Can EMC Restore Its Glory?
The hardware juggernaut's answer to its problems: software.
Monday, July 8, 2002
By Daniel Roth
In August 2001, two dozen coders, product managers, and marketers at storage kingpin EMC were sitting in a conference room on the company's sprawling campus outside Boston. The gathering was routine: The group was plotting a road map for the latest release of EMC's ControlCenter software. One programmer explained to Erez Ofer, who oversees the group, how a certain piece of the software would help power EMC's crown jewel: the Symmetrix box, a data-storage system that can cost several million dollars. Ofer stopped him. "I think you should build that for Hitachi instead," he said, invoking an EMC archrival. The room fell silent. Then came a reply: "Are you serious?"
He was, and so is EMC. In an attempt to regain its past glory, the beleaguered company is rolling out a bold new strategy. Since 1991, when it practically invented the modern storage market, EMC has defined itself first and foremost as a hardware company, albeit one that sold pretty darn good software. But the technology world has turned upside down, so EMC is attempting a flip of its own--emphasizing software instead of hardware. And if that means embracing Hitachi's or any other competitor's technology, so be it.
How big a change is this? Consider: In 2000, 70% of EMC's $8.9 billion in revenues and 92% of its $2.3 billion in operating profits came from sales of its storage hardware. By the end of 2004, CEO Joe Tucci wants that down to 50% of revenues, with the rest coming from software and services.
To get there he's pitting programmer against programmer and splitting EMC's famously hard-driven sales force into hardware and software camps as well. "Companies that are afraid to disrupt themselves almost 100% of the time end up being disrupted," he says. "I'm doing what our competitors never thought we'd have the intestinal fortitude to do."
Clearly, Tucci's gut has gotten quite a workout in the 30 roller-coaster-like months he's been with the company. In the late 1990s the firm was heralded as one of the "four horsemen" of the New Economy, along with Oracle, Cisco, and Sun. As bits replaced paper and people's lives moved to the Net, EMC was going to be the world's collective memory, stor-ing everything from grandma's meat loaf recipe to her entire credit history--and not just storing it, but replicating it and making sure it reappears instantaneously even after a disaster, man-made or natural. Companies bought into the idea--and so did investors. EMC was one of the top-performing stocks of the 1990s (which is saying something), returning a staggering 84,000%. For chief information officers looking to build fail-proof storage, EMC was the only option. Sure, IBM, Hitachi, and Sun had boxes that claimed to do the same thing, but EMC's blew them all away. The company knew it too. EMC's sales force developed a reputation for arrogance, frequently refusing to budge on prices or forcing IT buyers to relicense software when they upgraded machines. (EMC says it has since changed the way it licenses software.) By the end of 2000 the company controlled 71% of the lucrative upper end of the market, despite the fact that its prices were nearly double those of its nearest rival.
But in 2001, EMC found itself fighting a war on two fronts--and losing. As the Net bubble burst, the company took a beating on the demand side. Telcos and dot-coms that spent $1.7 billion on EMC products in the prior year disappeared. Other customers froze or slashed their IT budgets. On the supply side, EMC was caught flat-footed by the sudden arrival of real competition. IBM started bundling its Shark storage array (still considered inferior to EMC's boxes) with its mainframes at a price that budget-conscious CIOs just couldn't refuse. Hitachi, meanwhile, beefed up its Lightning storage arrays, making them faster than EMC's. Then Hitachi, through resale agreements with HP and Sun and on its own, offered the machines at just over half the EMC price.
At the same time, customers began grumbling that storage was getting too difficult to manage. As the machines became increasingly important to companies and started moving from simply being connected to mainframes to sitting on the network, the headaches for techies turned into migraines. The heart of the problem was a lack of standards in the storage market, which kept the boxes--unlike their PC and server cousins--from being able to communicate with one another. EMC machines speak fluently to other EMC boxes, but what they say often comes out as gibberish to, for instance, Compaq storage devices. Likewise, IBM storage software will show how much room is left in an IBM box, but it can't see into one of EMC's. And a Hitachi can back up another Hitachi, but--you get the drift. CIOs like Mike Prince at Burlington Coat Factory found themselves devoting precious IT resources to keeping on top of their setups: "We spend more time worrying about storage than any other aspect of the total computing environment."
To make life easier for corporate users like Prince, storage-software makers and a growing number of storage startups are scrambling to build lingua franca software that would sit on a server and manage all the storage, no matter the vendor or technology. If they succeed, a backroom full of Compaq, EMC, and IBM boxes would all work as one. Analysts call it storage's Holy Grail. But it's also the first step toward the commoditization of storage hardware. Once companies no longer care about which vendor's box they use, the real opportunities--and the big margins--will be in the software.
As software rivals worked to make EMC irrelevant in the future, hardware foes were inflicting plenty of pain in the here and now. Amid the industry's worst-ever price war, EMC in 2001 watched its hardware gross margins drop 25 percentage points, to 32%, and its share of the high-end market share drop 14 points, to 57%. For the first time in over a decade the company reported an annual loss, ending the year $508 million in the red on sales of $7.1 billion. Soon investors were seeing red too: EMC's stock now trades at $7.20, down 93% from its September 2000 high.
