastructure was creating business opportunities in virtually
every sector of networking. Ciscos unique vantage point allowed it to rapidly identify these new
markets. By early 1996, Cisco believed that a need existed for an inexpensive product to carry
voice, data, and video traffic from a companys local-area network to the wide-area network.
Cisco identified two principal customer needs. The first was to simplify and improve
management of network access equipment. The conventional approach to public network access
required cabling disparate hardware components (such as leased line modems, channel banks,
6

The $4.6 billion acquisition in April 1996 of StrataCom, which filled Ciscos hole in WAN switching products,
stands out as an exception.
7
The Art of the Deal, Business 2.0, October 1999.
8
Ciscos Secret: Entrepreneurs Sell Out, Stay Put, Inc. Magazine, March 1997.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 6

etc.) together, creating a complex hardware puzzle. Companies incurred high maintenance costs
and trouble-shooting nightmares because a different management system controlled each
component. The second was to optimize use of expensive WAN bandwidth. Despite the
industry buzz about high-speed ATM trunks,9 Cisco believed that these solutions would remain
expensive especially compared to LAN bandwidth where Ethernet technology10 dominated.
Cisco expected high-speed network access solutions, running at T3 (4.5Mbps) or OC-3
(155Mbps), to be confined to niche markets for the foreseeable future. Most customers would
choose the slower, more economical T1/E1 (1.5Mbps) link to the WAN.
These factors highlighted a market opportunity for an access solution that aggregated LAN-based
data, voice, and video traffic over the low cost T1/E1 ATM trunk. This solution would help
service providers:
Provide an integrated T1/E1 access solution that was cost-effective for wide
deployment
Contain costs by using a single product in multiple applications
Contain upgrade/conversion costs by using a remotely configurable product
Contain support costs by using a product with an interface familiar to both customers and
service providers.
This product concept was the genesis of Ardent Communications.
New Venture Strategy
In 1996, Volpi contemplated the traditional buy and build alternatives for the market that Ardent
Communications would serve. Building the product in-house had several advantages notably
not having to integrate two different organizations. The multi-service access business unit had
been doing similar things on a day-to-day basis, but lacked the human resources to devote to the
new project. Diverting resources away from current projects was not feasible. Building the
product in-house would take too long competition from 3Com, Ascend, US Robotics, and
Micom made time to market a priority. The business development team concluded that Cisco had
neither the time nor the resources to go after the new market on its own.
Buying a company whose products addressed this market was another option. However, Cisco
had a clear conception of the market need, but was unable to identify attractive companies that
were focused on this space. Volpis experience suggested that finding the right acquisition target
would be difficult in all cases Cisco would have to spend time and effort modifying the product
set and integrating the newly acquired company into the Cisco organization. Retaining key
employees post-acquisition was also always difficult.
Finding neither the buy or build alternatives satisfactory, Volpi mused, Why not custom make a
start-up to build exactly the product we want, and then buy them later if they succeed? This
solution would entail creating a new venture as a spin-in from day one build to buy. This

A trunk is an access line that connects remote offices or central sites to the service provider network. Asynchronous
Transfer Mode (ATM) is a data transfer technique where multiple service types, such as voice, video, or data, are
conveyed in small, fixed-size cells.
10
In 1996, Ethernet technology penetrated every corner of the Enterprise network with 10BaseT (10Mbps),
100BaseT (100Mbps), and the coming Gigabit Ethernet (1000Mbps).

