CRN Interview: Gary Bloom, Veritas

Michael Vizard, editor in chief of and Heather Clancy, editor, recently sat down with Gary Bloom, chairman, president, and CEO of Veritas, to discuss storage, channels and his company's move into utility computing.

CRN: You don't have a specific goal when it comes to building revenue through the channel. That's kind of unusual. Why is that?

Bloom: I have a pretty simple view on this, that the partners in the transaction somewhere along the way need to add value. And if they're adding value, if 100 percent of my revenue came from partners, if 100 percent of them were adding value to the transaction, then I would be fine with that. So I don't have a goal because in a way a goal can be limiting on the high end and it can constrain your activity. My big thing is that [partners] need to add value. And we see these partners, in some cases they have relationships we don't have, they offer financing we don't have. In some cases they offer software licensing and management that we don't have. And certainly in the case of Backup Exec, they reach an entire customer base that we certainly aren't interested in selling to directly, that they service very effectively. And so we try to leverage their value.

CRN: Do you think the enterprise level will be the level that gets into utility computing, or will all levels of the channel participate there?

Bloom: No, it will be much more the enterprise capability. If you think about a small- or medium-business user that has 10 servers running Backup Exec, the likelihood [they will embrace] utility computing is a strategic dud. They may get there . . . through technology innovation evolution or get there because they buy a packaged solution from someone like Dell that includes a bunch of things that run like a utility.

CRN: Is there a race to get to the next level?

Bloom: I think we're more in a race with ourselves than anything. The foundation of utility computing and the thing that today operates most closely to a utility model is storage. Typically centrally managed, typically centrally backed up for protection and recovery, and typically very automated. And, oh, by the way, most of the value has already shifted from hardware into software. Servers are lagging that trend a little bit. We're in a pretty natural position, and while some of these other companies have things we don't have, we have assets they don't have in order to get there.

We have the leading application performance software. When you think about utility computing, [you need to] think about it from an application users' perspective, which is what we're kind of doing here. We're not talking about a storage utility or a server utility. It's an application utility. The goal of the user is applications. And what do they want those applications to do? They want them to run really quick and they want them to be highly available at a low cost. . . . There were better revenue streams we could have bought for the same dollar as Precise. We bought it because it's strategic, because we believe the fundamental view of the user in a utility computing model is going to be through the view of the application. And what you ultimately provision and what you ultimately dynamically manage in a utility is going to get determined based on what the performance manager tells you needs to happen. It's just going to be a central driving force of how people view and run a utility.

CRN: There are companies that would consider you a small player that doesn't have the critical mass to drive this. What should people look at in Veritas to view you as credible in that space?

Bloom: I think the market has been underestimating Veritas for a number of years, so I love it that they are discounting us and thinking we can't compete there. Because that means they will focus on other players, and we'll come up and do the same thing we did to them in the backup market. Remember that Computer Associates didn't think the backup market could be owned by anybody but CA. In a period of two years, we took 25 percent market share. . . . If you think about it, that's part of our competitive advantage, continuing to be underestimated. Sun Microsystems thought they could offer clustering technology five years ago because they are the operating system company that supplies Solaris and displace our clustering. We outsell them about two for one on their own platform and it's a huge revenue stream [for us].

So if you look across the assets we have--and the fact that we have a complete offering of storage management software and we've built out the entire offering of the storage utility and we own that space--then you say, that's as good of a foundation in my view to go and own utility computing along with performance management software as coming in it as a hardware vendor, like IBM. What's IBM going to do? Sell a bunch of hardware components and offer the customer a multimillion-dollar services package to bundle and make it work?

CRN: What partnerships will you need to forge to make your strategy more complete?

Bloom: Clearly, I would be a hypocrite if I were to say anything different: We don't think one vendor will supply utility computing, which is why we are supplying building blocks to help enable it. It's not going to be IBM, it's not going to be Microsoft, it's not going to be Oracle's 10G offering, it's not going to be Sun's N1 strategy, because none of them meet all the requirements. Because the reality is that there is a heterogeneous world today. Sun N1 is a great strategy if you run StarOffice on the desktop and you have iPlanet application server in the middle tier and you run Sun storage. Now if you find me any customers that use any of those three components meaningfully, I'd be shocked. . . . The reality is that if you are going to implement a Sun utility, you are inherently going to [implement] a Windows utility as well.

So when we think about what partners we need, we think about the enterprise capability that we are going to need. We are going to need to work with Microsoft. We are going to need to work with the Microsoft channel because they sell and service to that enterprise side of the Microsoft market. The Microsoft solution providers will become incredibly critical right at that crux where you move from small business to medium business and above, and then going all the way up through the channel, it's a building approach, it's not a delivery approach. . . . What partners will we need? Well, we're not going to be an apps company, so we'll have to partner with application providers. We don't intend to be an application server company, so we will have to partner with the BEAs and the [IBM] WebSpheres of the world. We're not going to be a database company, so we're going to have to support and partner with the database companies. And we certainly, right now, don't have any intention of entering the security space because it's a pretty saturated and confused market right now.

