Vinod Khosla

"I’m not a fan of the Cerent deal [with Cisco]. I discourage founders from trying to sell the company too early."

June 1, 2000

10 Min Read
Light Reading logo in a gray background | Light Reading

The Light Reading Interview

In The Spotlight: Vinod Khosla

Vinod Khosla Many VCs working within the networking industry have successful portfolios. None can match that of Vinod Khosla, a partner at Kleiner Perkins Caufield and Byers http://www.kpcb.com.

Trying to introduce an interview with the man is a guaranteed exercise in frustration (not enough superlatives). Khosla is, simply, the holy father of networking investments, the big kahuna of the optical networking industry -- not to mention no.1 on Light Reading’s Top Ten List of the most powerful movers and shakers in the optical networking industry (see The Top 10 Movers and Shakers in Optical Networking).

Khosla serves on the boards of Corvis, Juniper, Redback, and Qwest. He also brought Cerent and Siara into the world, not to mention having his hands on Cosine, Lightera, and ONI. Then there are the service providers: big names like Concentric, @home, and Rhythms.

Those are just some recent successes. Moving forward, he’s putting even more emphasis on optical networking investments, working with companies like Zaffire (which makes metro DWDM), Kymata (optical integrated circuits), Iolon (optical components), OnFiber (fiber optic services), and Broadband Office (a building CLEC, or BLEC) – as well as a few others that he won’t talk about.

So who is this quiet, modest man with a sense of humor?

Recently, in an exclusive interview, Light Reading sat down with "the great one" to find out.

He told us:

· What Vinod brings to the deal
· Why he’s not a fan of the Cerent deal with Cisco
· Whether startups should be allowed to give shares to their customers
· How he picks his investments
· How to forecast the optical future
· What his three least favorite words are
Light Reading : Vinod, you’re famous.

Vinod Khosla: [Laughs] I don’t know about that. I’m only as good as the market, and people can change their mind pretty quickly. One day they’re giving you credit for things you didn’t do. The next they’re taking it away for something you didn’t do.

Light Reading : You’ve been busy recently. Vinod Khosla: Oh, sure. We have the Corvis IPO coming up, and Cosine as well.

Light Reading : Do you know Light Reading?

Vinod Khosla: Never heard of it.

Light Reading: Oh. Well, let’s get started. How long have you been working at Kleiner Perkins?

Vinod Khosla: Since eight-six. Too long [Laughs]. Before that I was the founder of Sun. Kleiner-Perkins was an investor. That’s the connection.

Light Reading : Why did you leave Sun?

Vinod Khosla : Well, I didn’t expect to be working this hard. I thought this would be a part time job. I’m not complaining, of course. It’s a very fun job.

Light Reading : The hours are long, I would imagine.

Vinod Khosla: You know, they’re pretty reasonable – but intense.

Light Reading : Here’s a question a lot of entrepreneurs would like to know the answer to: how do you pick your startups?

Vinod Khosla : It’s very intuitive. There really isn’t a science to it. It’s all about understanding the market.
The big companies ask their customers what they want, and what they get in return is a list of the next increment of features, or their customers’ immediate set of needs. I don’t ask the customers what they want now. Instead, I try to guess what they will need two years from now. Customers usually have no clue what they will be in two years time because they’re more focused on their immediate set of problems, which is what they should be focused on.
I was just on the phone with someone from SBC. At the start of 1998 we presented him with this list of [Kleiner Perkins] companies, and solutions: Corvis for the backbone, Lightera, which is now the Ciena CoreDirector, Juniper for routing, Cerent for metro networks, Siara for edge networks, and ONI for metro area optical DWDM. Two years ago he said ‘we’re not worried about those requirements.’ Now they think they should have planned their networks with those requirements in mind.
So I tend to envision the new environment. Then anticipate it, and stay ahead of the curve.
Another thing: look at that list of companies. Each one comprises a piece of the end-to-end solution, or a piece of the jigsaw. That’s another part of what I do, to take a holistic view of the network, and try to deliver the different pieces.

Light Reading : How successful is the Ciena CoreDirector product?

Vinod Khosla: It’s really the only product that’s selling at Ciena. It’s their lead product. I just talked to Pat Nettles and he acknowledged that it was doing really well. To be honest it’s the only significantly differentiated product in their whole line.

Light Reading : Have you always been this successful?

Vinod Khosla: Frankly, because I do my job this way I often get egg on my face. I don’t get embarrassed about that. I probably screw up as much as anyone else.

Light Reading : Example?

Vinod Khosla : “Go.” They were building a Pen computer in the ‘80s.

Light Reading : Well that was a while back. What have you messed up recently?

Vinod Khosla : [Long pause] Ok, I guess I didn’t do anything really bad in the nineties. Not when I was in charge, anyway. But next time I screw up my reputation will change. I recognize it.

Light Reading : What does Vinod bring to the deal?

Vinod Khosla : By and large investors don’t want to do anything but buy low and sell high. That’s why I don’t consider myself an investor. I actually don’t like the term venture capitalist. I consider myself a venture assistant.

