If you're looking for somebody who knows what a real technology cycle looks like, Dick Kramlich is the guy.



Kramlich, widely regarded as one of the classier characters in Silicon Valley, has been investing in technology for nearly 40 years.

Kramlich is certainly among the elite VCs, having chalked up multiple big deals three decades running. After investing for nearly a decade with partner Arthur Rock, Kramlich became a founder of New Enterprise Associates (NEA) in 1979, just in time for the 1980s technology boom, where he’s also a general partner. One of NEA’s first big deals was with the Ethernet company known as 3Com Corp. (Nasdaq: COMS), which is still public, though it’s a vastly different company now than it was since going public in the early 90s.

Bob Metcalfe, a general partner at Polaris Ventures and one of the founders of 3Com, says meeting Kramlich, whom he calls a “real gentleman,” changed his life.

”One of the best things that ever happened to me was meeting Dick Kramlich in 1980,” says Metcalfe. “He had just founded NEA, and I was looking for my first round of venture capital for 3Com. He invested and joined my board, on which he served through our IPO. A sweetheart. An entrepreneur's VC.”

Metcalfe also let us into one of Kramlich’s deep investment secrets: Screen the last names of potential founders carefully.”The first thing Dick told me when I asked what his criteria were for investing was that he never invested in a company if the CEO's last name was a fruit -- he had lost a lot of money on those,” says Metcalfe. "Whew, I was over the first hurdle."

Later on, amidst the monster networking deals in the nineties, NEA invested heavily in Juniper Networks Inc. (Nasdaq: JNPR), and Kramlich spent many years on the board of Juniper, departing last fall. In total, NEA has taken 147 companies through IPOs and had 193 companies acquired.

In our interview, we learned that Kramlich’s a straight-shooter, and he’s sometimes prone to large eruptions of laughter. It’s a refreshing switch from the over-earnest and humorless VCs that seem to proliferate venture partnerships these days.

Right now, Kramlich says NEA is practicing its own brand of venture capital “with scale,” in which it places very large bets on companies it thinks can be world-changers. These new deals are reflective of the current climate, in which large bets need to be made in order to make money, says Kramlich. He believes companies need scale and visibility to compete, and customers must be assured that their supplier is not some small fry that’s going to disappear overnight. Hence, NEA’s sizable investment in two companies familiar to Light Readers: Force10 Networks Inc., the 10-Gbit/s networking company, and Vonage Holdings Corp., the VOIP service provider -- both of which have raised hundreds of millions of dollars in capital (and counting). The size of these funding rounds -- combined with the fact that neither has met with a classic “exit” in the form of an IPO or acquisition –- has also made these two companies controversial.

We caught up with Kramlich in May in NEA’s 80s-style office, perched on the hills over Stanford University. He had just been featured in a Sunday New York Times article on a shakeout in the venture capital industry, portrayed as one of the experienced survivors. Kramlich refers to the reset of the venture capital industry after the Nasdaq crash as more violent than the Great Depression.

Read on, as we cram with Kramlich on the following topics:

— R. Scott Raynovich, US Editor, Light Reading

Light Reading: I see you were just featured in the Sunday NYT recently about surviving a VC shakeout. You must be a real pro.

Kramlich:That doesn’t mean you are a pro!

LR: How long have you been doing this?

Kramlich: I came to Silicon Valley in the mid '60s. I worked with a partner, Arthur Rock, from 1969 to 1977. Then [I helped found] NEA in 1978.

LR: Tell us about this article in the New York Times. It’s about the VC industry having its own bubble.Kramlich: I’ll say. It made the depression look like child’s play. Top to bottom, the venture capital indices went down 90 percent [Kramlich says that none of NEA’s funds, to date, have ever lost money, and they have had an average return on capital of more than 20 percent]. Today, our industry is a $20 billion industry and it’s in balance. I think there are now about 7,000 to 8,000 companies, about which half don’t have an economic reason for being. A lot of the overhang comes in supporting those companies. In the last year, we’ve been getting a lot more optimistic as a business and a lot more competitive in deploying capital. And there are some good companies being started.

LR: Will we see more people drop out?

Kramlich: I think we’re mostly through most of that.

LR: Let’s talk about Force10 Networks and the 10-Gigabit Ethernet market. Tell us how it’s going.

Kramlich: We are seeing very rapid adoption. The question is: What are the drivers of this market? It started with high-performance computing centers. It goes from that to data centers and clustering. Those are robust markets. Those are some areas where we find the economics to be compelling. The team at Force10 is exceptional. They are honed by a lot of real world experience and they have the technical skill set to build a powerful array of products.