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 7

spin-in model seemed to address three key issues: time-to-market, recruiting top talent, and
integration with the relevant Cisco business unit.
However, the Cisco business development team realized that the hybrid nature of the spin-in
solution raised difficult tradeoffs. What structure would allow the start-up to leverage Ciscos
strategic assets without quashing the entrepreneurial feel? How should Cisco structure the
venture to minimize the tradeoff between the virtues of independence and the need for smooth
ex-post integration? Could Cisco personnel coach the new team without stifling creativity?
Should Cisco invite other investors to participate in the financing? How large an initial
ownership stake should Cisco take in the venture? Incentives would also be a major issue: How
could Cisco provide the right incentives for the new ventures management and employees,
without upsetting the current Cisco employees who would help integrate the new venture?
Eventually, the new venture would have to live within an existing Cisco business unit and rely on
Cisco employees for success.
Structuring the Ardent Communications Venture
To develop a potential model for the new venture approach in 1996, Volpi and Kozel had
reflected on an earlier deal that Cisco had considered. In the spring of 1996, Wu Fu Chen, a
networking entrepreneur, was working with Sequoia Capital and two Cisco employees to launch
a new networking company. The idea for this company came from the Cisco employees, who
intended to leave their jobs at Cisco to build a solution that they hoped Cisco would want to
acquire. The product concept had potential, and the founding team was flush with engineering
talent. Wu Fu Chen had co-founded four companies since 1986, including Cascade
Communications and Arris Networks. Yet the Cisco business development team declined to
invest: Chambers believed that funding Cisco employees to go out and build new networking
companies would set a dangerous precedent.
Mike Volpi and Ed Kozel believed that Wu Fu would be an excellent person to recruit as
President and CEO of the proposed spin-in venture, which they would call Ardent
Communications. Kozel contacted Wu Fu and outlined the Ardent business idea and Ciscos
spin-in concept. Volpi later characterized the initial message to Wu Fu as simply, Make this
product and well give you lots of money. After a series of discussions, Wu Fu agreed to head
up the Ardent venture.
Defining the Ardent 101 Product
In June 1996, Kozel, Volpi, and Wu Fu outlined the basic functional specification for the first
Ardent product, tentatively called Ardent 101. For Cisco to buy the new company, Wu Fu and
his team needed to develop a traffic aggregation device for data, voice, and video with certain
functional requirements (Figure 2). The group also developed milestones that would set
expectations for the product timeline (Figure 3).

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 8

Figure 2: Ardent 101 Functional Requirements

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.

Ability to accept data, voice, and video traffic

Aggregate up to 2Mbps traffic on the WAN side
Support ATM, Frame Relay, and TDM trunks
Support standard office environments
Support Bridging, IP Routing for LAN data traffic
Support Circuit Emulation for Voice and Video Applications
Support voice and data compression
The target list price for the base configuration is about $5,000; Cost of
goods target is $800 or lower
Will consider support of data encryption
Support European requirements (E1)

Figure 3: Ardent 101 Milestones

1.

2.
3.
4.

Six months after the Effective Date of the Agreement, the Company shall have
completed the specifications for function, architecture, and design for the
Product
Twelve months after the Effective Date of the Agreement, the Company shall
have begun integration of the Product
Fifteen months after Effective Date of the Agreement, the Company shall have
begun the beta program for the Product
Eighteen months after the Effective Date of the Agreement, First Customer
Shipment shall have occurred

Capital Structure
By June 14, 1996, Cisco and Wu Fus team had agreed on a preliminary term sheet for the new
venture (Exhibit 4). Kozel and Volpi invited Sequoia Capital to participate in the financing to
create a more start-up feel. To foster an entrepreneurial environment with strong employee
incentives, Cisco gave the founding team and employees a large ownership position over 55%
on a fully diluted basis. Cisco sought an equity stake of only 32% for itself. This was a major
departure from the large equity shares other parent companies were requesting in their spin-ins
and spin-outs (arguing that their intellectual property, brand name, and other resources entitled
them to free equity). Sequoia Capital also took a relatively small equity position of 11%. All
parties agreed that a balanced board of directors would deliver the right control over the
companys direction. Initially, the board would consist of Wu Fu, Ed Kozel, and Sequoias Mike
Goguen.
Unlike most venture deals, the Series A and Series B rounds were negotiated simultaneously,
with closing dates less than two months apart. In the A round (July 11 closing) Wu Fu and the
other members of the founding team would purchase 3 million shares of Series A Preferred Stock
at $0.333 per share. The low share price was analogous to cheap founders stock in an
entrepreneurial venture. Neither Cisco nor Sequoia would participate in the A round. The
implied post-money valuation as of July 11 was $2.4 million.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 9

For the B round, the new company decided to issue 11 million shares. On August 30, Ardent
received the first cash infusion of the B round, in which Sequoia Capital purchased
approximately 2.5 million shares at $1.00 per share. Cisco also made its investment at $1.00 per
share, purchasing 7.535 million shares of Series B Preferred Stock on September 20. Seven days
later, the founders purchased another one million shares. The remaining equity capitalization
consisted of 9.25 million shares of common stock, of which approximately 3 million shares
would go to the engineering team as option grants. The implied post-money valuation as of
August 30 was $23.3 million. Exhibit 5 describes the rough capitalization table Cisco used.
Retaining Key Employees
Volpi knew that even though Cisco was creating Ardent to produce a specific product, it was the
people, not the product, that represented much of Ardents value. Cisco therefore laid out a fouryear vesting period for the options granted to employees 25 percent would vest after the first
year, with the remainder vesting monthly over the next three years. Upon a change in control,
like the planned acquisition by Cisco, only Wu Fu Chens vesting would accelerate (at most one
year of vesting would remain, but he was subject to a one-year lock up agreement which kept
him from leaving Ardent upon acquisition).
Facilitating the Spin-in: The Put/Call Feature
Cisco needed a legal mechanism that would allow Ardent to cleanly spin-in at some point in
the future. The founding team proposed a simple put/call structure that would give Cisco the
option to purchase the company at a pre-specified price, but would also obligate Cisco to
purchase the company if the new team succeeded in building the product. This was the first time
that Cisco had integrated a put/call feature into a strategic investment. John Chambers and Ed
Kozel viewed it as an innovative mechanism for developing a made-to-order company. The
Option section of the term sheet explained the call option:
Until the earlier of fifteen (15) months from the closing or one (1) month after the first
customer shipment, Cisco shall have the right to acquire either all of the outstanding
equity securities of the Company, or all of the Companys assets, in Ciscos discretion,
for a purchase price of $232,500,000, payable either in cash or equity securities of
Cisco.11

