Despite Microsoft Azure entering the cloud computing market later than competitors, such as Google and Force.com, its launch is likely to make large ripples, proving that first come is not always best served.
"If you look at this on a relative timescale, Microsoft is late in bringing this offering to the market. But if you look at it from an adoption perspective, there have been a lot of offerings available before Azure, but not a lot of adoption," says IDC Australia's associate director for research and consulting Linus Lai.
According to a Longhaus survey released last year of 110 medium to large Australian companies, only 9 per cent had adopted any form of cloud computing.
"Microsoft is entering at a time when we are still at an early adoption stage for the cloud in Australia. So that means they aren't really late to the party," Lai says.
One such developer is catering specialist CaterXpress. Director Anthony Super says that CaterXpress is currently in the process of moving all its clients using its apps from on-premise servers to Azure, which will mean being hosted on Microsoft's servers in Singapore.
According to Super, this means a 60 per cent saving in costs for the company. "You're looking at the monthly expenditure on running our own equipment at the moment, not even to mention the labour costs on top of that versus a fully managed version of Windows Azure," he says.
CaterXpress has had to ask its clients if they were willing to have the app hosted overseas, but Super says there has as yet been "no hesitation" because of the more rigorous redundancy and back-up Azure offers. Geography isn't relevant, he believes. "I guess because we deal with small-to-medium-sized businesses those sort of questions just aren't asked."
If some of the customers don't agree, CaterXpress will simply keep them running on the in-house servers, but those customers won't be receiving the cost savings that CaterXpress will pass onto the others.
Its local offering is aimed at Australian enterprises with 1-2000 employees and up, according to Fujitsu Australia general manager solutions Cameron McNaught.
"We see a huge advantage of investing and deploying the cloud in Australia. The feedback we had when we did our market testing towards the end of last year was that the domestic angle was extremely important and the number one factor for customers deciding whether they would use cloud or not was where would their data be located," he says.
It isn't just non-critical systems that companies are considering moving. "We're seeing a lot of interest around not just development and testing environments but also around production IT moving into cloud."
McNaught says that when comparing unit prices, a local-based cloud will go for a light premium, but if bandwidth costs are taken into account along with other costs of having an offshore cloud, he thinks it will be on par.
"We see it as a different market because your sign-up method is through credit card. The customers that we'll bring on with [our offering] will be customers who are verified by Fujitsu and then come into an enterprise-ready cloud," McNaught says. His definition of enterprise-ready is having high availability, security and dual datacentres providing the service. Fujitsu's cloud availability is advertised as 99.999 per cent.
Yet Fujitsu's model didn't impress Cloud Central's founder Kristoffer Sheather. "I think that model's old-fashioned really. The new way of doing things is low friction. The pricing's out in the open. Everyone knows what everyone else is getting and you can actually sign up to the services very rapidly. Cloud should be about getting on board rapidly. If you have to go through a negotiation process that sort of takes the nature of cloud away. I'd sort of question how cloudy that really is," he says.
Cloud Central's uptime is 99.9 per cent, although customers can negotiate higher availability. Sheather's customers are small- and medium-size enterprises at this point. He is also looking to expand into local government.
Such firms often don't have defined service levels, whereas Microsoft does, Higgins says. If they don't increase their economies of scale, the non-Microsoft route might be their only path, the analyst believes.
"It would not surprise me if we see some merger and acquisition activity in the local hosting market as more and more larger players enter the market like Microsoft," he says. "That said, many of Australia's larger hosting firms still have the advantage of local datacentres, which will continue to appeal to many local firms in the short term."
Google's success in promoting adoption cloud-based services such as Gmail and Google Apps, especially in universities as well as in companies such as AAPT, makes it seem like it should be well ahead of its competitors with its App Engine. Yet Google might not be fighting in the same ring as Microsoft, according to Higgins.
"We hear very little about Google App Engine from end users or local vendors," he says. Higgins goes on to say that Google's choice of languages excludes many enterprise-level apps that might be expected in the cloud. "Google App Engine is very much just an invitation to write code 'the Google way' in the cloud," he says. He believes Microsoft is more flexible with what its developers put on its platform.
And, as with any other development platform, keeping a broad range of developers happy is key, or the platform will be empty. So, despite its lateness into the fray, adoption and reach may mean Microsoft is onto a good thing. Add its reach to the fact that other companies, such as Salesforce.com, VMware and Amazon, are still bringing out new cloud offerings, it seems the game is not over yet.
