Today, there are about 100 small- to medium-size companies (20-100 employees) making storage racks and disk arrays, with sales ranging between a few million and $50 million dollars annually. They specialize in assembling components made by other manufacturers—hard disk drives, fan assemblies, RAID controllers, power supplies—and testing them before shipment to the final customer or systems integrator. These firms may install some company-specific management software, but otherwise their products are practically indistinguishable from one another.

While a number of these firms have proven successful to this point, if the history of the PC industry is any indication, the disk array industry soon may be headed toward consolidation and, in some cases, the elimination of the small- to medium-size manufacturer.

After IBM introduced the PC, thousands of small firms rushed to take advantage of its open architecture. These firms embodied a business model strikingly similar to that currently embraced by the disk array industry. They assembled PCs from off-the-shelf items, essentially designing only the layout of the components inside the case. These firms did not have any special technical competence, nor did they invest in any new innovations.

Any time there were technological improvements in any of the PC components—the microprocessors or the disk drives for example—they were simply incorporated into the product. These firms competed on the basis of price (mostly) and service (being close to one’s market), and catered mostly to their immediate neighborhoods. The software to run the PCs was not part of their operation (it was provided by Microsoft and others) and, except for the nameplate, the PCs themselves were hardly distinguishable from one another.

Fifteen years ago, there were over 3000 of these small PC shops in the U.S., and their market share was around 70%. There were no DELL, HP, Gateway, or Compaq. IBM was the only big player. Today there are barely 1000 such firms, and they have a share of about 35%. Many of the smallest companies have either disappeared or reduced their operations to maintenance and service.

Two major factors contributed to this shift. First, of course, is the natural consolidation that takes place in any industry. Over a period of time, the smaller firms either get absorbed by larger firms or simply close shop because they cannot compete effectively—on the basis of both customer acquisition and vendor pricing agreements. The second factor in the case of the PC industry was that very few firms had good technical resources. They were just assemblers of components made by other firms, which made them even more vulnerable as their stronger brethren emerged.

Which brings us to the disk array industry of today. For most small- to medium-size disk array manufacturers, the entire design effort consists of laying out components in a case while accounting for capacity, accessibility, power, and heat dissipation needs. The product is modular and the components are practically interchangeable.

This modularity does allow individual disk array manufacturers to assemble a wide variety of products by selecting different combinations of disk drives, power supplies, and RAID controllers. But, if these manufacturers are to survive where their PC predecessors have not, many important questions need to be addressed. The preeminent questions being—what is the technical competence of these firms? Do they have any original design capabilities in terms of developing new configurations? Do they have software capabilities that can be used to develop more efficient data storage and transfer systems? Or, are they just going to rely on their marketing capabilities to survive?

I have recently been talking with a number of CEOs of disk array manufacturing firms. These discussions have convinced me that these small- to medium-size firms currently have neither the software capability nor the hardware design skills to add significant value to the products they are assembling.

They are entrepreneurs who have experience in the storage industry. They are familiar with SAN, NAS, DAS and the types of interfaces available such as fiber channel, SCSI, IDE, etc. This knowledge enables them to design basic storage systems. However, as the system becomes more complicated they must rely on systems integrators to provide effective storage solutions to the customer.

As in the PC industry, this model works in an emerging market, but the disk array industry is now beginning to mature. Costs of components are dropping, capacities of storage and throughput are increasing (all at Moore’s law rate or faster), and margins are getting thinner.

With costs of storage running at less than a penny a megabyte, even large firms such as EMC and IBM that claimed special status by providing total solutions are being buffeted by rapidly falling profit margins as the hardware is increasingly commoditized. The focus and economics of the disk array industry are clearly shifting towards software and integration.

There is another interesting trend on the horizon—the move towards interoperability in the storage industry. Most components of a disk array system are specific to that system, or they have to be configured carefully to function well. Consequently, system integrators either make their own disk arrays or buy from vendors who make compatible arrays. In an interoperable world, instead of producing arrays for specific system integrators, and developing different designs for each customer, standardized products could be developed and sold to a much larger market. The competition in disk arrays would then be based more on quality, price and performance than on compatibility. Interoperability may actually benefit some disk array vendors, but competition in the array market could actually increase as well.

Given these factors, I see a number of trends that will affect the disk array industry:

• Many small firms will disappear, as they have neither the technology nor the market reach to compete effectively.
• Some firms will try to find a matching firm to acquire or an acquiring firm to sell to. This will lead to consolidation of the industry, but the focus of these consolidated companies will still be on basic disk arrays. Eventually, a few major firms will emerge as in the PC industry.
• Some bolder firms will try their hand at integration by acquiring firms that complement their strengths. They will provide a total storage solution by including hardware, software and implementation expertise.

In view of these trends, what should a small disk array manufacturer do to survive for the long haul? Unfortunately, the choices are limited. But there are a few things you could do, and I list them in order of ease of implementation.

• Sell out—You should choose this option if you do not have the resources or desire to acquire other similar companies and expand your market, or the technical competence to grow and offer expanded services to customers. To remain in business for the long haul you need resources to acquire and grow. If you don’t have the resources or the inclination to grow, then it is better to get out now while you still can.
• Specialize—Firms that are focused on providing storage arrays for a particular industry sector—such as small banks or retail chains—are positioned to survive a bit longer (the big guys may not come into these niches for some time). By specializing, you will still need to develop a total solution approach for that niche (see below), but you can do so without having to be all things to all people. This approach may allow you to thrive for awhile, or it may just buy you time to pursue the other strategies described below.
• Consolidate—If you have the resources and the inclination to grow, then you should start looking at other similar firms to acquire and expand your market. The market expansion could be either geographic or within product lines.
• Develop a total solution approach—This is the approach for the ambitious and the hardy. It requires investment in acquiring technical skills to complement the ones you have so that your company can develop and provide an integrated solution for the customer’s needs. You need to either employ additional technical staff to develop the solutions and implement them at the customer’s site, or get the needed skills by acquisition and consolidation as suggested above. You will need to focus on providing total solutions and acting as a general contractor.

These options may seem drastic, but as the PC industry illustrates, small- to medium-size manufacturers that don’t evolve with the industry risk becoming part of its history.

Dr. R. Balachandra is a professor at Northeastern University's Business School. His research and consulting focus on technology management, management of R&D projects, and new product development, and he is the author of “Early Warning Signals for R&D Projects” (Lexington Books). Currently, he is a Visiting Professor at UCSD, where he is associated with the Information Storage Industry Center (ISIC). At ISIC he conducts research on issues such as modular design and interoperability.

To unsubscribe, send an e-mail message to listserv@ucsd.edu with unsubscribe isicnews-l in the message body.