NetApp's Past and FutureBy Chris Preimesberger | Print
Desktop-as-a-Service Designed for Any Cloud ? Nutanix Frame
Q&A: CEO Daniel Warmenhoven has overseen Network Appliance's growth from a startup to a major storage player.
Over the past 13 years as CEO of Network Appliance, Dan Warmenhoven has overseen the company's growth from a small startup to one of the key players in the highly competitive data storage market. Warmenhoven recently sat down with eWEEK Senior Writer Chris Preimesberger to talk about NetApp's past and future.
Congratulations on your 13th anniversary as CEO of NetApp. That's an eternity in the IT worldin fact, most enterprises in general.
I've been CEO ever since I came to NetApp in October 1994 after working for IBM and [Hewlett-Packard] and serving as the CEO of another company called Network Equipment Co. At NetApp, we now have about 6,900 employees. When I started, it was 45 employees, and the quarter was about $3 million in revenue. This quarter, we're forecasting somewhere north of $750 million in revenue, with almost 7,000 employees. So it's been a pretty good run so far.
What do you see as your greatest accomplishment in your time at NetApp?
That this company is number six on the "Great Places to Work" list. We've participated in the "Great Places to Work" survey, which is conducted by the Great Places to Work Institute and published by Fortune, for five years. Thousands of companies participate. For the last five years we've been in the Top 100; this last year, number 6.
People like it here. It's an exciting company. It's fast-paced; this is not a country club. It's not as though we earned our way to number six through perks like on-site massages and stuff. People work hard here, but they really enjoy it. They feel like they're appreciated, they have the freedom to pursue their objectives the way they feel they should. There's a lot of collaboration, not a lot of infightinga totally apolitical environment. Ask anybody, and they'll say this is good, this is fun.
To read about NetApp's suit against Sun over the WAFL file system, click here.
We've had tremendous success in terms of putting the company on the map, identifying a market strategy. The investor community has done wellthe company was valued at about $40 million when I joined and it's now worth about $10 billion or $11 billion. It's got an enormous compound annual growth rate.
What are the key reasons for that kind of growth and success?
A lot of it's luck, and a lot of skill. It's a combination of the two. I think we've identified a trend in the market that we've been able to expand upon since the company was founded back in 1992.
It was in the mid-90s that the two storage networking technologies, SAN [storage area network] and NAS [network attached storage], both started to come about. We were one of the pioneers of the NAS world. While people think it was a technology play, it was really an opportunity to develop a pure-play company. Once you disassociate servers and storage into separate styles of computing infrastructure, you had the opportunity to create a pure-play company that could be just a storage provider.
Our belief was that would have significant value over just disks attached to servers. The initial play was, well, that we could build a better file server. We said, "Don't use your general-purpose computer as a file server." But it expanded from that very quickly to, "Storage is a large portion of the IT spend." It happens to be the only component that has any state, meaning its content is permanent, and therefore has real value in the customer's eyes.
Page 2: NetApp's Past and Future
Storage is also the one area that's growing the fastest and therefore has the biggest challenges in management. We always felt that ... [if] the customer's got a problem, he's going to look for a solution. As long as we were at the forefront of it, we had to opportunity to be it. And that's worked well for just about 15 years. I still have about another 15 years to go, by the way.
What is the market's biggest misconception about NetApp?
Well, we started as a small file server company. We sold departmental systems. ... In many cases, first impressions are lasting. In many ways, our competitors have tried to position us that way, too, as a departmental solution, not ready for prime time. In fact, most of our sales now are to major corporations for mission-critical environments. We suffer from the image of "file servers for departments."
But if you look around, the Oracle outsourcing operation where they host their customers' European environments, that's pretty mission-critical, and that's all NetApp. The SAP Business ByDesign environment, that's all NetApp. These are mission-critical deployments of the first order, but the perception is: "Nice little file server company." We're going to change that.
To know us is to love us. If you are not a customer of us, in fact, we have been positioned well by our competitors. I think that has been part of the challenge. It's time for us to go break out and establish our own image in the market.
What new challenges does NetApp need to conquer in the future?
There are two or three. I really think one of them is products that are suitable for the medium-size enterprise. We've started a real strategic [small and midsize business] initiative. We've been really good for providing a value proposition for large enterprises, but medium-size enterprises have those same kinds of challenges, and I don't think we've presented them with any solutions scaled to them until just recently.
[SMBs] have a different style of buying. I think we have more work to do in enabling the channel partners and other parts of the ecosystem to be as effective as they need to be.
Click here to read about Sun's countersuits against NetApps.
My sense is that, given all the new style of projects that are going on in the enterprises today, both large and small, whether it be use of disk-to-disk for reduction of tape or whether it be encryptionwe can go through a thousand different project stylesthey are all new, and the customers have no experience, which means they have a dependency on the vendor or whoever they're dealing with to help them through those transitions. We've done a good job with that in our own professional services, but we've not enabled our channel partners to offer those same services.
Part of the strategy this year is to train them, make all our internal training available to our partners so they can provide that same of level of consulting services and support services to the medium-size enterprise customer. It's a whole set of solutions, not just a product. [It includes] even enabling our partners to be consultants to our customers. It's all around the evolution into the midmarket; it's also broadening out in the enterprise.
We also still have a lot of untapped opportunity. We are at about 10 percent market share. Part of the question is: Where does the next piece come from? The medium-size enterprise is a component, but I think the high end is another component. We have done some research internally of what we call the "storage 5,000"the 5,000 largest consumers of storage in the world. And we are a supplier to about 1,000 of those.