Since Cisco would also write a put option, the shareholders in the new venture could force it to
purchase the company at the pre-specified price, as long as the ten specific functional
requirements were met. To keep matters simple, the put and call would have the same strike
price. The put option read:
if First Customer Shipment occurs within (15) months after the functional
requirements for the Product are first defined, and in Ciscos reasonable judgment, the
product meets the specifications set forth, each of the security holders shall sell its
Securities to Cisco, and Cisco shall be obliged to and will purchase such Securities, in
accordance with the purchase price and other terms of purchase12

11

Excerpt from Memorandum of Terms for Private Placement of Series A and B Preferred Stock of Ardent
Corporation, June 1996.
12
Ibid.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 10

Cisco believed that although the put/call structure truncated the upside for investors and
employees, it mitigated enough risk to make the investment or employment decision attractive
from a risk/reward standpoint. The option agreement turned out to be a very effective recruiting
instrument. If the product requirements and milestones were met, the 15-20 person engineering
team would share a $30 million payout in less than 15 months. The five person founding team
would do even better: delivering on the product would allow them to share more than $100
million.
Leveraging Ciscos Assets
IOS
The Ardent product would complement Ciscos existing multi-service access products (called the
3800 product family). To facilitate interoperability, Volpi decided to license Ciscos IOS
software to Ardent free of charge until Ciscos option to buy the company expired. IOS was to
be the architectural foundation for Ardent 101. Ardent would focus on adding the technologies
of ATM and Frame Relay over a T1/E1 connection, circuit emulation for digitized voice over
ATM or Frame Relay, voice compression, and telephony capabilities. These changes were not on
the official evolutionary path of IOS, though they were similar to development work being done
within Cisco. Many Cisco employees had also created and modified IOS for use in the various
products Cisco sold, but not to be sold as a shrink-wrapped software product. The procedures to
use and adapt IOS were not well documented, which would present a challenge to the Ardent
employees. Under the terms of the licensing deal, Cisco would retain all ownership of IOS,
including software Ardent developed to interact directly with IOS.
Licensing IOS software to Ardent for free was contentious it upset Cisco personnel who felt that
the company was giving away the crown jewels, the real value-add in Ciscos solutions, and then
paying to buy it back.
Engineering talent
Ardent would also need engineering help to integrate IOS into its new product. To address this
issue, a few Cisco employees were selected to work with the Ardent team throughout the
development process. This was not unusual, because Cisco had provided consulting services in
the past. Ardent paid the standard fee for these engineering resources, $250,000 per engineer per
year. Since this was regarded as a temporary assignment, these people remained employees of
Cisco under the same terms as they had before Ardent surfaced.
Testing and certification facilities
Cisco also provided testing and certification facilities. Cisco allowed Ardent to use its testing
and certification facilities free of charge until the option period expired, after which Ardent
would pay a nominal fee.
Business unit expertise
A key issue was the extent to which the multi-service access business unit should coach Ardent
through the development process and stay informed about what progress had been made. Cisco
had similar products in the pipeline, but none overlapped significantly with Ardent 101 as
defined in the product requirements document. Volpi and Kozel decided not to involve the
business units until after the Ardent product was completed. This approach seemed appropriate
because the product specification had already been narrowly defined, minimizing the degrees of

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

freedom and therefore the need for frequent coaching or updates.

stealth mode through early 1997.