I was fortunate to be involved in Juniper Networks. I left the board a year ago in October. There are some parallels [with Force10] that I find intriguing. In Juniper, that team was lead by Pradeep [Pradeep Sindhu, one of Juniper founders and it’s current CTO] and Pete Wexler [Juniper’s former VP of engineering] did a terrific job. And then, there was a second generation of ASICs that were done. This team at Force10, which is run by a gentleman named Sachi Sambandan [Force10’s VP of engineering], they have done equally well. They are fantastic, in my judgement.

LR: Sachi is a quiet guy, we don’t hear much about him. Where did he come from?

Kramlich: I helped recruit him. He came from Cisco and before that, Intel. He is just an exceptional guy.

LR: Aren’t you worried about Cisco driving down pricing?Kramlich: Oh yeah, I’m terrified of it. I’ve been through it at least twice before. I was involved in Ascend Communications, which was run by Mory Ejabat, and after that Juniper, which was very direct and competitive with Cisco. In the Juniper case, they entered the game with a purpose-built system based on ASIC designs and they were able to focus on one sector of the market and migrate to the top carriers.

In this case, it’s a different set of customers, although we are not ignoring carriers, because they will be adopting it and we will have customers to announce. But I think it’s really an enterprise-based for the most part.

LR: So, the defense against Cisco is ASICs?Kramlich: That’s certainly powerful. You have to develop the software and the hardware and the features. At Force10, we started shipping product in 2002. We did $3 million then and last year we did $27 million. This year we should do upwards of $60 million. So we have direct sales and partnerships and we have more than 125 customers, so it’s really broadly based. I can’t release customers, but I can tell you they are the most demanding customers in the IT world.

LR: Right, haven’t you said publicly that one of your customers is Google?Kramlich: I don’t know if we’ve said it or not, but it’s definitely true –- talk about a demanding customer.

LR: Have you seen the Google data center?

Kramlich: I have not seen it, but recently I went to Huawei. But we can talk about that later.

LR: Okay. So, the skeptics on Force10 say that so much money went into Force10 that it’s going to be hard to make money on the investment. What's your response?

Kramlich: That’s a fair question. But, I would ask, what does it take to get into this business today? I think it’s a result of a purposeful development of a family of products. In earlier generations, you could approach in a more serial manner. Today, the big concern of the customers is whether the company will be around. These are multi-generational risks. So in order to do that you have to have not only financial stability in the company, you need a product line that transcends from one generation to the other while optimizing the architecture you put in place with software and feature sets.

So, there is a sizable community of interest from the vendor to the customer and that has to be extremely candid and well thought-out on both parties' parts in order to have something that’s going to prosper for a long period of time. In order to do this, it requires a lot of preparation and simultaneous development, which costs money.

LR: So are you saying as long as you keep Google happy, you’ll be fine?

Kramlich: [Laughing] I didn’t say that, but I would respect that. I said we have over 125 customers. And they, including Google, are among the most technically demanding in the world. So you have to have a broad product line, you have to have a series of features sets, good today and better tomorrow, and you have to have an international services organization. Couple that with partners who can incorporate it within a sensible way with their own product line. So you have all those demands, which you don’t do on a shoestring.

Embedded in that question is the assumption that there will be enough of a victory to share all around, investors and the company. I’m assuming all that can happen. I’m not predicting that.

LR: So these kind of projects are harder now?Kramlich: Tremendously harder. The easier part is there hasn’t been a whole tribe of imitators. Even though we have a lot of intellectual property, it’s a high hurdle. That part is better. The part that’s harder is that it has taken a lot of money to so this and it begs the question, is there a payoff at the end of the day. Thirdly, on the part of the customers, there is a question on their part of whether or not you are going to be around. We have to answer all those questions. I’ve been on sales calls where they ask that question. LR: So you think this is a company that you think can make it to IPO?

Kramlich: I really do. I believe we have multiple options to pursue on the way to liquidity. We are guardedly confident.

LR: Let’s talk about Juniper now. You were an early investor in the company, and you were on the board. Was there a particular reason why you left the board?

Kramlich: No, my time had come. We had distributed our shares and we had gone back into the marketplace during the dark days and bought more shares and then you go through the reporting process and we did. We began to lighten up, we distributed those shares. We’re not in the business of managing public capital so it was time to get out of there.

LR: Yes, you are talking about the story we did when you went in after 9/11 (see NEA Turns to Trading) and bought Juniper shares on the open market. That’s quite an unusual thing for a VC to do. Is that a point of pride for you?

Kramlich: The SEC gave a window of opportunity where they said you don’t have to abide by the [insider trading] windows. So we felt that allowed us to go in and legally acquire shares at what we thought were attractive prices. So we said, yep, we will be part of that initiative to restore confidence in our economy. So we stepped in and bought those.

LR: What was the price?Kramlich: I think we started at $4 and $5. We started buying in the low single digits and ended in the low [teens]. And we distributed at $15 to $20.LR: $4! Wow, that’s a good deal.

