Justifying the cloud AccMan

Paul Miller has an intriguing post citing a study (PDF) into cloud computing adoption:

Also interesting was the relatively small impact of the economic situation upon Cloud adoption, with only 13% suggesting it had ‘helped’ adoption plans and 58% reporting ‘no effect.’ In my conversations with Nick Carr and others, there’s been an underlying presumption (on my part, as well as theirs) that cost-saving arguments with respect to Cloud Computing would prove persuasive and compelling. It would appear not.

Bear in mind this was a study among larger companies so the same drivers may not be true for smaller businesses where cost is often a deciding factor, albeit the analyses I’ve seen are far from scientific. Paul juxtaposes his piece with a link to an IDC report which says that cloud computing should be seen as a stop gap:

“Cloud costs need to come down much further to be a realistic long term option,” said Matthew McCormack, IDC analyst, at the company’s recent Cloud Computing Summit in London. “It could be useful in the short term financially for companies with severe cost overruns.”

“Your datacentre would have to be really poorly run for it to be more expensive than cloud in the long run,” he added.

Businesses would be wise, he said, to look at their capital and operational costs over at least the last five years, to analyse their datacentre workload efficiency, and to take a long term view, before making any move to the cloud

I’d love to know what sort of numbers IDC is talking about or how the calculations are done. I’m betting there’s a swathe of factors they’ve missed or under-represented.

There is plenty of evidence that cloud based data centers are an order of magnitude cheaper to operate, hence why Amazon has a rapidly growing business with its ‘elastic cloud.’ Listening to my colleague Vinnie yesterday, cloud centers are way cheaper to run than the traditional ‘cold war bunker. As in 3x to 10x cheaper. This was something Robin Bloor explained the other week when I attended the Netsuite event. So how IDC thinks that cloud costs need to come down (even though Moore’s Law suggests that will happen anyway) is beyond me.

In the conversations I’ve had, while cost is definitely a top of mind factor, others are equally or more important. Such as the ability to transfer cost from capex to opex, the ability to easily provision for collaboration between business partners, fast track provisioning for new applications (which can slice months off deployment) and the ability to rapidly scale up and down as circumstances dictate. Those reasons are not restricted to the larger users but run right across the board though as I said earlier, cost matters a great deal to the smaller business. There are of course nuances but these secondary and some might argue ephemeral benefits weigh heavily in favour of the cloud.

Tagged as: Amazon.com, Cloud computing, Data center

Nara

I think there are some fundamental misunderstandings out there about the cloud and why it's cheaper or better.

If you talk about IaaS, I can see how as datacenters reach scale and are appropriately virtualized that the economic advantages start to reduce or go away. However, there's still the benefit of almost instant provisioning of new compute resources which can make a huge difference in terms of time to market or the ability to do things that couldn't have been done before.

At the PaaS layer, the benefit is much faster development time, scalability without significant incremental cost and lower overall TCO.

At the SaaS layer, we can argue about TCO but there's no argument about initial costs being lot lower and typical much higher user adoption.

We'll see how the arguments pan out over time but it's weird to have IDC make statements like the one above at the same time that another part of IDC does research showing that development on the Force.com platform is 5x faster and less than 1/2 the cost of traditional apps.

share flag

spam
offensive
disagree
off topic

Dennis Howlett

I'm glad you brought up the 3 aaS layer argument because in recent times there's been a lot of conflating around saas/on-demand/cloud (hence my saying SOC somewhat tongue in cheek.)

I'd be interested in seeing your rational for IaaS because all the metrics I'm seeing say bunker style data centers don't catch up anytime soon without significant re-engineering.

As regards IDC - that sounds a bit like my 'double dipping' comment the other day (lol) but again, I'd like to see their numbers.

alastair

I'm interested in the evidence, but then I am a (particularly cynical) accountant - to my mind the only persuasive evidence is economies of scale. Perhaps the better message than its cheaper would be why its cheaper.

I guess that part of the cynicism in this argument is that its not about cost, its about what you have to pay the supplier.

"Expertise" or "Fast Track" is an argument that works in supply chain management, and arguably "cloud" is more complex so lends itself better to this sort of argument. But it only really works when you have either a track record or a particularly gullible target.

It depends how you want to define 'economies.' I could for instance argue there is a value in collaboration, there is definitely a value in fast track implementation (think less consultants, less or more timely training, less messing around with processes building) so perhaps the better measure is not pure cost...in reality that's a movable feast but about value delivered.

I'm loathe to put calculations out there because every case is different and in any event the broad cases I have seen in the public domain are at best 'questionable' leading down all sorts of ratholes.

But to continue with broad strokes, fact remains Amazon can deliver storage (for example) at a fraction of what you'd pay an HP/IBM/Accenture and with better availability. That's just one element of many.

The amazon case is interesting in that it breaks one of the "rules". Are they competing on cost or are they differentiating? I think it is the latter - the rule is you compete on one or the other - never both. You have had conversations on here about how accountants muddle cost and value. To my mind this is the same error.

Recent Stories

Recent Comments

Cloud Computing

Subscribe to all comments

Interesting Links