[December 1982] The story of Vinod Khosla and Computervision ...

Date view Thread view Subject view Author view

From: Adam Rifkin (adam@KnowNow.com)
Date: Sat Nov 25 2000 - 21:55:04 PST

Man, my background is too technical. I *just* discovered the canonical
Harvard Business School case study of Vinod Khosla and Computervision:


Go back to 1982 when you read the case study (below). So many questions
to answer...
  1. Does Vinod Khosla have a real chance at changing Computervision's decision?
  2. Does Sun have a better product?
  3. Can Sun be regarded as a reliable long-term vendor?
  4. How should Vinod Khosla respond to Computervision's president?
  5. What should his long-term and short-term objectives be?
  6. Specifically, what should he offer Computervision?

Ok, start reading...

> One area in which Sun did not make much progress in its first year was
> the OEM market. Sun considered gaining share in the OEM market to be a
> major, long- term goal and adapted its product and pricing strategies
> accordingly.
> One fundamental part of our strategy is to get as many CPUs out there as
> possible, and get them popularized. Market share is very important to us
> and revenue is not. Even though we did really good in revenue last year,
> our goal was not revenue.
> So if somebody is better off buying our CPU and putting his own disk
> drives and tape drives, because they can't afford ours, we will let
> them. We have created sort of an airline pricing structure. People who
> want full support and service and don't want to hassle with it pay full
> price. People who are sort of willing to book two weeks in advance and
> sort of leave on Thursday evenings get a price break. Everybody puts in
> their level of pain to get their price point. Our goal is to skim as
> much of the money and also get as many of our CPUs and operating systems
> out there as we can. We can capture the rest of the system revenue
> later. And if our co-developers--all these other guys who put in the
> other 70 or 80 engineers in various companies--get a lot of revenue,
> they will invest more in Sun.
> In spite of Sun's efforts, and even with the endorsement of computer
> science departments, OEMs had proven a very tough nut to crack.
> Daisy is the leader in its market segment, and Computervision and
> Integraph are leaders in their market segments. These three have decided
> that they are going to build their own computers. Calma, Autorol, and
> Mentor Graphics are the second tier, but the most important second tier
> after the top three, and Apollo has them all. Calma, Autorol, and Mentor
> account for 60% to 70% of Apollo sales. Mentor, I think, accounts for
> 40% of Apollo and is really catching up on Daisy, and its growth has
> been very good for Apollo. Apollo has been going through the roof
> because of these customers. And because of that, all the smaller people
> in the business are picking Apollo.
> Apollo is the successful platform. They have this huge momentum. They
> have $20 million or $30 million of financing versus our $4 million. They
> have well-connected senior management. We have diddly squat. We clearly
> have the better technology, and people recognize it, but Apollo's
> customers were, apparently, satisfied customers. As Mentor Graphics,
> Apollo's first OEM customer noted in its prospectus.
> The use of Apollo hardware enables the company to concentrate on
> applications software and special- purpose hardware development, while
> benefiting from Apollo's research and development efforts in general-
> purpose hardware and operating systems software. The company believes
> this strategy enables it to stay at the leading edge of technology in
> computer hardware, operating systems, and applications software. In
> addition, the company has benefited from market acceptance of the Apollo
> computer for general applications, as well as rapid improvements in the
> price and performance of Apollo's hardware.
> As Datamation had noted: "Everything seems to be going like clockwork
> for young Apollo; the original business plan [which] called for launch
> in 1980, sales in 1981 profits in 1982, and a public offering in 1983
> [is] right on schedule."
> In December 1982, Sun saw an opening into the OEM business.
> Computervision announced that it was considering buying a workstation
> instead of building its own.
> Computervision was a company in transition. From the time it was founded
> in 1969, it had followed an ambitious strategy. It offered complete
> "turnkey systems"- - all the hardware, software, and training that users
> needed. Instead of designing specialized systems for one or two
> industries, Computervision aimed for a wide range of domestic and
> international customers. And unlike most of its competitors, who added
> software to Digital's or IBM's minicomputers, Computervision designed
> and built its own.
> In the seventies, according to a New England Business article,
> "Computervision was a high- tech company on a roll. Everything was in
> superlatives. It was the fastest growth, the newest technology, the
