StockTalk:
TMF Interview With
Network Appliance Founder Dave Hitz
With Brian Graney (TMF Panic)

September 23, 1998

In 1992, Santa Clara, California-based Network Appliance, Inc. <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: NTAP)") else Response.Write("(Nasdaq: NTAP)") end if %> created a new file server category for network storage with its high-performance appliance designed for specific network functions. Since then, net sales have jumped from $2.2 million in fiscal 1994 to $166.2 million in fiscal 1997. We caught up with Network Appliance V.P. and founder Dave Hitz, who brought us up to speed on the company, its products, and its plans for the future.

TMF: What are you trying to do at Network Appliance?

Hitz: The best way to understand Network Appliance is to talk a little bit about Cisco <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: CSCO)") else Response.Write("(Nasdaq: CSCO)") end if %> and the TCP/IP routing industry. The reason for that is that there's actually some really strong similarities. In fact, the chairman of our board, Don Valentine, is the vice chairman of the board at Cisco and came on to Network Appliance's board largely because he saw the similarity in what we were trying to do.

If you look back in the middle of the '80s, there was a large TCP/IP routing market, but nobody recognized it. And the reason for that was because it was just another thing that general purpose computers do. If you wanted to do TCP/IP routing, you would go buy a Sun <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: SUNW)") else Response.Write("(Nasdaq: SUNW)") end if %> workstation or a Sun compute server and configure it as a router.

In the mid-'80s Sun actually had over 50% market share of TCP/IP routing, but nobody recognized it as a market. Cisco came along and said TCP/IP is over 10 years old, it was invented in the '70s, it's an important protocol and it's here to stay, and it's time that somebody built a dedicated black box that does nothing but TCP/IP routing. That's all it does. It's not running a general purpose operating system. It's a simple device that does just that one thing and as a result it's fast and fully reliable. So, our goal at Network Appliance is to basically repeat Cisco's success but in the network storage area.

I should say what network storage is. Network storage is basically the idea that instead of having your disk storage attached directly to your computer, you should access that storage out over the network. There are some standard protocols for doing this -- NFS is the protocol in the UNIX space and CIFS is the name of the protocol in the NT space. Both of these are old protocols; they've been around for a decade, just like TCP was around for a decade when Cisco started. Both of these functions are currently served by general purpose computers. If people want an NFS file server, they'll go buy a Sun workstation or a Sun compute server and configure it to serve files out over the network. And similarly for NT -- you'll go buy an NT server from Compaq <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: CPQ)") else Response.Write("(NYSE: CPQ)") end if %> or Dell <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: DELL)") else Response.Write("(Nasdaq: DELL)") end if %>. So, our goal is to provide that functionality in a dedicated box. We call it an appliance -- it does just one thing.

"We encountered a very rare opportunity, which is we invented a new way of solving a problem. We solved the network-attached storage problem..."

TMF: Is this something that Cisco's working on, too? Or is this a niche that Network Appliance pretty much has for itself?

Hitz: No, it's not something that Cisco's working on. There is a particular product area that we overlap, but there's a very large difference between routing packets and moving packets to and from disk drives. We describe our fundamental technical competency as moving data at high speed between disk drives and networks.

TMF: What are the types of organizations that would be using your appliances rather than another competitors'?

Hitz: Well, we've actually got a large number of customers. The initial customers that we had were largely technical companies and doing technical applications, so especially in the ECAD area, we've got customers like AMD <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: AMD)") else Response.Write("(NYSE: AMD)") end if %> and Cirrus Logic <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: CRUS)") else Response.Write("(Nasdaq: CRUS)") end if %>. Motorola <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: MOT)") else Response.Write("(NYSE: MOT)") end if %> and Intel <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: INTC)") else Response.Write("(Nasdaq: INTC)") end if %> are actually customers.

We moved from there into other kinds of highly technical customers doing technical design work. We've got 3Com <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: COMS)") else Response.Write("(Nasdaq: COMS)") end if %> as a customer, Bay Networks <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: BAY)") else Response.Write("(NYSE: BAY)") end if %>, Cisco, Qualcom <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: QCOM)") else Response.Write("(Nasdaq: QCOM)") end if %>. Again, highly technical companies. After that, we started branching out into other areas, so we've actually got government contracts like Lawrence Livermore and Los Alamos National Laboratories and even banks. Bear Stearns <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: BSC)") else Response.Write("(NYSE: BSC)") end if %>, Citicorp <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: CCI)") else Response.Write("(NYSE: CCI)") end if %>, and NationsBank <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: NB)") else Response.Write("(NYSE: NB)") end if %> are all customers.

So, if you look at the industry as a whole, it's actually pretty widespread -- the need for large amounts of data. And increasingly people are wanting to take those large amounts of data and make them accessible over the network. The market for this stuff is incredible. Like I said, that market is mostly served by Sun servers, Hewlett-Packard <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: HWP)") else Response.Write("(NYSE: HWP)") end if %> servers, or Silicon Graphics <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: SGI)") else Response.Write("(NYSE: SGI)") end if %> servers on the UNIX side, and Compaq servers or Dell servers on the PC side. That market today is $10 billion.

TMF: That's been really evident just looking at your financial statements, as your revenue growth has been really extraordinary the past five years. I was wondering if you could comment a little bit on how that's come about and whether you think it will keep growing at the 40% to 45% compounded annual growth rate that you've seen.