Suddenly rivals and analysts were talking about EMC's very survival. "EMC lives or dies by its ability to continue to sell into its traditional FORTUNE 500 base," says Jon Oltsik, founder of IT consulting shop Hype-Free Consulting in Acton, Mass., and a former marketing manager at EMC. "If it can't figure out how to do that, then EMC as we know it goes away."
As it turns out, EMC as we know it is going away--but by design. When the company began taking its hits, Tucci, a 54-year-old former programmer, started having flashbacks to his last job. Between 1993 and 1999 he was the CEO of Wang Laboratories, a once dominant minicomputer player that had slid into Chapter 11. He led Wang out of bankruptcy and eventually sold the company, but during his tenure he saw firsthand the dangers of sticking with a good thing. Scared of cannibalizing itself, Wang in the late '80s fought a losing battle to protect its market as the PC revolution grew. And while he thought EMC was far from being Wanged, Tucci wasn't about to take any chances.
So last November he acted. In an off-site meeting at the International, a private golf club in Bolton, Mass., he announced a plan to his top lieutenants that would in effect split EMC in two. To compete against its traditional rivals, he put 15-year EMC vet Dave Donatelli in charge of the half responsible for making sure EMC's hardware lines were the best in the business. To handle the threat from independent software vendors, Tucci tapped Ofer, a speed-talking ex-Israeli Air Force captain. Ofer's division would create and market "open software''--software that would make using an IBM Shark as easy for EMC customers as using an EMC Symmetrix. And both men would have to hit their targets with less staff: Last fall Tucci announced a 20% downsizing of EMC to 19,000 people. "The software guys shouldn't worry about protecting our hardware leadership role,'' says Tucci.
Therein lies the rub. Selling software separately could undermine EMC's hardware sales. "For a hardware company to get into this game, they need to write software that will enable customers to buy less hardware," says Jeremy Burton, chief marketing officer of independent storage-software leader Veritas. "That's like Microsoft coming out with a set of Linux products.'' But Tucci sees things differently. He's betting that both halves of EMC will steal market share from rivals before they start bumping into each other.
On the software side, Ofer's group already knows its way around a Symmetrix; the division's sizable share of EMC's $800 million R&D budget, the largest in the industry, will help it figure out how to program competing hardware. Plus, hints Ofer, the company isn't ruling out acquiring promising software companies. On the hardware side, Donatelli thinks he can make money by pushing EMC hardware into new markets. In October the company inked a deal with Dell allowing it to resell EMC's lower-level Clariion storage. The PC Goliath is particularly good at selling to small businesses and government--two areas where EMC has made little headway.
At the high end, Donatelli is also counting on customers to stick with what they know works. So far they appear to be doing just that--eschewing commodity hardware for the convenience of having a company like EMC or IBM manage their storage networks. "Maybe [buying cheaper boxes] is where many of us will end up," says Fred Dingraudo, director of technical services for a division of car-parts manufacturer Dana Corp., an EMC customer. What's keeping him from going there now is the cost, complexity, and time required to cobble together his own storage system. "In the short term, I'm looking at EMC as a full-blown turnkey solution for my storage needs," he says.
One last piece Tucci needs to work out is realigning his sales force. With their commissions on the line, the sales group has little incentive to push software in the hundreds of thousands of dollars instead of hardware in the millions. So Tucci is planning on splitting the force in two. "Here I'm using an evolution, not a revolution," he says. To keep big spenders from being inundated with EMC business cards, Tucci's putting "relationship managers" in charge of each account. "The largest customers will see only one EMC," he says.
Tucci's strategy, if it works, should enable EMC to play in both the less profitable, but still growing, hardware market and the fast-growing, high-margin software arena. Toni Sacconaghi, senior research analyst at Sanford C. Bernstein, thinks it's an important part of a strategy that could boost EMC's operating margins to 17% by 2005, compared with minus 8% this year.
Not that he's recommending the stock. "We like the secular story," he says of Tucci's plan. "The problem is EMC's valuation isn't supercompelling." At just over $7, the stock is trading at over 40 times expected 2003 earnings, vs. the high teens to the mid-20s for its peers.
And not one of those peers plans to just hand EMC the rights to both storage hardware and software. Veritas is committing itself to developing software that will manage the entire storage infrastructure. Hitachi says it's working on an open-software project, and the new HP, which competes with EMC in the midrange with its own devices and at the high end by rebranding Hitachi's Lightning, thinks it has the size to take on EMC. "Granted, EMC's building a credible software portfolio, but we'll compete highly with them in that space," says Mark Lewis, who oversees storage for HP. "Because we're doing very much the same thing."
All of which means that the chances of EMC's ever fully reverting to its late '90s form are gone. But then, those are the memories of the old EMC. Tucci's moving on.