p. 11

Hence, Ardent operated in

Cisco form factor

Both Cisco and Ardent wanted the new product to have the look and feel of the Cisco product
line, although the product specification did not require it. Instead of designing a tailored box for
Ardent 101, Wu Fu decided to adhere to the existing Cisco product line and required his
engineers to adopt the form factor for the Cisco 2500 series. This solution was compact and
familiar to Ciscos carrier customers. Squeezing Ardent 101 into the 2500 box would be
engineering and manufacturing challenge, but the Ardent engineers felt confident it could be
done.
Accounting issues
The Ardent spin-in model had implications on the accounting methods Cisco could employ to
account for the venture and complete the spin-in. Many of Ciscos acquisitions had used the
pooling-of-interests method. For acquisitions when Cisco paid far more than the assets book
value, the pooling method was preferable.13 Pooling-of-interests accounting required, among
other things, that the acquisition occur in a single transaction where over 90% of the company
was acquired, and that no prior control be exerted on the company. Since Cisco took an initial
32% stake in Ardent, and the call option translated into significant control, pooling was never a
possibility but the ventures strategic value outweighed the accounting effects. Volpi
commented, We certainly look at the accounting impact in our decision process, but I dont
think that accounting issues should ever dominate the strategic issues in making decisions.
The accounting for Ciscos investment in Ardent was relatively straightforward. Since Cisco
owned more than 20% but less than 50% of the company, GAAP required that Cisco use the
equity method of accounting. Hence, the Ardent investment was recorded on Ciscos books at
acquisition cost plus its pro-rata share of Ardents earnings (or losses).
The Ardent Acquisition
In June 1997, Volpi and Kozel discussed whether Cisco should exercise its option to purchase
the outstanding shares of Ardent. In many ways, the decision was immaterial, because Wu Fu
and his team were likely to meet the acceptance criteria and exercise their put option, obligating
Cisco to purchase Ardent. Hence, the key consideration was timing: Should Cisco wait until the
end of the option period to spin-in Ardent or do it now? Volpi and Kozel determined that doing
the deal sooner rather than later would deliver two principal benefits. First, Cisco could begin to
integrate Ardent into the Cisco family, speeding time to market for the Cisco branded solution.
Second, purchasing early would avoid confusing the marketplace. Ardents marketing people
had begun to put their own spin on the product, now called Integress. However, Cisco might
want to use a different marketing approach. If Ardent educated the marketplace, Cisco would
either have to continue the same marketing program or re-educate potential customers.
Historically, re-education had not worked well. In fact, competitors were highlighting Ciscos
inconsistent messages in their white papers and marketing materials.

13

Net income is generally lower under the purchase method because significant goodwill, an intangible asset which
represents the excess of the purchase price over the assets book value, must be amortized over a defined period.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 12

On June 24, Cisco announced its intention to acquire Ardent. The press release stated: Under
the terms of the acquisition agreement, shares of Cisco common stock worth approximately $156
million will be exchanged for the outstanding shares and options of Ardent (Exhibit 6). This
was consistent with the agreed upon total acquisition price of $232.5 million because Cisco
already owned 32% of the company.
Cisco paid approximately $10.00 per share. The founders received approximately $102.3 million
more than 100 times their initial investment. Sequoia Capital received $24.6 million, a
relatively small sum but still 10 times money invested in less than 12 months.
Although the Ardent deal had several flaws, Cisco had learned a lot about how to structure future
deals from the acquisition. Volpi turned his attention away from the Ardent acquisition and back
to the Pipelinks opportunity.

JUNE 1997: VOLPI CONSIDERS THE PIPELINKS OPPORTUNITY

Several of Volpis colleagues at Cisco had identified the market opportunity for a Sonet/SDH
router capable of simultaneously transporting circuit-based traffic and routing IP (Internet
Protocol). Cisco would target the product at many of its service provider customers who were
struggling to bring their networks into the New World of unified networks, but wanted to make
the transition without scrapping their existing circuit-based TDM infrastructure.
One factor was Cisco's expertise in optical routing. In 1997, it was limited. High-speed
Sonet/SDH networking solutions were much larger and more expensive (with price points in the
hundred thousand to multi-million dollar range) than Ciscos traditional products. Because this
was an unfamiliar market for Cisco, Volpi and his team looked externally to pursue the market
opportunity, but none of the existing players met Cisco's criteria.
Volpi's team believed, however, that Amit Shahs idea might address this market need. Shah,
who had sold his first networking company to Cabletron Systems, a Cisco competitor, realized
the increasing need for bandwidth on the access points of the Internet infrastructure the metro
space near large population areas. Shah conceived a Sonet/SDH router product similar to the one
Cisco envisioned. He called the idea Pipelinks. After a series of discussions, Volpi and Kozel
determined that Cisco would like to bring Shahs product to market using the Ardent spin-in
model. Shah was intrigued by the idea it seemed like an effective way to: (1) raise funds, (2)
recruit the right people, and (3) execute with customers. Once Shah had agreed in principle to
structure Pipelinks as a spin-in, Volpi sat down to outline the terms of the deal. As he dug
through the license agreements, term sheets, and product requirements in the old Ardent file,
Volpi identified several potential improvements and modifications that he would make.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 13

