I missed Doug Erwin's "7 reasons why web hosting is hot" session yesterday. If you were there, I'd love to hear from you - what did he say? What did you think??
Earlier in the day, though, I had a long conversation with Rurik Bradbury from Intermedia.Net about why he doesn't buy the hype. Everybody says the industry is in a great place. I, for one, have been dazzled by Rackspace's 59% growth last quarter. And Don Warnet recently mentioned that October was DataPipe's best month ever. (I *love* DataPipe's "operational empathy" tagline, BTW.)
But Rurik points out that web hosting is heading towards utility-land. Yes - ecommerce and SaaS and Web 2.0 and SOX/HIPAA compliance will continue to drive demand for data center space. Yes, Rackspace's Fanatical Support has been a winning formula. And yes, there's been lots of excitement surrounding grid computing. But let's fast forward to a point in time when web hosting reaches dial tone reliability. Once that day arrives, running a web hosting company won't be very different from running a power company.
The most important implication is, there will no longer be room for large numbers of small players. As a point of reference, you don't choose from a wide variety of electricity providers when you move to a new town. You just sign up with whomever the local utility happens to be.
In addition, there will be fewer opportunities for differentiation. Rurik has never called any power company, and I don't even know the name of my local utility. Fanatical Support doesn't apply in the electricity world, because the technology works too consistently
Last but not least, with no occasions for customer interaction, web hosting providers will become invisible to the average end user. Instead, application developers - such as YouTube's founders - will be the ones who drive innovation and build customer relationships. Continuing with the local utility analogy, do you give your power company much thought when you shop for new electronics? That, Rurik says, is why ideas people will have billion dollar cash out opportunities - and infrastructure people won't.
Rurik thinks Intermedia is in a relatively good position, because Hosted Exchange is more idea than infrastructure. On the other hand, the idea belongs not to web hosting providers, but to Microsoft.
So what now? We didn't come up with an answer. But right after our conversation, I ran into a guy who's getting ready to ditch the shared hosting model and release a bunch of hosted apps he's developed. He knows that the risk is much higher, but he's hoping his payoff will be as well. His business has been making good money, but to him, apparently, web hosting might not be hot enough.
Eventually, of course, every industry reaches saturation, when demand exceeds supply. Consolidation into a few large players follows, and they become as boring and undifferentiated as, ahem, a power company and a phone company. Sic transit gloria mundi.
But, before this happens, we need to reach saturation first. Andy Schroepfer from Tier1 and I independently estimated what is the percentage of all servers that are hosted today. We both came with numbers in the the 15% to 20% range. If we are not completely out of wack, this means that the hosting company has between 500% and 670% room to grow, even assuming that the total number of servers remains unchanged. But server sales are also growing in the 15%+ year to year range.
Assuming the hosting industry keeps growing at 50% annual rate, this means we are about 10-12 years away from saturating the market. Plenty of time to make money and have fun before 2016 when it will become, as Rurik righfully notes, a boring utility industry.
As a point of reference, I've been reading a lot about SK Telecom of South Korea. Like every cell phone provider, it's battling mobile number portability. And its execs no doubt weren't pleased about Google CEO Eric Schmidt's pronouncement last Saturday that cell phones should be free.
On the other hand, SKT owns Cyworld, a social network on which 96% of 20-something South Koreans are active members. Not only does Cyworld generate US $300K in daily revenue, it's also responsible for boosting SKT customers' average bill and additional service usage rate.
I'm not saying that every infrastructure provider should be building social networks, but I do think SKT has done a better job than most competitors in monetizing its assets and mitigating risks from the changing marketplace. Might there be ways for web hosting providers to do the same?
The parallels between electrical utilities (or utilities of any kind) here are a bit of a stretch. Utilities operate under government mandates and provide reliable service because their Public Utility Commission (PUC) forces them to, not because the customer can switch service to someone else. The end user in effect is the PUC and that creates a very different dynamic in the utility world.
Utilities by their very design have razor thin margins and cannot make more money with great customer service, so they focus on operational efficiency. Customers don't bother calling their utility not because the service is so reliable (thats just not true), but because they know its pointless. Our parent company just didn't get us and it was a train wreck from the very start because of our obsessive focus on customers.
Web hosting is going to become more commoditized and dominated by low margin services at some point, but it will take a while. This stuff is still complicated enough for customers to want to set their hair on fire, which is why support and expertise is a major differentiator in web hosting. Technology changes fast (RoR is going out of style any day now) and rarely reaches the maturity level that utilities are used to, so I expect successful customer service to dominate for a good long while.
While utilities have nothing in common with web hosting, parallels in other markets, like the cellular and deregulated telecom businesses, which serve as customer aggregators for a wide variety of value add services are far more relevant.
The opportunities for growth and scale in those industries are massive, customer churn is fierce and progressive players differentiate themselves on value added services and support (sound familiar?). Whether its ringtones and most feature rich phones for cell carriers, or security and business communication tools from telcos, there is plenty of ways to win a customer because you offer a differentiated product.
That's exactly what BT is doing while all the other carriers are stumbling all over themselves trying to enact net neutrality laws and tax Google into oblivion:
http://money.cnn.com/magazines/business2/business2...
You are right - hosting is heading toward the cell phone industry model. There is one thing that can make the dynamics different, though - unlike carriers, who depend on massive network buildup to deliver the service, a hosting operation does no need to spend much capital ahead of demand. This may change the economics away from consolidation - I am not sure that the big guys in hosting have inherent advantages...
Vlad