> highest income."
> Although Computervision was still the market leader, the CAD pioneer's
> position had begun to erode by 1983. IBM, Digital, Prime, Perkin Elmer
> and several other minicomputer manufacturers had entered the market with
> software from third party vendors, and they enjoying some success. IBM
> had become the number two player, almost overnight, with the help of
> Lockheed's popular software. And the effectiveness of Computervision's
> response had been undermined by delays in the development of its next
> generation (32- bit) minicomputer.
> In addition, dozens of start- up companies were gnawing at low end of
> the market with cheap workstation- based systems aimed at specific
> niches. Using hardware from companies like Apollo and Sun, these start-
> up companies were able to sell CAD systems for under $100,000, compared
> to the $500,000 typical of Computervision's systems.
> Computervision was taking several actions to shore up its position. In
> August 1982, founder Martin Allen gave up the titles of president and
> CEO to James Berret, formerly a group vice president at Honeywell. Allen
> said that he had "wanted someone who came from a multimillion dollar
> environment and who had experience competing against IBM."
> The company also seem to be edging away from its commitment to vertical
> integration: it had bought Cambridge Interactive Systems, which sold CAD
> software to go with other vendors' hardware; it was negotiating to buy
> IBM's System 4300 computer line, as a backup to its own 32- bit
> minicomputer development; and, to respond to its new low- end rivals, it
> was planning to buy a workstation platform instead of developing one.
> To gear up for the Computervision contract, Sun hired a veteran
> executive and group operations manager, Owen Brown, away from Digital.
> We needed credibility. The East Coast establishment didn't take me or
> Sun too seriously. They thought we were quirky . . . because of things
> like . . . I hated smokers, and for the first 20 people we didn't hire a
> smoker or anyone who drank coffee.
> Besides there was a real crunch on my time. I was chief salesman, CEO,
> and everything. Engineering was repor ing to me. Finance was reporting
> to me. So we asked Owen to be VP of sales and marketing. He said, "Give
> me the title of President so I can deal with these Computervision guys
> and do big corporate account selling and go play golf with them.'' I
> didn't play golf, and in some situations it is very important. So we
> hired him and formed an Office of the President.
> In the months that followed, competition for the Computervision order
> narrowed to Apollo and Sun. "We had our friends in Computervision," said
> Khosla, "who were mostly the technologists, and Apollo had theirs, who
> were mostly the business people and all the senior executives. And there
> were some people who wanted Computervision not to go outsite at all."
> At 4:00 one afternoon in July, Khosla received an ominous call.
> It was from a purchasing guy from Computervision and that was
> terrible. You never want to hear from a purchasing guy, because that
> means you are getting a rejection. He thanked us very much for bidding
> on the contract but said they had chosen another vendor.
> I was livid. Had smoke coming out of my ears. Owen Brown, who was
> supposed to be running that account, was off for two weeks on his Navy
> reserve training. How could he be off on Navy reserve training or
> anywhere else when the future of the company depended on this one deal
> and he didn't even know the deal is happening in this two- week period?
> I took over. By 6:00 p.m. I had sent off a letter, Federal Express, to
> about 30 or 40 people at Computervision. I said we would do anything for
> their business.
> I didn't go home; I had my wife bring my clothes to the office, and I
> caught a red- eye to Boston.
> The next morning, I was in the Computervision lobby, making phone calls,
> trying to see someone. Nobody would talk to me.
> After four hours and over 50 phone calls, I got through to the VP of
> sales and marketing, who was a Sun friend. Apparently, Computervision
> had made the decision long ago. By now, there were Computervision
> technical people from Europe who had arrived here for technical
> training, at Apollo. It had gone that far. They were on crossing the Ts
> and dotting the Is on the contract. So people were avoiding me for
> political reasons.
> The sales VP said that, "If you go back to the Sun office in Boston, I
> will make sure somebody calls you and talks to you." With that promise,
> since I wasn't getting anywhere, I went back to the Sun office there. I
> showered and shaved there and waited.
> Finally, the president called.
> He was saying, "'We have decided, and here is why." He was giving me all
> the good reasons. "You are a 40- person company, and you have an
> incomplete product. We love your technology, but there is no way you can
> supply it. Apollo is the standard in the industry, well financed and
> well managed."