Exhibit 1
Cisco Systems Historical Financials

BALANCE SHEET
FISCAL YEAR ENDING JULY 31,
ANNUAL ASSETS (000s)
CASH
MARKETABLE SECURITIES
RECEIVABLES
INVENTORIES
OTHER CURRENT ASSETS
TOTAL CURRENT ASSETS
NET PROP, PLANT & EQUIP
INVEST & ADV TO SUBS
DEPOSITS & OTHER ASSET
TOTAL ASSETS

1990

ANNUAL LIABILITIES (000S)

ACCOUNTS PAYABLE
$
ACCRUED EXPENSES
INCOME TAXES
TOTAL CURRENT LIAB
OTHER LONG TERM LIAB
TOTAL LIABILITIES
MINORITY INTEREST
SHAREHOLDER EQUITY
TOT LIAB & NET WORTH

INCOME STATEMENT
FISCAL YEAR ENDING JULY 31,
NET SALES
COST OF GOODS
GROSS PROFIT
R & D EXPENDITURES
SELL GEN & ADMIN EXP
OPERATING INCOME
NON-OPERATING INC
INTEREST EXPENSE
INCOME BEFORE TAX
TAXES
NET INCOME

Source: Cisco Systems

1992

1993

1994

1995

1996

1997

35,842 $
21,102
15,874
3,701
1,673
78,192
4,114
367
82,673 $

40,323 $
51,104
34,659
6,078
8,797
140,961
12,665
519
154,145 $

39,955 $
116,477
61,258
9,142
20,244
247,076
28,017
46,866
1,974
323,933 $

27,247 $
53,567 $
284,388 $
279,695 $
61,738
129,219
279,754
758,489
129,109
237,570
421,747
622,859
23,500
27,896
81,805
301,188
26,702
59,425
116,466
197,409
268,296
507,677
1,184,160
2,159,640
48,672
77,449
172,561
331,315
274,260
457,394
583,871
1,060,758
3,985
11,174
51,357
78,519
595,213 $ 1,053,694 $ 1,991,949 $ 3,630,232 $

269,608
1,005,977
1,170,401
254,677
400,603
3,101,266
466,352
1,630,390
253,976
5,451,984

4,973 $
6,290
1,976
13,239
123
13,469
69,204
82,673 $

7,743 $
17,965
542
26,250
436
26,686
127,459
154,145 $

16,262 $
46,953
15,108
78,323
78,323
245,610
323,933 $

24,744 $
31,708 $
59,812 $
153,683 $
77,492
130,846
257,099
445,776
17,796
42,958
71,970
169,894
120,032
205,512
388,881
769,353
120,032
205,512
388,881
769,353
40,792
41,257
475,181
848,182
1,562,276
2,819,622
595,213 $ 1,053,694 $ 1,991,949 $ 3,630,232 $

207,178
656,707
256,224
1,120,109
1,120,109
42,253
4,289,622
5,451,984

1990
$

1991

69,776 $
23,957
45,819
6,168
18,260
21,391
2,088
23,479
9,575
13,904 $

1991
183,184 $
62,499
120,685
12,687
41,809
66,189
4,567
70,756
27,567
43,189 $

1992
339,623 $
111,243
228,380
26,745
72,248
129,387
6,719
136,106
51,720
84,386 $

1993

1994

1995

649,035 $ 1,334,436 $ 2,232,652 $

210,528
450,591
742,860
438,507
883,845
1,489,792
44,254
106,680
306,575
130,682
276,995
485,254
263,571
500,170
697,963
11,557
22,330
40,014
275,128
522,500
737,977
103,173
199,519
281,488
171,955 $
322,981 $
456,489 $

1996

1997

4,096,007 $ 6,452,000
1,409,862
2,243,000
2,686,145
4,209,000
399,291
1,210,000
886,048
1,370,000
1,400,806
1,629,000
64,019
262,000
1,464,825
1,891,000
551,501
840,000
913,324 $ 1,051,000

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 14

Exhibit 2
Cisco Systems Monthly Stock Price Chart: February 1990- June 199714

$18.00
$16.00
$14.00
$12.00
$10.00
$8.00
$6.00
$4.00
$2.00

Source: Cisco Systems

14

Prices adjusted for all splits since IPO, based on January 25, 2000 stock price.