Light Reading : How long do you stay involved with your companies?

Vinod Khosla : I view everything as a five-year plan. Juniper is the perfect example of the way that I like to work with a company.

Light Reading : Not Cerent? [Ed.note: Cerent was sold to Cisco last year for $7 billion]

Vinod Khosla : I’m not a fan of the Cerent deal. I discourage founders from trying to sell the company too early. I’m not into the quick turnaround. Words I dislike are “deals,” “transaction,” and “exit strategy.” These are just things I hate. When a client starts pitching me on when they are going to IPO that’s when I stop listening.

Light Reading : You must be rich. Really rich. Why do you still do this?

Vinod Khosla : It comes down to whether you have a passion for a vision, or just want to make money. Everything I do is about the about the vision. I build companies that way, and I build teams that way. There’s a distinct difference between company building and investing. You won’t find me worried about dilution or giving away too many stock options -- things like that. I’ll do whatever it takes to build the company.

Light Reading : How many business plans do you get in a month?

Vinod Khosla : Too many. Way too many. Hundreds.

Light Reading : How many make the cut? Vinod Khosla: Less than one in a hundred. I don’t invest every month. Light Reading : Is the business plan an important factor in which companies you back? Vinod Khosla : The quality of the people is a bigger determinant than the plan.

Light Reading : Do they have to have a track record of being at successful startups?

Vinod Khosla : It doesn’t have to be a track record of startups. It could be a management track record. Or great intuition. Like the Excite guys. I liked their thinking, and they were just out of school.

Light Reading : But there is no secret to getting your attention?

Vinod Khosla : No, it’s judgment, on my part. That’s why what I do is not scaleable. That’s why you don’t see me adding 100 people to the firm to replicate what I do.

Light Reading : How big is your fund?

Vinod Khosla : We don’t talk about that.

Light Reading : How many new optical networking startups is Kleiner Perkins funding?

Vinod Khosla : I prefer to call them networking startups. We have about a dozen altogether. Six are fairly recent, including Zaffire. And a few others that haven’t been announced yet.

Light Reading : You seem to be doing a lot of work with component vendors.

Vinod Khosla: Both Kymata and Iolon are component vendors. We haven’t announced either of them yet. We have another one also.

Light Reading : Let’s suppose that someone came to you and said they could build the world’s first, true photonic router -- one that routes packets in light, rather than converting them to electrical signals first. Would you believe it was even possible? And would you invest?

Vinod Khosla : That’s certainly difficult to do. But I’m an optimist. If they had the right people I would absolutely fund them. If they had experience of building commercial products, and came and in and pitched it, I would take the de facto view that they can do it – then try to convince myself otherwise.

Light Reading : How common is it for VCs to get involved with setting up deals between carriers and the equipment startups that they are funding?

Vinod Khosla : We help. That’s what I mean by venture assistance.

Light Reading : You make an introduction?

Vinod Khosla : Yes, that’s all we can do. We never try to cross the line of forcing someone to do something that is not in their interest.

Light Reading : Some people say that some of these deals are little more than marketing tools, that they are just done for the publicity.

Vinod Khosla: How do you mean?

Light Reading: An example would be the deal between Siara and Broadband Office. [Ed.note: Siara announced a contract to sell equipment to Broadband Office last year, at a time when it did not actually have a product]. That wasn’t a real deal, was it? Vinod Khosla: That’s not true. There’s generally a pretty solid deal behind these things. Some of them are flaky, but not the Siara deal with BBO. Look at the deal between Juniper and Uunet. That was announced eighteen months before the product shipped. And that has become the single largest chunk of revenue for Juniper. And these sorts of deals are great for the [carrier] customer. The equipment vendor has to meet certain specs specified by the customer [in order to get the contract]. And the customer takes no risk, because if the product doesn’t get delivered exactly to the spec, they don’t pay for it.

Light Reading : What about the trend for startups to give equity to individuals at their carrier customers? Is that ethical?

Vinod Khosla : I’m sorry?

Light Reading : The sort of deal that Matt Bross had with ONI [Ed.note: Matt Bross is CTO of Williams Communications, a customer of ONI Systems. Bross owns 322,460 shares of ONI's stock -- see Williams' CTO Profits From His Position ].

Vinod Khosla: From my point of view it is ethical. It really depends on how a company handles it. From an equipment vendors’ point of view, it really helps them to get input from someone who’s in the line of fire -- it helps build the right product. It’s really important.
But there are two important things here. The first important question is who is making the purchasing decision [at the carrier]? The person making the decision should not be the one getting the equity.
Second, it all has to be above board, which means every one should have full disclosure. You have to do it very carefully, and with caution, and most importantly with full disclosure.

Light Reading : You have a sense of humor. Do you think that’s unusual in the VC world?

Vinod Khosla : I hope not. I would not say I am somehow unique. I just have a blast. I guess I don’t take work too seriously. It’s a hobby -- not a job, or career.

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like