Q&A: EMC's Tucci On A New Storage WorldQ&A: EMC's Tucci On A New Storage World

With its pending acquisition of content-management firm Documentum Inc., EMC Corp. will become one of the world's largest software vendors. EMC, which bought Legato Systems earlier this year, saw a need to create a more diversified business that could solve a wider array of customer problems. Establishing stronger links with software used to create and manage content is logical. EMC CEO Joe Tucci, who engineered the deal, was at Documentum's user conference in New Orleans last month to fill customers in on the reasoning behind the acquisition.

information senior editor Tony Kontzer discussed with Tucci the buyout, its implications for the storage and content management markets, and how it's likely to affect EMC's and Documentum's customers.

information: Documentum customers have said that they felt the Documentum acquisition was a good way for EMC to differentiate itself, but analysts have been cautious about hardware companies buying software vendors. Hardware vendors tend to get a little greedy and have the acquired software vendor write products specifically to their platform, and that's where they go wrong.

Tucci: People love to put labels on things--hardware company, software company. Constantly, I'm being asked, oh, you're turning EMC into a hardware company? And I say, nope. And then another group says, oh, you're turning EMC into a services company, since that's the fastest growing part of the company? And I say, nope. And of course, there's this group that uses the label "hardware company."

If you look at our revenue mix, we are 50% hardware, 25% software, and 25% services. If you looked at the services piece, that's software maintenance--pure maintenance, no professional services whatsoever--you'll see that we have $2.2 billion in software revenues. That's a huge software company.

So, if you look at where the bulk of R&D is going, more than 75% [went to software] before Legato and Documentum, north of 80% after Documentum. On the software side, we're spending 20% of revenue on R&D, and, overall, we're spending 12%. That doesn't look like a hardware company. Hardware companies do not spend 12% of revenue on R&D. We're dedicating ourselves--and Documentum gives us a perfect piece of this--to what we call information-life-cycle management. If you want to expand it, you might say information-life-cycle-management solutions, which is where we'll take one from column A and one from column B--hardware and software from outside of EMC, and services from inside and outside of EMC--to help customers solve business problems. So I'll call it an information-life-cycle-management-solutions company.

Again, everyone loves labels, but this isn't like we haven't done this before. It's not like we don't know how to separate and make sure we don't peanut-butter everything, which is the fear.

Documentum's mission is very different than the division I have that does the hardware. Will our software work with our hardware? Of course. Will our software work with other hardware, and other pieces of software? The answer is, equally, of course. That's just the way the fundamental company is set up and the way it's organized.

I've read the books of the past, too--what's worked and what hasn't. But the fact is that the biggest hardware company of all times for years in this industry was IBM. Now you call them a services company, because everybody loves labels and they're not spending enough on R&D to get into double digits, so they're now a services company. But if you look inside IBM, it's one of the world's biggest software companies, as well as the world's biggest hardware company, and it's the world's biggest services company. They've bought a tremendous amount of software companies and they have been very, very successful.

information: From a storage-market perspective, what were some of the drivers you were seeing that led to this acquisition?

Tucci: As we watch information built--and of course that information needs to be stored, protected, and accessed, and a whole range of other things--the majority of that information that's being created is what we call unstructured data--data outside the database. We recognized that the needs of this information were fundamentally different than that which the industry had grown up on, which was the structured data under the control of the database.

This move would drive what I would describe as the intelligence layer--and, in effect, increase the size of our pie, because any big company, and any good leader, wants to drive an organization to do two things: They want to get a bigger slice of their market, and then they want to redefine their market into a bigger market and get a bigger piece of that. This was just tremendously logical to us. Recognizing this, we had built systems like Centera, which is fundamentally very different from the ground up than any storage system before it. It shows we've been recognizing this for years, and of course, we've been keeping our eyes on this market, and we felt now was a good time to expand. And as we took a look into this market, it wasn't even close who we wanted to marry here.

information: Why should customers care what type of storage holds their content?