Feb-97

Aug-96

Feb-96

Aug-95

Feb-95

Aug-94

Feb-94

Aug-93

Feb-93

Aug-92

Feb-92

Aug-91

Feb-91

Aug-90

Feb-90

$-

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

Company

p. 15

Exhibit 3
Summary of Ciscos Acquisitions as of June 1997
Date
Purchase Price
Description

Crescendo
Communications, Inc.
Newport Systems
Solutions, Inc.

September 1993

$95 million

July 1994

$93 million

Kalpana, Inc.

October 1994

$240 million

LightStream Corp.

October 1994

$120 million

Combinet, Inc.

August 1995

$132 million

Internet Junction, Inc.

September 1995

not public

Grand Junction, Inc.

September 1995

$400 million

Network Translation,
Inc.

October 1995

not public

TGV Software, Inc.

January 1996

$138 million

StrataCom, Inc.

April 1996

$4.666 million

Telebit Corps MICA

Technologies

July 1996

$200 million

High-performance work group CDDI

and FDDI switching solutions.
Software-based routers for remote
network sites of small/medium sized
networks.
Manufacturer of modular and stackable
LAN switching products extend the
usability and data capacity of existing
Ethernet LANs.
Jointly held company formed in 1993
by Bolt Beranek and Newman and UB
Networks offers enterprise ATM
switching, workgroup ATM switching,
LAN switching and routing.
Supplier of ISDN (Integrated Services
Digital Network) remote-access
networking products useful for
telecommuting and other networked
applications.
Developer of Internet gateway software
connecting central and remote office
desktop users with the Internet.
Inventor and leading supplier of Fast
Ethernet (100BaseT) and Ethernet
desktop switching products.
Manufacturer of cost-effective, low
maintenance network address
translation and enterprise Internet
firewall hardware and software.
Internet software products for
connecting disparate computer systems
over local area, enterprise-wide and
global computing networks including
the Internet.
Leading supplier of Asynchronous
Transfer Mode (ATM) and Frame Relay
high-speed wide areas network
switching equipment transporting a
wide variety of information, including
voice, data and video.
Modem ISDN Channel Aggregation
(MICA) technologies will deliver highdensity digital modem technology with
Ciscos dial-up and access product
lines.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 16

Exhibit 3 (contd)
Summary of Ciscos Acquisitions as of June 1997
Company

Date

Purchase Price

Nashoba Networks, Inc.

August 1996

$100 million

Granite Systems

September 1996

$220 million

Netsys Technologies

October 1996

$79 million

Metaplex, Inc.

December 1996

not public

Telesend

March 1997

not public

Skystone Systems Corp.

June 1997

$102 million

Global Internet Software

Group

June 1997

$40 million

Ardent Communications
Corp.

June 1997

$156 million

Source: Cisco Systems

Description
Token ring switching technologies
for providing users with a wide
choice of employing highperformance switched workgroup
and backbone Token Ring
environments.
Standards-based multilayer Gigabit
Ethernet switching technologies
for developing a wide choice of
backbone network technologies.
Network modeling and design
software intended to help common
customers base design and plan for
networks ideally suited to their
unique business requirements.
Specialist in network product
development in the enterprise
marketplace, gives customers the
ability to migrate from SNA to IP.
Specialist in wide area network
access products, gives
telecommunications carriers a
more cost-effective way to deliver
high-speed data services for
Internet and intranet access
applications.
Innovator of high-speed
Synchronous Optical
Networking/Synchronous Digital
Hierarch technology to carry
information to high-capacity
backbone networks, such as those
operated by telecommunications
carriers and ISPs.
GISG is a pioneer in the Windows
NT network security marketplace
with its Windows NT Centri
Firewall for small/medium
businesses.
Pioneer in designing combined
communications support for
compressed voice, LAN, data and
video traffic across Frame Relay
and ATM networks.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 17

Exhibit 4
Preliminary Ardent Term Sheet, June 1996
______________________________________________________________________________
Memorandum of Terms
For Private Placement of
Series A and Series B Preferred Stock of
Ardent Communications Corporation
June 14, 1996
______________________________________________________________________________
This memorandum summarizes the principal terms of the Series A and Series B Preferred Stock
financing of Ardent Communications Corporation.
Offering Terms
Issuer:

Securities to be Issued:

Price:

Ardent Communications Corporation, a California

corporation
(the "company")
3,000,000 shares of Series A Preferred Stock and
11,000,000 shares of Series B Preferred Stock
$.333 per share of Series A and
$1.00 per share of Series B

Terms of Series A and Series B

Preferred Stock
Dividends:

Annual $.03 and $.08 per share dividend,

respectively, payable when and if declared by
Board; dividends are not cumulative. For any other
dividends or distributions, Preferred Stock
participates with Common Stock on an as-converted
basis.