Well, history has shown that Khosla figured out how to steal that deal
back from Apollo, thereby saving Sun Microsystems:


According to legend, he got the deal by sitting in Computervision's
lobby and not leaving until they signed the deal. Wow.

It's nice to hear that he believes he owes his success to wooing Lady
Luck, even though he had the requisite skill and tenacity needed to succeed...

> The cofounder and first CEO of Sun Microsystems, Vinod Khosla, once told
> a Churchill Club audience a story of his early days at Sun that I will
> never forget. One fine (presumably sunny) day, Vinod was busy working at
> his desk when an employee walked into his office and said something very
> simple like, "Gee, Vinod, what are you going to do for an encore?" It
> was actually quite a good question. Not only was Sun on a very steep
> sales ramp at the time, but this was Vinod's second successful high-tech
> startup--and he was not even 30 yet. The question began to haunt Vinod,
> ultimately sparking a yearlong depression. He started to wonder whether
> he would ever be able to achieve anything greater than he had with
> Sun. But before I get to the lesson of this tale, allow me to tell
> Vinod's whole story. I think it's worth hearing.
> While Vinod was growing up in modest circumstances in India, he
> habitually scoured the pages of Electronic News, dreaming of the day he
> could follow in the footsteps of heroes like Intel cofounder Bob Noyce
> and start his own Silicon Valley company. His first chance came after he
> arrived at Stanford Business School, when a local venture capitalist
> solicited his help researching and writing the business plan for a new
> company called Daisy Systems. By the time the ink dried on the plan,
> Vinod had been enshrined as the company's first chief financial
> officer. Daisy, a design automation hardware and software vendor founded
> by Aryeh Finegold (currently chairman of Mercury Interactive),
> eventually went public. In the end, after a debate with Aryeh over the
> virtues of open systems, Vinod took his seven-figure booty and left to
> start Sun at the age of 27.
> Sun was not as easy to launch as one might think, Vinod revealed in his
> speech. In 1982, when the company was founded, there were at least five
> other startups focused on building single-board Unix computers. Vinod
> also talked about "Black Monday," when Sun's venture funding temporarily
> fell through, and the day he refused to leave the lobby of Sun's biggest
> customer, Computervision, until it signed a company-saving deal.
> But by the time the employee had arrived in Vinod's office two years
> after Sun's founding, the company was skyrocketing. Vinod turned the
> reins over to Scott McNealy and took a year off to ponder his new
> burden. After learning everything from hang gliding to windsurfing,
> Vinod awoke one day and found himself at peace. What finally occurred to
> him was that the biggest reason for his success was luck. While he was
> indeed a bright and tenacious entrepreneur, circumstances beyond his
> control could just as easily have caused him to fail.
> The point of this story? Don't take life so seriously. If you're a fair
> and reasonable person, good things will generally follow. And luck did
> continue to follow Vinod: he and his lifelong sweetheart, Neeru, started
> a great family, and he's been a partner with Kleiner Perkins Caufield &
> Byers, Sun's original backers, ever since.

It's interesting that the story of Sun and Computervision is so
canonical that business schools use it as required reading:


Hey, maybe we should go to this...

> Session 14: February 22, 2001
> Real-Time Decision Making
> Quote of the Day
> "Never, never, never, never give up." ~ Winston Churchill
> Summary
> In addition to a compelling vision and an effective marketing
> strategy, start-up leaders must excel at real-time decision making in
> order to manage the growth of hot high technology start-ups. A crisis
> at Sun in its early days is used to illustrate this. The case details
> how Vinod Khosla got Sun Microsystems started, and then describes a
> critical marketing challenge the company faced at the end of the first
> year. Our objective is to illustrate how bold creative thinking and
> leadership can make or break start-ups.
> Required Readings
> Case Study: Vinod Khosla & Sun Microsystems, Inc.
> Article: "A Quiet Man Puts Some Sizzle In Latest Deal Involving
> Cisco," [WSJ Interactive, August 27, 1999]
> Article: "Starting Up in High Gear, An Interview with Vinod Khosla",
> HBR July-August 2000, #R00403
> Skim the following readings:
> Geoff Moore, Inside the Tornado
> Study Questions
> 1. Does Vinod have a real chance at changing Computervision's
> decision? Does Sun have a better product? Can Sun be regarded as a
> reliable long-term player?
> 2. How should Vinod respond to Computervision's president? What should
> his long-term and short-term objectives be? Specifically, what should
> he offer Computervision?
> 3. What are the consequences of not doing the deal?
> Team Case Analysis
> 1. Computervision (has/has not -- choose only one) made a mistake in
> selecting Apollo because ...
> 2. As Vinod, my offer, if any, to Berret (the president of
> Computervision) will be to ...

Here's one sample answer...


> The Computervision order presents Sun Microsystems with a number of
> problems for which there is no immediate solution. The order is
> perceived to be too large for Sun to handle, the product is regarded as
> incomplete and the company has no established reputation. Sun's major
> competitor, Apollo, provides the industry standard, is reputable and has
> ample funding to be able to service Computervision's order.
> In order to win the contract with Computervision, Sun Microsystems must
> convince Computervision of the superiority of their system as well as of
> their abilities to fulfil the order, within a single phone call. These
> are the points that should be addressed.
> 1. Explain the superior processing power of Sun compared to the competitors
> 2. Higher gross margin available than if dealing with Apollo. Sun is
> considerably cheaper which will enable Computervision to sell their
> products on at the same price as Apollo.
> 3. Open Systems are the way forward, as opposed to the proprietary
> systems of their competitors
> 4. Plans of imminent staff expansion
> 5. Product is complete, use CADLINC example
> 6. That Sun has managed to achieve additional financing for this
> expansion. Vinod should employ "Ham & Egging" tactics.
> 7. That Sun Microsystems is willing to enter into a formal agreement
> with Computervision, to ensure an ongoing relationship.
> Assuming that Vinod managed to convince Computervision to accept the Sun
> bid, he must then immediately address the following issues.
> Sun needs a dedicated marketing and sales force. Until now Sun has been
> receiving orders and inquiries without any active
> marketing. Capitalising on their strengths, Sun should make a determined
> attack on the OEM market.
> Vinod must also ensure that Sun attains the capabilities to service the
> Computervision order. He will need to enact the second part of the "Ham
> & Egging" process in order to obtain the necessary financing so as to be
> able to increase his manufacturing capacity. Vinod must not waste time
> as Sun's reputation within the market place is at stake, before it
> has even become established.
> If Vinod experiences any form of delay in the whole procedure of
> expanding his capacity he should investigate the outsourcing
> opportunities in Silicon Valley, enabling Sun to complete the order.

