When storage companies look to the horizon, they see clouds of big data rolling in, and the view brightens their day.
The world is generating phenomenal amounts of data. IDC projects that the "digital universe" will grow 44 times from 2010 to 2020, with the converging trends of cloud computing and big data driving enterprises to store and exploit vast seas of information. It's a good time to be in the storage business. And it's a good time to be storage industry leader EMC.
EMC's revenues have more than tripled since 2002. The Massachusetts company has been shoulder-to-shoulder with its larger rivals in snapping up smaller storage technology companies. EMC has also continued to invest heavily in research and development. The company's technology strategy is more aggressive than its rivals' strategies, but by no means out of place in a rapidly growing market where the players strive to have an offering for every buzzword-laden customer query.
Last month, EMC CEO Joe Tucci told a University of Washington audience that between 2003 and 2010 his company spent $10.5 billion on R&D and $14 billion on acquisitions. EMC spends 11 to 12 percent of its revenue on R&D, and company officials have said that it will continue to be opportunistic in acquisitions. "The companies that have done well are the companies that have played an offensive game," Tucci told the audience. "We're spending almost $5 billion a year on technology."
EMC is playing offense in three games these days. EMC is the data-storage market leader attempting to fend off strong advances from its competitors. It's also a company with "only" $17 billion in revenues battling 12-figure behemoths IBM, HP, and Hitachi. And it's the crusty 32-year veteran facing off against nimble younger competitors like NetApp and a host of niche players.
"There is a healthy degree of paranoia inside EMC," said Jeremy Burton, EMC Executive Vice President & Chief Marketing Officer, the company's storyteller-in-chief. Burton held up HP as a cautionary tale of what happens when a company scales back its technology investments. "You cut that to a certain point, and you end up stealing from the future, and it's very, very difficult -- once you've stolen from the future for a few years -- to catch up."
The competition is certainly heating up. Last month, Dell decided to stop selling EMC gear in favor of its own storage hardware. And just last week, venerable supercomputer maker Cray joined the party with an integrated storage system that can scale up to 50 petabytes.
As the big guys jockey for position, they've gobbled up all sorts of smaller players. HP paid $2.4 billion for cloud storage vendor 3Par. IBM bought data warehousing vendor Netezza for $1.7 billion. NetApp bought storage array maker Engenio for $480 million. And Dell bought storage area network vendor Compellent for $960 million. Not to be left behind, EMC paid an estimated $350 million for data warehousing vendor Greenplum, and $2.2 billion for big data storage cluster maker Isilon.
"Most of these vendors are checking boxes in their portfolio and making sure that they don’t lose ground to competitors by lacking a play in a hot new area," said Andrew Reichman, a principal analyst at Forrester Research. "They tend to wait to buy them until they are pretty well proven sellers, rather than taking early chances."
EMC’s purchase of Isilon shores them up in large-scale file storage, a market segment that took time to emerge but is now a high-growth category, said Reichman. HP had already bought in with its Polyserve and IBRIX acquisitions, and NetApp bought the technology eight years ago with its purchase of Spinnaker, he said.
Buy vs. build
Storage industry technology investments have evolved to be similar to those of drug companies, said Reichman. "Firms don’t seem to be willing to spend money to build out products or features, rather they tend to let small emerging firms prove out the technology and then cherry pick the best of them."
The storage companies bid up acquisition deals to the point that they probably end up spending more than they would to develop the technologies internally, said Reichman. "But it seems that Wall Street can tolerate one-time huge cash outlays for acquisitions more than ongoing expensive research programs," he said.
Wall Street is a major factor in the acquisition decision because EMC, like its competitors, is on the hook to deliver earnings. But the math makes sense, said Burton. Building organically takes operating capital that figures into a company's profit-and-loss statement, he said. "Whereas an acquisition often will take cash off the balance sheet, and if you believe that your distribution channel can get you scale on the revenue for the product, it will pay for itself."
Another reason to buy technology is that startup companies are usually unencumbered by legacy technologies and applications. Internal development efforts often focus on adapting existing technologies to new problems, and the result usually involves compromises, said Burton. This is one reason EMC acquired Isilon's technology, a storage array that EMC could have built internally, he said.
Companies are also obligated to put internal development resources toward maintaining and improving existing products, said Burton. An example is EMC's addition of automated tiering to the company's core storage product, Symmetrix. "That takes time away from building the new thing the way God would have intended you to build it," he said.
R&D, Not Just M&A
So what has EMC been spending its billions of R&D dollars on? A lot of it goes to furthering the technologies the company acquires, said Benjamin Woo, Program Vice President of IDC's Worldwide Storage Systems Research.
EMC put considerable resources into the deduplication technology it acquired when it bought Data Domain, said Burton. "What a lot of people don't realize is that product is 175 times faster than it was when we acquired it," he said.
A current project that falls in the category of R&D following M&A is Greenplum's work on a Hadoop distribution, said Burton. The goal is to allow users to perform data analytics across structured and unstructured data, he said.
For all EMC's focus on post-acquisition development, the company's strategy has a weak point, said Forrester's Reichman. "EMC has an extremely comprehensive portfolio of products, but none of them work together because they all came from different engineering departments," he said.
Not all of EMC's R&D is tied to acquisitions. The company has had some recent success with homegrown technology. The Atmos cloud storage product, launched three years ago, has become an important product for EMC. The research project was a fairly big bet, said Reichman.
Overall, EMC's R&D efforts can't hold a candle to IBM Research in scale or prestige, but it's really not fair to compare the two, said IDC's Woo. "While IBM Research has spent a lot of time on the development of next-generation technologies all the way down to the component level, EMC has focused more on the systems level," he said.
On the Menu
Acquisitions are likely to drive EMC's technology investments, and there's lots to choose from. Cloud is king, and technologies that support hybrid clouds -- cloud gateways and federated workloads -- will become very important, said Burton. Companies in the cloud gateway space that might catch EMC's eye include Panzura, Nasuni, TwinStrata, StorSimple, Ctera Networks and Riverbed, said Reichman.
Tierless SSD architectures from Solidfire, Nimbus Data Systems and Pure Storage, and server-storage hybrids like Nutanix' technology are also ripe for acquisition in the storage industry, Reichman said.
And the reigning buzz phrase, big data, is also likely to factor in EMC's future, said Burton. "We feel like big data is where cloud was two or three years ago," he said. "You'll absolutely see us kicking around looking for technology, and potentially bigger M&A targets, in that space."
Photo: aliwest44/Flickr