Liquidation Preference:

First pay cost plus accrued dividends on each share

of Preferred Stock. Thereafter Preferred and
Common share on as converted basis, until such
time as the Preferred Stock has received an
aggregate of two times cost, thereafter all proceeds
shall go to the Common Stock.
A merger, reorganization or other transaction in
which control of the company is transferred will be
treated as if a liquidation.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 18

Exhibit 4 (contd)
Preliminary Ardent Term Sheet, June 1996
Conversion:

Convertible into one share of Common Stock (subject to

antidilution adjustment) at any time at the option of the
holder.
Automatically converts into Common Stock upon
consummation of underwritten public offering with a
price of $5.00 and aggregate proceeds in excess of
$7,500,000.

Antidilution Adjustments:

Conversion ratio adjusted on narrow weighted average

basis in the event of a dilutive issuance. Proportional
adjustments for stock splits and stock dividends.

Voting Rights:

Votes on an as-converted basis, but also has series vote

as provided by law and on (i) the creation of any senior
or pari passu security, (ii) repurchase of Common Stock
except upon termination of employment, (iii) any
transaction in which control of the Company is
transferred, and (iv) any adverse change to the rights,
preferences and privileges of the Series A or Series B
Preferred.

Terms of Preferred Stock

Purchase Agreement
Representations and Warranties:
Standard representations and warranties by the Company.
Assignment of Inventions and
Confidentiality Agreement:

All employees and consultants shall enter into companys

standard form inventions and proprietary information
agreement.

Terms of Investor Rights

Agreement
Registration Rights:

(a) Beginning earlier of June 28, 2000 or six months

after initial registration, two demand registrations upon
initiation by holders of at least 30% of outstanding
Preferred Stock for aggregate proceeds in excess of
$10,000,000. Expenses paid by Company.
(b) Unlimited piggyback registration rights subject to pro
rata cutback at the underwriters discretion. Full cutback
upon IPO; 30% minimum inclusion thereafter. Expenses
paid by Company.

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 19

Exhibit 4 (contd)
Preliminary Ardent Term Sheet, June 1996
(c) Unlimited S-3 Registrations of at least
$1,000,000 each upon initiation by holders of 20%
of the Preferred. Expenses paid by Company.
Registration rights terminate (i) five years after
initial public offering or (ii) when all shares can be
sold under Rule 144, whichever occurs first.
No future registration rights may be granted without
consent of a majority of Investors unless
subordinate to Investors rights.
Right of First Refusal:

Cisco Systems shall have the right to purchase all

securities issued in subsequent equity financings of
the Company, provided the Option, as defined
below, has not expired.

Financial Information:

The Investors shall receive standard information

rights including audited financial reports, quarterly
unaudited financial reports, monthly unaudited
financial reports and annual budget and business
plan, as well as standard inspection rights.

Board of Directors:

Board shall consist of four members. Board

composition at Closing shall be Wu Fu Chen, Ed
Kozel, and Mike Goguen. One other representative
will be designated by a majority vote of the Series B
Preferred Stock.

Post-Closing Capitalization
Series A Preferred Stock Outstanding
Series B Preferred Stock Outstanding
Common Stock held by Founders
Common Stock Reserved for Employees
(however, an additional 3,750,000 shares
shall be available for grant after expiration
of Option, held by Cisco):
TOTAL :

3,000,000
9,000,000
6,250,000
3,000,000

shares
shares
shares
shares

12.9%
47.3%
26.9%
12.9%

23,250,000

shares

100.0%

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 20

Exhibit 4 (contd)
Preliminary Ardent Term Sheet, June 1996

Other Matters
Common Stock Vesting:

Restrictions on Common Stock Transfers:

Common Stock shall vest as follows: After twelve

months of employment, 25% will vest; the
remainder will vest monthly over the following 36
months. Repurchase option on unvested shares at
cost. No acceleration in the event of a Change of
Control, except for Mr. Chen, whose vesting shall
accelerate in the event of a Change of Control such
that at most one year of vesting shall remain.
(a) No transfers allowed prior to vesting.
(b) Right of first refusal on vested shares until
initial public offering.
(c) No transfers or sales permitted during lockup period of up to 180 days required by
underwriters in connection with stock
offerings by the Company.