Hitz: We encountered a very rare opportunity, which is we invented a new way of solving a problem. We solved the network-attached storage problem -- the server for NFS or CIFS. We invented a new way of solving that problem, but it's for a giant existing market. Usually when you get a company that invents something new, they invent some cool, new thing but there's not a big existing market for it. So, first they have to evangelize not only their solution but they also have to evangelize why you need such a thing in the first place. So, we've got this advantage that there's a big existing market, just like Cisco had when they started.

Our run rate today is about $230 million, if you take our last quarterly report, and we're selling into a $10 billion market. Well, if you've got a better, new solution in a giant market, then you've got an opportunity for a really dramatic growth rate. I do think we have an opportunity to sustain that growth rate going forward.

Obviously, we have to do an awful lot of things right, but the market is clearly there for our product. And when you look at other companies in situations like we've been in and a new product that really meets a market need in an important, new way, you do see dramatic growth. Companies like Sun, Cisco, Bay Networks, back when they started, all of those companies underwent very high growth rates -- up past $1 billion (in annual revenues) -- and I think we have a similar opportunity. I really believe we can be the Cisco of the network storage market.

"I really believe we can be the Cisco of the network storage market."

TMF: I was wondering what Network Appliance was going to do in the future to get their name out there like those other companies.

Hitz: It's a little bit tricky being an infrastructure company. We sell stuff that goes in the wiring closet and so there's a very focused market of people that we want to know about our stuff. On the other hand, we have done some advertising and we're continuing that and expect to do more going forward.

I've talked to a lot of sales guys because we sell primarily direct and we've hired a lot of sales guys from companies like Cisco, Sun, and 3Com. And I've asked them, "So, how big did your company get before there started to be name recognition, before people started to have heard of you?" The answer I generally got back was somewhere in the $200 million to $400 million (revenue) range per year. So, I really believe that over the coming year or year and a half we'll start to get a lot more name recognition. People will have heard who we are and know what our story is about.

TMF: So, in this industry the issue of name recognition is maybe secondary to the performance of what the product is going to do?

Hitz: You know, as we branch out into more industries, I think name recognition will get more important. Today our primary customer is a highly technical user and a technical company and those guys are really focused. They're technical folks and they're focused on products that work. And you need to get your name out enough so that they'll consider you as they're looking at what they want to purchase.

Our VP of Sales likes to point out that, for a company like this, our primary competitors are people like Sun and Compaq. We would not exist us a company. There's no way that we could exist having grown at our growth rates if our products weren't better. We'd be dead if our products were just as good because everyone would just buy the products they're already buying from the salesman they already know.

"Many people view the Internet as largely a networking phenomenon. I do not. I view the Internet largely as a storage phenomenon. The question is, 'We've got all this data, how do we give people access to it?'"

TMF: This industry seems to have just popped up basically 10 or 15 years ago. It's really not that old. What do you see it looking like in 10 or 15 years down the road?

Hitz: I really believe that a high percentage of people's data is going to move out to the network. I believe that we're just entering the era of storage. If you look at the computer industry, I think there's some very, very broad trends that you can identify.

If you look back in the late '70s and early '80s, the fundamental focus of the entire computer industry were the questions: What is a computer? What is an operating system? What does it mean to be sitting in front of computer and interacting with it? What are the applications I need? What's the model of computation? And out of that emerged a collection of companies that focused on those issues -- companies like Intel defining the chip set, Microsoft <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: MSFT)") else Response.Write("(Nasdaq: MSFT)") end if %> defining the operating system, application systems like Oracle <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: ORCL)") else Response.Write("(Nasdaq: ORCL)") end if %> and Lotus.

I believe that in the mid-'80s and late '80s we really shifted to asking questions about networking. What does it mean to have a network? Do we use TCP/IP? Do we use bridges or routers or switches? And companies like Cisco and Bay emerged to deal with those issues.

I think in the late '90s and moving into the next century, the fundamental question will be storage. If you look at the amount of storage, the amount of information that we've got these days, it's incredibly large. In fact, many people view the Internet as largely a networking phenomenon. I do not. I view the Internet largely as a storage phenomenon. The question is, "We've got all this data, how do we give people access to it?" I see large amounts of data moving out into the network and people focusing increasingly on the concept of how to design and build their storage infrastructure.

We've got two classes of products. I mentioned earlier there's an area where we compete with Cisco. The first class is storage products that store data permanently. The second class is caching products. We've got a family of products that do Web proxy caching.

Basically the way these products work is if you've got a browser and you want to talk to an Internet server way out someplace -- a Web server somewhere on the other side of the Internet -- it's going to be a long, slow trip. The Internet is still pretty slow. If you go through a Web cache, the Web cache can check its local disk and see whether anybody has looked at that page recently and just give you that cached copy. It gives you much faster performance as an individual. It saves your company or your ISP network bandwidth, and it also offers opportunities for control like parental access or different kinds of control that people might want.

We actually see the two together -- network storage and network caching -- as really fitting hand in hand as people are moving forward trying to design their network infrastructure or their network storage infrastructure.

TMF: I'm really glad that we found out about you and we were able to sit down and talk about this today.

Hitz: Great, thank you very much.

Related links:
Network Appliance website
Networking message board

Got an idea for StockTalk? Who would you want us to interview and what should we ask? Drop us a note at [email protected].

Transcript Archives