That means there's 4,000 left to go. That's a lot of market opportunity. That's 80 percent still untapped. That 80 percent represents at least 60 percent of the total dollars. It's not just sizeyeah, we're going after the big ones, but it's largely vertical-based. We've focused our energies on eight specific verticals, and there are others beyond those that are really quite large.
What are your strongest verticals at this point?
Leading the way is probably [technology]. It's kind of our core business; it's where we started in 1994-95. And the big tech companiesOracle, Texas Instruments, Cisco [Systems], Intelthey've all been big customers for a long time.
Page 3: NetApp's Past and Future
Our largest customer is Yahoo. Everything you see on Yahoo is coming off a NetApp storage system. The mail, the GeoCities stuff, the calendar, everything. You want a mission-critical environment? Yahoo Mail. When was the last time Yahoo Mail was not available? That would be a headline you'd cover.
Financial services is really big for us. It's about neck and neck now with tech. It's mostly broker-dealer: Goldman Sachs, J.P. Morgan, Morgan Stanley, Citigroup ... and right behind that is the federal government. Our U.S. federal business has just been on fire. It's been one of our fastest-growing components. It's now well over 10 percent of our total business.
What would be your message be to a SMB who was trying to decide between buying, say NetApp, and HP to redo a data center?
HP? Well, that's no contest. Let me take this from a different perspective. I think that the customer community is moving to a place where they think about the storage infrastructure as very different than servers or whatever. I view the transition as similar to the transition they went through in the networking world, where there used to be application-specific networks like IBM-SNA for transactions, TechNet for engineering, NovellNet and NetWare for office functions.
As they looked at their operations, they said, "I've got to put all those together for purposes of economy and efficiency." And now customers think about networks as a very independent component of the infrastructure. They're getting to that same place with data management and storage.
Read here about Sun's second set of countersuits against NetApp's.
Now, they still buy it in an application-specific sense: 'I'm going to upgrade my Oracle environment, I need new servers, I need new storage. I need to upgrade my Exchange environment, I need new servers, I need new storage. I need to do server virtualization, I've got to rethink my storage infrastructure.' They think about [storage] still as an extension of whatever is going on the server side.
Step one of this analysis is: Look for a solution where you get the maximum amount of value-add in your storage infrastructure to support your application environment. So we've invested heavily in integration with [Microsoft] Exchange, SQL Server, Oracle, SAP, VMware, and we actually integrate the services in the storage infrastructure with what you're trying to achieve on the application side.
Step two, then, is to think about where your highest leverage points are in terms of getting the maximum from your asset, the best [return on investment], and maximum efficiency of your management staff. Storage is growing in most enterprises at 50 to 80 percent per year, and the amount of manpower is going up at the rate of zero. This is an area where you need to get 50 percent productivity improvement every year. How are you going to do that?
Consolidation is where it starts. Virtualization, deduplication, thin-provisioningthis is how we start getting more productive. We just sold our 1,000th deduplication license the other day.
Page 4: NetApp's Past and Future
What advancements in data storage do you anticipate in next three years?
This all moves at glacial speed. The communications transition I mentioned earlier? Well, local networks came out about 1985; by 2000, everybody's running total-mesh interconnected LANs over IP. Fifteen years.
The movement from direct-attached storage to what I think is the end of the evolutionary period in this case is at least 30 yearstwice as long. Two reasons: Number one, this is people's data. If you have a service-center interruption in the network, you have a pain, an outage, but it's easy to recover. If you screw up your data, it could be really injurious. So people move more cautiously.
Step two: In the networking world, there was this great drive to move everything to the Internet. It took everybody in the same directionvendors, customers, everybody was headed towards an IP-based infrastructure, and we don't have the same compelling kind of forcing function in the storage world.
One final topic: I love the blog wars between [NetApp co-founder] Dave Hitz and [Sun Microsystems CEO] Jonathan Schwartz when they talk about the IP lawsuits your two companies have filed against each other [regarding the origin of the cutting-edge Zettabyte File System].
The problem with Jonathan's [blogs] is they're just full of lies, you know? He says they never presented us with a demand: Dave published the demand note. Jonathan never retracted it or said anything, he just kind of persisted.
Both sides appear to have very strong cases [regarding the origins of the ZFS, and whether it should be in the open-source community].
It's very compelling. What gives them [Sun] the right to steal our intellectual property and publish it to the world? I think Jonathan was out of touch; I don't think he knew what he organization was doing. That's my suspicion. This was largely done by a bunch of guys on the StorageTek legal team that he inherited. And they persisted in their activities that started prior to them being acquired by Sun.
My guess is that they were still marching to their set of objectivesthey present us with a demand, Jonathan doesn't know it. And then, you know, we presented them with a set of demands around our intellectual property, and they never surfaced them. Then they stopped returning calls; it's like all their phones broke. And then, the question is, what do you do? You only have a couple of choices, but given Sun's history, we decided we'd fire the first shot.
Why did you file the lawsuit in Texas?
Expertise. The federal court system has, in fact, developed areas of expertise. And guess what? The judges that hear the most cases and have the teams that understand these things the best are in Texas. How it got going, I don't know, but talk to any attorney around here and they'll tell you. There are two places like this in the U.S.Texas and, I think, in Minneapolis.
What you really want is a judge who really understands all the intellectual property law, and who has immersed himself in all these issues over the years and can separate the facts from the hyperbole, and the wheat from the chaff, and be a very reasoned adjudicator of the issues.
Check out eWEEK.com's for the latest news, reviews and analysis on enterprise and small business storage hardware and software.