Option:

Until the earlier of fifteen (15) months from the

Closing or three (3) months after First Customer
Shipment, Cisco shall have the right to acquire
either all of the outstanding equity securities of the
Company, or all of the Companys assets, in Ciscos
discretion, for a purchase price of $232,500,000
payable either in cash or equity securities of Cisco.

Closing Conditions:

Closing subject to negotiation of definitive legal

documents and completion of legal and financial
due diligence by Investors.

Source: Cisco Systems

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 21

Exhibit 5
Ardent Capitalization Table

Preferred A
Cisco
Sequioa Capital
Founders
Engineering Team
Total
Valuation ($/shr)
Valuation ($)
Cash Inflow

Preferred B

3,000,000
3,000,000
$
$
$

0.33
2,400,000
999,000

Acquisition Price

232,500,000

Return
Cisco
Venture Capital
Founders (5 employees)
Engineering Team (20 employees)

$
$
$
$

Cash Out
75,350,000
24,650,000
102,500,000
30,000,000

Return (Cash in - cash out) Per Employee

Founders
$
Engineering Team
$

20,299,550
1,499,850

Common

7,535,000
2,465,000
1,000,000
11,000,000
$
$
$

1.00
23,250,000
11,000,000

$
$
$
$

Cash In
7,535,000
2,465,000
1,002,250
3,000

6,250,000
3,000,000
9,250,000
$
$
$

0.001
23,250,000
9,250

Option to Acquire

Cisco Cost/Head
Cisco Cost

$
$

6,286,000
157,150,000

Conditions
- No accelerated vesting for employees or founders, except for Wu Fu Chen
- Commitment from Wu Fu to stay one year post acquisition
- Right to future offerings in the company or direction of those offerings

Source: Cisco Systems

Multiple
10.0
10.0
102.3
9,978

Total
7,535,000
2,465,000
10,250,000
3,000,000
23,250,000

Ownership
32%
11%
44%
13%
100%

Cisco Systems: A Novel Approach to Structuring Entrepreneurial Ventures EC-15

p. 22

Exhibit 6
Press Release for the Ardent Communications Acquisition

CISCO SYSTEMS TO ACQUIRE ARDENT COMMUNICATIONS CORP.

Further investment in Data, Voice and Video Integration For Public and Private Networks
SAN JOSE, Calif. June 24, 1997 Cisco Systems, Inc. today announced it has signed a
definitive agreement to acquire privately-held Ardent Communications Corp. Previously, Cisco
and Sequoia Capital held minority equity stakes in Ardent. San Jose-based Ardent is a pioneer in
designing combined communications support for compressed voice, LAN, data and video traffic
across public and private Frame Relay and ATM networks.
Under the terms of the acquisition agreement, shares of Cisco common stock worth
approximately $156 million will be exchanged for the outstanding shares and options of Ardent.
In connection with the acquisition, Cisco expects a one-time charge against after-tax earnings of
23 cents per share in the fourth fiscal quarter of 1997. The acquisition is expected to be
completed by late-July 1997 subject to various closing conditions, including clearance under the
Hart-Scott-Rodino Antitrust Improvements Act and Ardent shareholder approval.
CISCO STEPS UP INTEGRATION OVER FRAME RELAY AND ATM NETWORKS
With the continued pace of deregulation of the telecommunications service industry, carriers are
increasingly offering services which integrate communication channels of voice, video, and data.
As a result, the demand for low cost, easy to use, multiservice access products for new carrier
services is rapidly expanding. The acquisition of Ardent will complement Ciscos 3800 series
within carrier service offerings for branch offices and remote sites by extending leadership in
integration of voice, video and data. Based on Cisco IOS software, Ardents low cost platforms
will natively support multiservice traffic and implement voice compression using high
performance Digital Signal Processor (DSP) technology. Ardents early affiliation with Cisco has
resulted in a complementary product platform offering superior interoperability with existing
Cisco multiservice access and switching product lines.
ABOUT ARDENT COMMUNICATIONS
Ardent Communications was founded in 1996 by CEO Wu Fu Chen. Mr. Chen has co-founded
four other companies since 1986 including Cascade Communications and Arris Networks.
Ardents approximately 40 employees will remain in San Jose and become part of the
Multiservice Access Business Unit led by Vice President and General Manager Alex Mendez
within Ciscos Service Provider line of business.
Ardent Communications is on the leading edge of integrated access equipment design. Founded
in 1996, Ardent designs, manufacturers and distributes advanced access products for integrating
voice, video and data on public or private Frame Relay or ATM networks.
Source: Cisco Systems