Tucci: Storage is kind of like, watch out what you wish for or you might get it. EMC was the company that created the category of IT infrastructure called storage. And then, of course, people still think, OK, great, that's an array where information lives on disk. But again, if you look at our definition of storage, it's the back-end network which we call SAN [storage-attached networks]. There are different ways to access data, whether it be through NAS [network-attached storage], through the SAN itself--direct attached--or through objects which we call CAS--content-addressed storage. We have the software that manages the movement and replacement of data. The software actually resides in large part on servers. We have software that resides in the network, we have software that manages this thing as a single pane of glass.

So we're way beyond what people normally view as storage. On the physical disk, I think there's a lot of reasons people should care, and it all comes around several things. Can you best protect it for me? Can you give me the best ability to share it? Can you give me the variety so I can lower my costs by understanding the differences in the value of my information and dynamically moving that physically stored data to different cost media? What kind of service? You go to almost any user survey, done by almost anybody, and we invariably are either No. 1 or No. 2 in every survey I pick up. If you were going to build a house, would you pick the lowest-cost builder? Would you pick the highest-cost builder? You'd probably say, who builds a house that I can reference-check at a competitive price that has real quality? Those are all reasons why customers pick EMC. We have a tremendous focus on quality and customer satisfaction and service.

information: If you project yourself into the customer's seat, what are the best parts of having your storage vendor play a part in your content-management decisions?

Tucci: We have specialization inside of specialization. I believe, without a doubt, if you ask any customer, a macro trend in IT is customers want to deal with fewer vendors. IBM would have you believe that customers would like to deal with one vendor, but I happen to disagree with that, because when vendors had that in the '60s, '70s, and part of the '80s, customers didn't like it. They wanted choices. I believe the answer is fewer vendors.

So what we've done as a company is say, OK, what are we doing around information so we can be a one-stop shop for customers--with total openness so that we can work with other choices that they make? We won't lock you out of a choice. If you want to go use you-name-it storage, go ahead. We'll still work with you. Or if you want us to do this little piece of the information infrastructure, we're there. We're focusing on the networked information piece.

Then you go down a layer, and we have specialty in content management and unstructured data. We have specialties in how you layout your SAN. We have specialties in many areas of storage management. And we have specialties in protection, which is totally backward today, because it's now called backup. And backup is the bane of any IT executive's existence. It's just terrible. What they really want is recovery. And different types of information need different types of recovery. So you can have stuff on tape--it's fine, if you can live with hours and hours or even days of recovery. Some applications are just fine there. Other applications, you can't be down ever, and unfortunately, more and more applications are falling into that category. Therefore, you want to get into disk replicas. But again, we're dialing in the whole thing for you.

We're going to help you find out the value of your information, give you lots of statistics about your information. So that's why I'm saying watch out what you wish for or you might get it. We're a lot more than classical storage. I think the word "storage" has come to mean these things. But still, some people think of storage as the physical hardware array, and we are so far beyond that.

information: You mentioned the commitment to the open architecture and the ability of customers to work with other platforms. But the logical assumption is that, over time, EMC's and Documentum's products will develop a tighter integration than EMC will with other content-management products, or than Documentum will with other storage platforms. So from an evolutionary perspective, even by default, aren't these two product lines going to move closer together?

Tucci: There's a yes and a no. There's a really interesting phenomenon happening now. If you looked at every other wave of information technology, the wave was led by newcomers. So you had the mainframe wave, and that bunch, and then you had the minicomputer wave, which was none of that bunch--it was DEC and Wang and Prime and Data General. Then, all of a sudden, you had a microprocessor wave, which was none of the minicomputer guys. Here you had Microsoft and Intel and Compaq and Dell and Apple, and then you had a network wave, and you had Cisco and what became Bay Networks. It's really odd what happened here.

Now what's happening is, the bigger companies are moving faster now. They're saying, no way do I think my guys are going to have all the good ideas in the history of man, so I'm going to move really fast. I think we blew [Documentum CEO] Dave DeWalt's mind with how fast we boogied in this transaction, that a big company can move this fast. So what I'm telling you is, I've seen the last chapters of the last movies. I'm not going to let these guys on the open side of our business get so reliant on the hardware, or this side of the business, that we lose that veil of openness. I don't want it to be a veil. I want it to be a reality.