EMC at the crossroads

Most Wall Street analysts believed storage king EMC would finish the fourth quarter with a slight loss or break-even results. But much to their delight--and surprise--EMC last week said earnings would be between 1 cent and 2 cents on revenue of $1.47 billion.

You could practically hear some folks at the New York Stock Exchange break out into song as the stock shot up more than 23 percent. But happy days are not here again. The truth is that EMC's results were based primarily on successful cost-cutting. This remains a critical make-or-break time for EMC, which could face serious trouble if it fails to come up with a radically different business model before the start of 2004.

Why the negative outlook? Let's start with hardware.

EMC has thrived on sales of one product: Symmetrix. But the old workhorse has grown old and the days of winning 80 percent margin deals are over. Hitachi Data Systems offers superior technology and lower prices while IBM bundles its Shark subsystem into big services and server deals. The February introduction of a new version of Symmetrix will help but still won't be enough to reverse market share losses or boost margins.

EMC's midrange CX system is a great box, and its partnership with Dell will help, but CX still competes in a cutthroat market that includes Hewlett-Packard, IBM and Sun, where competition turns each deal into a low-margin battle.

EMC also relinquished its leadership in network-attached storage (NAS) to Network Appliance in the fourth quarter. In a segment dominated by smaller systems, EMC only offers a high-end NAS solution, called Celerra.

The company's newest strategy for picking up market share is to lower pricing while seeding the customer base for its software products that only run on EMC's proprietary boxes.

The company needs bold action, but what can it really do?

On the software side
Management's bold claims about the company's ability to profit from sales of software and services are a stretch. Although Gartner Group estimates EMC as the leader in storage management software, that is only true in terms of revenue. The company's two leading software products, SRDF and TimeFinder, are based on machine-level software written in microcode within the Symmetrix and Clariion platforms. In layman's terms, this is more like the functionality hard-wired into your microwave oven than the multiple CDs you buy for your PC from Microsoft. The software works, but it remains hard to add new features without ripping apart hardware along the way.

Symmetrix and Clariion have the same software functionality but each is incompatible with the other and the software can never be ported to other storage or server platforms. This means that if the industry shifts and storage software moves from disk subsystems to switches or servers (a distinct possibility), then EMC loses big.

EMC's software vision of a multilayered storage management platform is dead on, but the company will be competing against a cast of thousands including Veritas, IBM, Computer Associates, BMC and HP. These companies write code in C++ and Java and have huge organizations already skilled in software engineering, support and distribution.

As for services, it's hard to understand what EMC is doing here. According to IDC, the company led the market in the second quarter with $288 million in storage services revenue. Still, it's difficult to understand how much of this is pure services revenue, as each product installation carries a services fee with it (a software fee as well). Aside from this, EMC service includes a maze of third-party subcontractors, an Accenture partnership, junior engineers--and no clear strategy.

Up for sale?
The company needs bold action, but what can it really do? Rumors have circulated that EMC is for sale, with Dell, IBM, Sun, Cisco and HP topping the list of likely suitors. This won't happen. Dell doesn't do big acquisitions. It prefers to sign partnerships and wait until products become commodities--then it attacks.

As for the others, EMC's in-your-face, macho culture is a mismatch for button-down IBM or the Silicon Valley coolness of Sun. Disk subsystems are too far from core communications for Cisco, and does anyone really believe HP has the stomach for another big acquisition?

What EMC should do is spin off a storage subsidiary and create a new software-centric company and culture. This would allow the company to play by real software rules internally and in the field while it continues to shave costs and change the distribution model for its hardware.

What EMC should do is spin off a storage subsidiary and create a new software-centric company and culture.

EMC could jump-start this process by acquiring a reasonably sized management software firm with an already existing development model, partnering strategy and distribution channel. Rumor has it that EMC was blindsided and outbid by Veritas for Precise Software. With that opportunity gone, the company must open its wallet a bit wider and look at companies like Compuware and NetIQ.

The software spin-off strategy is unlikely. EMC will probably continue on its course and cherry-pick the market for distressed start-up companies or those with bargain basement valuations. Last year EMC bought the remaining assets of failed start-ups Cereva and Sanrise and acquired storage resource management company Prisa Networks. These companies may have intellectual property of value, but for EMC, this strategy is akin to a pedicure when an extreme makeover is needed.

Without a radical move to boost revenue and margin, EMC will be forced to cut manufacturing, real estate, and personnel costs to the bone. Not a pretty picture. If this happens, the company's fate will be a lot like that of another Massachusetts high-tech legend, Digital Equipment. Like Digital, EMC could fade from its leadership position into years of mediocrity.