Kramlich: [Laughing] It was a good deal. That was part of the program with our limited partners. What we try to do is destroy the J-curve -- do you know what the J-curve is?LR: No idea.

Kramlich The J-curve is when you start an [investment] partnership, you start drawing down capital, you have some expenses, and then you have losses, because lemons often ripen early. So what you have to do is work your way back up and get over the J-curve and you’re in the black.

The point is that tactics like I just described after 9/11, it helps out when a quality company is caught in the downdraft like everybody else. We were encouraged to do something and it made a difference. That doesn’t make a dent in the long term, but it gets you on the right track early.

So, anyhow, Juniper is a great company and I was proud to have been a small part of that. It was terrific for NEA. Two of their founders are limited partners of ours, Scott Kriens and Pradeep Sindhu.

LR: Now, Juniper seems to be entering the next phase of warfare with Cisco.Kramlich: They certainly do. We happened to be involved with a company Netopia, we merged the company into Netscreen, so we ended up with Netscreen. That was an interesting strategic move on the part of Juniper because it allowed them to get into enterprise. I’ve always felt, knowing what I know, that Juniper had pound-for-pound the best product line in the space, but they were outmarketed and outsold by Cisco who got a much earlier start and Cisco does a terrific job. But I’ve always felt that if they kept doing what they were doing that you’re going to gain market share over time, that’s what they’ve done.

LR: But to go from saying “We are the service provider company” to “By the way, we’re now an enterprise company, too” -- isn’t that a tricky marketing proposition?

Kramlich: You know, they already made the transition from ISVs to carriers, that was a pretty big transition. That’s what the world didn’t see much of.

LR: You mean because they were selling through third parties and now they’re selling more direct?

Kramlich: Yes. When the company went public, they had two of the top 25 carriers. As of 2004, they had 24 of the top 25. So they made a huge transition in that period that largely went unnoticed. Now the transition into enterprise with the help of Netscreen gives them an unusual time advantage. Do I think it’s tricky? Yes, I do. Cisco has an extremely strong and embedded relationship with enterprises.

LR: Some descriptions I’ve heard of the market just this week say that Cisco and Juniper are desperately trying to work their way as high up the networking stack, because if you are lower in the stack the more susceptible you are to competion from China. The hardware is being commoditized. Is that what you see happening?Kramlich: Well, I think earlier we were talking about Huawei. In my last trip I had a chance to go through the facilities at ZTE and Huawei. They are both pretty good companies. ZTE is more like a Silicon Valley company, and Huawei is almost like no company I’ve seen before. ZTE has a lot of intellectual property. Huawei has some intellectual property. Huawei has billions in revenue.

LR: Yes, and we hear they work 24 hours a day.

Kramlich: I wasn’t there for the full 24 hours, but it looked to me like they were doing that. It’s an impressive campus. And the thing that’s amazing is I believe all of their products were developed organically. Now, I can’t get into what the lawsuit with Cisco was all about because I don’t really know. But I can say they’ve turned out a comprehensive product line. What they’ve done is shipped into a lot of underdeveloped countries of the world. A lot of other companies would be reluctant to ship to [these places] because of the credit risk.

LR: But here in North America they are hitting margins.Kramlich: I would say they are force to be contended with. It sure looks that way to me.

LR: How does that affect your investment strategy? I assume you are still doing networking hardware startups?

Kramlich: Not really hardware. LR: Are you thinking more about software?Kramlich: Well, we have one company called NextHop, which licenses software which is a pretty good company. LR: Any others that our readers might be interested in?Kramlich: We are involved with Vonage. Peter Barris and Harry Weller are on the board. I had somebody tell me they haven’t seen anything like this since the early days of Cisco. There are people sitting on chairs in hallways, it is just jam-packed. They’re about ready to pass [one] million subscribers. Again, you might ask the question, “Is an investment in that company ready to be rewarded because they’re using quite a bit of capital?" Jeffrey Citron has invested quite a bit of his own capital. I’m guardedly confident that that will work out just fine.

This and other things we are doing fall into something we call whitespace. These are exciting deals that don’t fall into the traditional VC model.

LR: How’s that? Kramlich: Well, you know, put in a couple million dollars, see it through next round, and let some other guys put money in. But we believe in staying with winners all the way through and try to build a big position over long periods of time and really add value and help to build these companies. We call it specialized venture capital with scale. That’s our challenge, to deploy sizable capital with top-level performance.

LR: What do you think about the telecom market?

Kramlich: The telecom freeze is thawing, but it’s a very demanding climate and the customers really demand and expect value add that is very clear to them and they expect substantive companies that have longevity. So it’s a real testing environment. But around these vectors, and in the case of Force10, data centers are a big part of it, security and a lot of sectors in addition to telecom.

I think it’s a pretty normal time, actually.