[Do these people realize there was no such thing as outsourcing in 1982? :]


The reason behind KPCB's success is a focus on great management teams and breakthrough technologies and using them to build durable companies instead of a Wall Street-like focus on buying and selling stock. -- Vinod Khosla, http://www.sipa.org/sig/ventcap/msgboard/archive/0134.html

Most good business plans have a focus on markets and people. The fundamental economics make so much sense that detailed financial analysis is generally unnecessary. -- ibid

Most entrepreneurs aren't clear about their own goals. It is important to know whether goal is fame, fortune, family business or working with friends. Each of those goals will result in a different strategy for the company, and they are all acceptable goals. -- ibid

Lack of an M.B.A. has seldom, in my view, been a handicap. Once the entrepreneurial idea and energy is present, a complementary team can be built by the entrepreneur that includes M.B.A.'s, accountants, marketers and other needed functions. -- ibid

I believe the success factors [for a startup] are people, a business model that includes leverage, a strategy to manage risk, paranoia and persistence, and an inclination towards experimentation. -- ibid

Start with the key messages. Understand the emotions of your audience and the emotional path you want them to travel. It is more important to be coherent -- address their emotions and answer their questions -- than to be 100% complete in your presentation of facts. -- ibid

Good management is all about a clear vision. Keeping a consistent culture and keeping the new employees motivated are small parts of the total management equation. -- ibid

I personally believe vision is more important than management skills. As long as the entrepreneur is smart enough to surround himself with good managers. -- ibid

The most difficult part [of starting Sun] was not knowing what I didn't know and gaining credibility with other "normal" businesspeople. -- ibid

[With Excite] I funded a couple of entrepreneurs whose idea I did not like, but where I liked the people. They had no previous work experience. We incubated them, helping them develop their business plan, turned down some acquisition offers along the way that looked very attractive at that time. We continued to focus on building a long-term, sustainable business advantage and eventually announced a merger with AtHome at about a $7 billion market cap in the space of a few years. -- ibid

Valuation is very much a gut call. It depends on the people, it depends on the unfair advantage a start-up has in its market, and it depends on the size and dynamics of that market. -- ibid

Generally, the industries of greatest interest to me are technology segments that are undergoing rapid change. -- ibid

I do believe luck plays a big role in one's life. Somebody in the early days of Sun had a sign in his office saying, "Success is when opportunity meets preparation." Which means that you have to be doing the right things, approaching it the right way and having the right people. And then, some things have to just break your way. You have to be prepared and then you have to get the lucky breaks for it to really work. So it's a combination of smart efforts, preparation, and then luck. Anybody who denies this is not being honest. -- Vinod Khosla, http://www.siliconindia.com/magazine/Aug97vent.html

When people ask me what I do best, I say, charm Lady Luck. -- ibid

95% of the people I meet are not willing to make the compromises necessary to really get great people into a venture. They'll want a better financial deal or wouldn't want to give in too much to get a great person involved. That's the single biggest mistake I see -- not appreciating the value of really good people when it comes to the nuts and bolts -- what do you pay him, how much you compromise on what you want to do, your job, your title. -- ibid

I love the notion of building something that will be around 10 or 20 years from now. -- ibid

My goal is to be the best assistant there is for anybody trying to build a large technology oriented company. And I spend less than 10% of my time evaluating new deals. I spend less that 1% of my time looking at the financial aspects of a deal. I spend 90 percent of my time helping companies after I have made the investment -- in strategy, in tactics, in transactions, in relationships, in recruiting, in implementation of something -- in the process of building a company. One of my CEOs calls me his VP of Business Development. And I really like that, because in some sense, its much more appropriate to what I like to do for a company, than being called a board member, or something like that. -- ibid

We spend a lot of time discussing every mistake. And that's something that should be extremely valuable. In fact it is far more valuable to entrepreneurs than any money we bring to the company. -- ibid

I think that though the Web has had an enormous impact, we have not started to see the real impact it will have on business and business economics, yet. -- ibid

Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Sat Nov 25 2000 - 22:00:12 PST