One of the big negatives surrounding the Dell/EMC merger is that it takes two legacy companies that don’t appear to have a competitor for AWS and puts them together without directly addressing this seemingly big shortcoming.
However, both Dell and EMC are executing a similar strategy of creating solutions to sell to cloud services vendors rather than becoming cloud computing providers themselves. Both paths—offering services and selling to service providers—have hazards. With AT&T we saw the "do-it-yourself" path flame out rather badly in telecom, while selling in that market has proven, over the decades, to be far more lucrative.
The old AT&T was likely the largest vertically integrated vendor in a similar space we have ever seen short of Amazon. There are two problems with this strategy at scale, and we saw both do horrid things to AT&T. The first occurs if the company stops innovating and focuses on milking their customer base. This will undoubtedly result in either competition or regulation, and AT&T got both. In addition, it can lead to customer disloyalty and anti-trust action—and AT&T got both of these as well and eventually failed.
So, in extremes, vertical integration can result in company failure. We didn’t just see this with AT&T; we first saw it with the original Standard Oil and then later RCA. All three of these firms eventually failed. (The AT&T we see today was an attempt to recreate the old company, and Standard Oil is simply one of the retained names from when the original company was broken up).
Getting to dominance isn’t very easy either, particularly in a mature market. Standard Oil, AT&T, RCA and IBM all started out more than equal as their respective markets grew and were able to maintain leadership pretty consistently until they achieved dominance. Even Google, which you could argue is in a similar position, came up as the search market matured.
Today in Web services, Amazon is the closest to the AT&T/RCA class, making it very unlikely someone other than Amazon could reach a position of such dominance. It could be done—particularly if the market makes a turn as it did against the mainframe for IBM and against Windows for Microsoft—but it is really difficult for an existing company to time this right. Often a new startup (Compaq/Dell/Sun for IBM, and Google/Amazon for Microsoft) is more likely to take advantage.
A technology supplier, if they can hold onto critical mass and demonstrate value in terms of both capability and economies of scale, is in a less-risky position. This is because they can sell to the firms until (and if) the market settles on one vendor. The cloud vendor in this case focuses on their customers, and the parts supplier focuses on the needs of the cloud vendor. Each is optimized for their own focused areas. The parts supplier is free to move to other markets as they mature, and because they are end-supplier independent, they are more secure against the failure of any one cloud services provider.
So the problem with the vertical integration path is that we already have Amazon Web Services becoming dominant. Catching them from behind will be incredibly problematic. The problem with becoming a supplier is that cloud service providers like Amazon are often not seeing the value of buying a solution from an OEM; instead, they are often working directly with component suppliers to build their own solutions.
IT has flirted with "build-your-own" on and off for decades. It always seems to come back to the OEM model because the OEMs have been able to demonstrate both competitive prices and unique technology that better assures the stuff works. OEMs also provide someone IT can point to when there is a failure, and cloud services have had a lot of embarrassing failures of late. So OEMs provide the dual sustainable benefit of both assuring the system is reliable and making sure the IT organization doesn’t get shot regularly when there is a problem.
Both of these advantages should transition from IT to cloud providers like Amazon over time. At least that would be the bet for any company not going into cloud services themselves.
In short, being a supplier makes better use of the skills an OEM actually has, while being a cloud services company would mean they would have to develop a whole new skill set and then chase a firm that is already expert. Granted, Amazon is having issues with both services and culture that suggests you could move against the company, but this would also suggest a strategy of selling to them could work. Here too an OEM would have the stronger skills.
There are two approaches to the cloud market. One is to develop or buy a cloud services vendor and compete directly with Amazon and their peer firms. The other is to build solutions that make these firms more successful and perform the role of kingmaker.
Both paths have advantages and disadvantages, but the supplier path is closer to the skills Dell and EMC have already. This is why they are both on, and will likely remain on, this path. VCE in particular presents a powerful alternative to the build-it-yourself approach that cloud providers are using, and Dell in particular should be expert at selling against that strategy.
The strategy that both EMC and Dell are currently using, and are merging to enhance, comes out of their existing strengths and should be less risky. But like any strategy, it’ll come down to execution, market conditions and competitive issues (against cloud services companies building their own stuff and against competitors building their own cloud solutions).
One thing will be sure: the cloud computing market is about to get a lot more interesting.
Photo courtesy of Shutterstock.