Now that Microsoft has launched additional services for Windows Azure, providing a generic, scalable service for hosting websites, and has improved support for running generic virtual machines, things are bound to change for the entire hosting market.
The point is, Amazon has offered virtual machines for ages with EC2, and we all know how successful they are. Amazon is, however, sort of a black swan. So far at least. Now there are 2 huge players in the space, with one – Microsoft – being very diversified in terms of what it has to offer. Basically Azure, like EC2, can host anything but, unlike EC2, it can do so in a variety of ways, with a variety of levels of control over the underlying (virtual) hardware and infrastructure.
I’m not familiar with EC2 pricing, but I can see it’s not much different from Azure (it seems to be slightly more expensive, at least for Windows machines). In both cases hourly pricing favors intermittent, or fluctuating usage of the service. The idea is that you can easily spin up more “virtual hardware” just as you need it, even for a couple hours a day, and then shut it down. This is very powerful, and surely much more convenient than static monthly or yearly up-front billing.
Typically, long-term contracts give you some kind of advantage if you keep using the service – phone, power, connectivity, hosting, you name it – over time. The truth is, Amazon’s and Microsoft’s hourly pricing is cheaper than long-term billing offered by old-school VPS competitors. Even worse, it’s cheaper than hourly pricing offered for similar services – Rackspace and GoGrid come to mind.
Microsoft and Amazon are able to achieve such economies of scale that it’s almost certain that they’ll crush smaller providers. Both are diversifying their offering, and keep lowering prices, and I don’t see any way for others to keep up with the breakneck speed at which such giant enterprises are pushing forward, not to mention the huge marketing resources available to them. Microsoft is also able to push the Azure brand through their developer products and events, and I suspect it will eventually pop out even in consumer products – most probably Windows 8.
I have been a loyal SolarVPS customer since, if I recall correctly, 2009. That’s a whopping 3+ years, which is a lot in the IT field. The problem is, the more I think about it, the less reasons I see to keep using older, smaller providers, even when it comes to cheap shared hosting. The one I recall and sort of stands out from the crowd is still Rackspace, but it’s quite expensive and, frankly, I don’t pay for “Fanatical Support” when I’m pretty sure I won’t be using it.
I already mentioned this in the past: I surely put more trust in Microsoft’s or Amazon’s capabilities of building and running world-class datacenters. I am certain that Microsoft can run Windows servers better than anyone else, and despite the outages Amazon suffered in the past, I’m sure they know better than most – if not all – other providers. As with any other commodity – because cloud hosting is becoming a commodity – standardization is king and the war is on price and breadth of reach.
Perhaps I’m wrong and the expanding market will leave enough room for everyone, but I suspect we’ll start seeing changes very soon, as small providers start to struggle for air. The only area where small players can – in theory – win some ground is customer support. Despite Azure support is relatively good and fast, you’re still interfacing with a giant behemoth of a corporation. The feeling is not entirely good, but as I said it’s non-issue for me. I’m perfectly fine when I’m left alone with a decent control panel.
All of this, of course, is valid for hosting services. For end-user services, I don’t see Microsoft or Amazon making a big dent in the market. Surely they’re able to absorb freemium pricing costs much better than others, but I’ve the feeling that they’re not able to innovate much. Just look at Microsoft SkyDrive, Amazon Cloud Drive, Google Drive. They can’t even innovate on the damn name. Then look at Dropbox: another story altogether.
No, those lame attempts won’t kill Dropbox. Instead, big players will provide the infrastructure small player will use to provide services to end users, and perhaps we’ll see former hosting providers turning to service providers, running on Amazon or Azure.
Comments
2 responses to “Will Large Cloud Providers Kill Smaller Ones?”
Back in 2008, I was already predicting that cloud computing was for big players. Apparently, I wasn’t too off. :-)
Yes probably what’s hard to predict is the timing at which new products are introduced. At this point, however, I can’t see anyone else entering the market. Sun – or rather Oracle – is less relevant than ever, and seem more interested in lawsuits than building technology. Google App Engine seems to be sort of dead, and I can’t think of any service that runs on it, although there are rumors about something new being launched at I/O in a few days. Amazon is a bit silent lately, so perhaps they’re preparing something new. Very interesting future! There are smaller players that are struggling to build something more scalable than “regular” VMs, and perhaps they’ll tap into their existing customer base, but the big ones are surely going to win.