CRN: Storage is in its ascendancy, and servers in some ways are almost $1,000 peripherals to massive storage devices. How does that change the dynamics of storage, and have you given thought to your role in data management, given that?

Bloom: We look at the compliance issue. . . . I think Veritas is in a great position to become a huge enabler of that ability to switch from backup and recovery to archival and retrieval of information. One of the reasons we're on such a strong foundation is that with all the NetBackup copies that have been sold through the world, we can go back in and index all that data as well. So, we're not talking about starting with a compliance strategy for Veritas customers from today going forward, we're talking about a compliance strategy that says, 'As far back as you want to go in NetBackup, we'll go ahead and index and archive that data and make it retrievable at a lower cost all the way up through the data you are collecting today.' And that's an enormous advantage.

CRN: What are you working on in the realm of content management?

Bloom: One of the things you haven't seen here is a project that we have talked about publicly. There's really two projects, predominantly centered around dynamic distributed data. We call it our 3-D project; it's also sometimes referred to as content-addressable storage. [It's about] the dynamic movement of data around a network because ultimately the way data is going to travel around the network is predictive technologies put the data and information in front of the end user with some predictive qualities. . . . Chances are that if you are out in a remote location, you are not going to be waiting for data to transmit broadly across the network, the data is going to be flowing through the network pretty freely. . . . We have prototypes up and running and engineering under way on it.

CRN: What do you think of the EMC-Documentum deal?

Bloom: It's going to put EMC in competition with application providers because Documentum is really nothing more than yet another form of application. It's an application that manages content and manages distribution of documents. And document management from an application perspective is something that SAP is going to see as competitive, something that Oracle is going to see as competitive. So I don't think it's a very friendly cannonball across the bow of the ship for very few people. Our belief is that it ultimately creates, just like the Legato acquisition did, another opportunity for us which is to go back to all of Documentum's competitors and say, 'Here's an alternate file system for all your management infrastructure to be provided by somebody who is not going to compete with you.'

CRN: Microsoft is talking about creating a dedicated storage product. What are they these days, friend or foe?

Bloom: The nice part about Microsoft is that you always know where Microsoft is going because they typically let the entire world know two or three years in advance. Will they offer storage, will they offer some form of backup capability? Absolutely. But what's the other thing we know about Microsoft? They're not going to do any technology to support Unix. They're not going to do any technology to support Linux because they have to feed their operating system addiction. If you think about them being hooked on their operating system, then it becomes a matter of what customers want. Do customers want to manage Windows storage and Unix storage and Linux storage all differently or all the same? Do they want to back up data differently on Windows than the way they back up data in their Unix environments? I don't think so.

CRN: If you think about what Veritas will look like a year from now, how would you describe it? What is it?

Bloom: It will be a storage management company, and we still will be the leader there. It's clear that EMC's strategy here is, 'We hope that Veritas continues to expand outside storage.' [They hope that] somehow we'll forget about storage and they can own that space. If they think we are in any way conceding or walking away from the storage market, everything we [recently announced] is right in the core of our storage and backup/recovery business, a direct hit on preventing them from getting any traction whatsoever with the Legato acquisition. We have just raised the bar. We are the leader; we intend to continue to be the leader. Will we do other things besides storage? Will we offer broader technologies that help customers evolve toward utility computing? Well, absolutely. We'll have to figure out the challenge of what tagline we put on that. . . .

If you think about it, look at some of the industry giants whose growth has stalled out. Why did it stall out? Because they didn't add a second line of business or a third line of business that was successful. You know, Oracle desperately needs an applications business. Their database business isn't growing and they haven't been successful in applications.

CRN: Are you worried about them becoming interested in the storage business, because databases are just software to manage storage in their minds?

Bloom: Yeah, but their minds are completely Oracle-centric because it makes an assumption that all information is in an Oracle database. [In Larry Ellison's mind], there is no consideration for the fact that the vast majority of the world's data is unstructured and not stored in a database. That data doesn't exist in his mind-set.

CRN: People get confused about the phrase 'utility' and see it not just as a computing model but also as a way to pay for things.

Bloom: It's a legitimate question. We are doing some interesting price models and experiments with customers based on capacity-based pricing; based on pricing based on consumption, or the amount of disk or data under management or the amount of data backed up. So we are doing these kinds of pricing experiments with multiple different customers. One concept that is very clear about utility computing, one thing they want to be able to do is they want to be able to bill back the users for what they consume. It's not that they want to collect money from their users, they want to bring attention to what the cost is of what the user is doing. If the guy knows that it will cost him money to back up all of this data, maybe he will only back up a subset of his data and he'll discriminate between what's his personal data and what's his business data and only be backing up his business data. One of the things they want to be able to do is align what something is costing with who is consuming it and be able to tell him that. And that will require different price models. . . .

CRN: How much time are you spending, personally, with channel partners?

Bloom: I probably spend as much time with channel-type guys as I do with end-user customers. About evenly split. I've been going to more and more of their conferences and speaking at them. More one-on-one type of meetings with them. Starting to get to know some of these people. . . . As much as a CEO running all of sales can cover, I'm getting in with them.