Rackspace knows a lot about the cloud. Maybe they know more than the pundits do, and very possibly more than the consortium of investors (led by Apollo Global Management) that are taking them private. The question now is whether they know more than those who think that the managed cloud services space is a great, latent, independent, opportunity.
For at least two decades, the world of technology has been driven far more by hype than by reality. Arguably this started with the Frame Relay Forum, which was essentially a marketing promotion activity hiding under the label of “standards group”. In most cases, including that of frame relay, there was at least some real substance underneath the covers, but the hype tended to blur things so much it was hard to ferret it out. The cloud is like that today. There is no question that the cloud has substance (no pun intended) and in fact it could develop into the IT advance of the age. There’s so much hype, though, that the market often responds to that more than top opportunity.
Rackspace was a hosting company when the two-decade hype-dominating period I just referenced began. It transformed itself into a cloud company as the cloud wave exploded, but here I think the hype started to catch up with it, and the latest going-private step is just the latest reaction.
Being a cloud provider of any sort is putting yourself in a vice. On the one hand, cloud services are attractive to the extent that they’re cheap—cheaper than traditional purchased IT devices. On the other hand, you have to make a profit to stay in business, so you have to sell the cloud for more than it costs you. If your clients are big companies, you can expect their own economies of scale (which rise according to an Erlang C curve, meaning it rises slowly, then more quickly, then plateaus) to nearly match yours. How do you respond?
Managed cloud services, meaning professional services offered to facilitate cloud adoption. These can admit a company to the sacred field of cloud computing without requiring them to invest in massive infrastructure in competition with others.
Professional services and managed cloud could work, but it faces an early and serious challenge—marketing it. You can’t sell managed cloud door-to-door like it was lawn service. That’s particularly true with SMBs, because the SMB market is a very large number of prospects with a small sales value each. Imagine knocking on a couple million doors. No, the customer has to come to you because of marketing, and where Rackspace fell down was in its ability to stay in the public eye. Amazon got all the good ink, and companies like IBM were able to leverage their brand and account presence.
Well, that explains why the Rackspace strategy didn’t take off. Their stock peaked in early 2012 and has generally trended downward as “real” cloud momentum built. But taking a company private is a bet that they’re seriously undervalued and you can turn things around. Is that true? Actually, it might be.
Like most tech companies, Rackspace has undervalued marketing. The hype around things like the cloud makes it seem to vendors that there’s some great ecosystemic tidal wave sweeping buyers toward inevitable adoption, and nothing need be done but toss a windmill in the stream and use it to drive a press to print money. “Build it and they will come,” as the saying goes. Had Rackspace been smart in marketing they could probably have built a business to rival Amazon.
That was then. What about today? Well, the truth is that the cloud is kind of old news. Every day, reporters have to write stories people will click on. “The Cloud Saves You Money” might have worked 20 years ago, but it’s been said too many times now to generate much interest. If the only thing that Rackspace and its new investors do is better marketing against the traditional cloud messages, I don’t think they’ll succeed.
Obviously, then, there would have to be other things that might make this deal sensible, and there might be two of them. One is the systemic change to cloud usage that we’re now starting to see, and the other is the potential entry of a bunch of new cloud aspirants, some of whom might be happy to do some M&A.
IaaS stinks as a business model, and even Amazon knows and is proving that. There’s minimal differentiation possible, and because all you’re doing is hosting a VM you’re not displacing much in the way of opex. The problem is that PaaS in the true model (Azure, for example) isn’t as versatile in addressing opportunity. What’s emerging as the alternative is creating an ad hoc PaaS by adding a set of web services to IaaS. These web services, accessible to applications that are built to link to their APIs, can provide horizontal or vertical tools to build cloud-centric or cloud-specific applications. Because customers pay for their use, they add to provider revenue and that’s a direct benefit.
The indirect benefit for a managed-cloud provider is that the web services can facilitate application-building for customers, for specialized VARs and resellers, or even for the cloud-MSP itself. For example, Amazon and Microsoft, both of whom offer these services in their clouds, include tools to build IoT applications. These kinds of tools could be expanded to provide a much bigger slice of application functionality, which would make them even more attractive to VARs or SMB users. From them, Rackspace could even build a kind of IoT shell application that could be customized as needed.
The challenge for Rackspace, as a “cloud services provider” rather than a cloud hosting company, is that they really don’t have a great place to put such services. They do have servers for hosting, but not the kind of structure optimal to host web services. They resell other providers’ public cloud services. That reduces their chances of profiting directly from hosting web-service enhancements or specializations like Amazon or Google or Microsoft. Thus, the goal has to be developing super-skills that value you highly. Those skills are credible only if we step beyond IaaS to web services.
However, leveraging skills doesn’t always provide a fast path to profit. The challenge of marketing remains, and so does the challenge your own cloud-host partners could pose if they decided to get into the same space. Which, if it’s a good one, they should. That’s what raises the second issue, of selling yourself off. To exit going private, you can either go public again at a profit, or sell out.
The telcos are the only players in all the world who actually have a path to building massive distributed clouds that could fully open the opportunity for new-style cloud applications. They’d be crazy not to want to exploit those kinds of applications to build value on their clouds, which is why I think “carrier cloud” is the real goal for telcos and not “SDN” or “NFV”. Exiting to a telco could be a very smart play for Rackspace, providing that Rackspace builds the essential skills to validate the carrier cloud—then sings the song beautifully and builds brand recognition there.
Here, the problem is that telcos already have favored integrators—the integration arms of their big equipment providers and particularly Ericsson, who has been almost a professional services company with limited equipment sidelines. Why would telcos want to buy a telco professional services company when so far they’ve been content to hire one? Rackspace would have to answer that question the only place it can be answered—in the media via marketing and positioning and singing pretty.
That brings us full circle, doesn’t it? Rackspace didn’t push themselves effectively, and still doesn’t. They now need to develop a new asset to push, and expertise in “fog-cloud” applications would surely qualify. They can then position that new asset and make themselves attractive as a target. All of this is possible, but it’s going to take a lot of enlightenment to drive it. Do they have that? Look at Dell, who went private. With carrier cloud the largest single server opportunity out there, has Dell been revolutionary in their offerings or positioning? Not by my standards for sure. Can Rackspace do better? We’ll have to wait and see.