Maybe the world isn't ready for cloud
Summary: Though the arguments for cloud adoption are often compelling, buyers are often too mired in old habits to move forward
I heard some utterly compelling arguments for adopting cloud computing at the recent Business Cloud Summit in London. The strides being made by the UK public sector through their use of cloud computing and applications make a seemingly irrefutable case for going cloud.
- Liam Maxwell, Deputy CIO of the UK Government, cited a torrent of examples of how IT procurement had gone wrong in the past, such as a project that had that cost £55 million ($89m) yet only had 79 users, or a system where each transaction ended up costing £790 ($1280) to process. All of this adds up to the astonishing statistic that the UK spends one percent of the country's entire GDP just on government IT — "on pushing paper around in government," as he put it. It's a number so huge that several reporters have assumed he must have meant IT spending throughout the economy, but I've checked and this figure is the public sector proportion alone — in the region of £14 billion ($22.7bn) a year. Around £1.2bn ($2bn) alone is spent just on hosting.
- Maxwell explained that the aim of the government's G-Cloud initiative was to dismantle the prior system of complex, lengthy procurement processes that effectively gave large suppliers a license to create these massively costly projects. Now that public sector buyers can browse the CloudStore to source pre-accredited cloud providers and services, there is competition among a much wider spectrum of suppliers. He gave the example of a project that had been quoted from a traditional supplier at £57 million ($92m). On being told that the same could be sourced from CloudStore at less than £1m ($1.6m), the supplier provided a new quote of £2m ($3.2m). "The cost savings we're able to identify at the moment are enormous. Those have come through CloudStore," he said.
- Dominic Campbell from 20-person small business FutureGov (SME) described the impact of its Patchwork cloud app, which allows more frictionless co-ordination among different workers in multiple agencies involved in child protection in the county of Staffordshire. Putting the app on CloudStore overcomes procurement barriers, while using the cloud to connect care workers, teachers, police and other professionals allows them to share crucial information in a timely way.
- Mark O'Neill from the Government Digital Service said that he used to manage 120 separate servers for a user base of 500, and yet for what he used to spend on each server he can now have thousands of servers available in the cloud. He went on to cite an example of secure hosting for a highly sensitive project being sourced from G-Cloud for one-tenth the cost of a traditionally sourced alternative.
Yet despite the cost, convenience and process improvements available by adopting CloudStore solutions, buyers are still resisting attempts to persuade them to use CloudStore. It's as though they feel the promised benefits are too good to be true, or that it's simply too easy and there must be a catch.
Denise McDonagh, Home Office director of IT, said that many public sector buyers were so used to the complex red tape of established practices that they find it hard to break out of them. Many simply assume that using CloudStore is somehow against the rules that apply to their own organisation. One buyer who had complained CloudStore was too difficult to use was found to be going through 47 separate steps imposed by internal procurement processes, most of which were irrelevant for CloudStore. McDonagh spoke directly to buyers in the audience, seeking to reassure them: "G-Cloud is legal, it's here, it's an open competition, therefore it's very, very competitive ... What can be commoditized should be bought from the cloud."
The problem is that the vast majority of people still seem to believe that cloud somehow isn't for them. Despite the growing maturity of cloud solutions and the backing they receive from thought leaders across government and industry, the world is not yet ready for cloud.
Meanwhile, even such well-meaning initiatives to speed take-up of cloud run the risk of getting derailed. Maxwell warned of the danger that the recently announced European policy on cloud could reintroduce barriers to competition: "We as a government have a strong disagreement with the EU cloud policy," he said. "We don't think recommending a single certification will work. It will drive government back into the arms of the big SIs."
One of the European Commission's lead policymakers on cloud, Ken Ducatel, was in the audience and later explained that, "We're just trying to make sense of what's going on. It's not our intention to impose certification schemes." But the committee set up to advise Digital Agenda Commissioner Neelie Kroes on a framework for public sector cloud procurement is dangerously skewed towards established big ICT suppliers, including Accenture, ATOS, SAP and Telefonica.
So even when the argument for cloud seems overwhelming, it may still fail to win ground against the reticence of cautious buyers and the inertia of established processes and power relationships. There is still a long road to travel before this battle has been won.
As a reminder of what we'd miss out on if cloud gets bogged down like this, I'll leave the last word to Mark O'Neill and his vision of what IT procurement could become:
"This is about fundamental business disruption. G-Cloud, the CloudStore, is starting to become, not just a store to provide IT services, not just about the technology stack, it's becoming more and more about meeting a particular user need. I'll be able to fulfil business needs with services from the CloudStore. That's a fundamental shift."
Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.
Talkback
What about the Network?
He had to admit that they are now having to look at improvements in that area. I'm surprised that this isn't addressed more often.
Oh I'm not suprised.
So so true!
P.S. I'm not about to allow all my data to be stored on cloud servers where I can't control it. And when these massive servers fail they often fail to take responsibility for the loss of data. Plus, considering how slow the computing world has been to adopt widespread use of encryption, I don't want them having control of my data or ability to look at my data without my permission (Dropbox for example is able to look at one's data without your permission but there's absolutely no reason for such a setup).
Or maybe the cloud isn't ready for the world.
inertia of established processes and power relationships
If the initial price was 47 million, and it was reduced to 2 million, where do you think the difference was going to go? It was going to be split between the contructor and different goverment officials.
Just do not be afraid of saying 'Corruption' loud and clear - after all Britan is democracy, and you should not be afraid, should you?
The reason is blindingly obvious
To do so defeats the concept of capitalism
One unified standard set up to keep downtime minimal, and not housed in one region... But then, globalization is only used when it is convenient to those doing it...
The problem with Religion...
Look, the arguments you felt were "compelling" were compelling to you because you've already bought into the idea. They reinforce your preconceptions. However, to people who do not agree, they may not seem compelling at all, and dismissing those concerns as 'holding onto old ways' is disrespectful and doesn't really move the discussion forwards.
There are lots of very basic, equally compelling arguments against SaS:
1. You have to constantly pay for something rather than getting it one chunk.
2. You may not be able to control updates.
3. You lose control of your data.
4. You can't access your software when you're off the web (which happens a LOT if you don't live in downtown San Francisco or some other major *American* city.
5. You become tightly coupled to a vendor you can't control. If they go out of business or just don't want to support that SaS product anymore - you're screwed because not only have you lost your software (which isn't the case with local software), you've lost access to your data.
There are plenty more reasons why SaS is an iffy proposition for any serious or long term use, but these are the biggies.
and then there are counter arguments
So here are the counter arguments
1. The cost of a server itself is a drop in a bucket in the overall cost. You are forgetting the cost of sysadmins supporting the servers (with gov salaries, benefits, and pentions), power backup systems (and their maintanance), airconditioning systems (and thier maintanance), rent for the physical spce where the servers are set up.
And then there are costs of replacing outdated servers with newer ones. At some point MS or Oracle or whoever is your software provider will tell you that the new version of their software requires newer hardware, and they will stop supporting the old version within twelve months.
I may be missing something, but there is a lot of labor involved into maintaining a comm room.
2, 3, 5 - these things just need to be writtent in contructs. It is not like you or me signing up for some free google service. Paying customers have more influence onto their suppliers, and they have lawers to draw contracts for them.
4 - you are mixing up wireless and SaS. This is about the gov contracts with SaS providers. It may or may not include wireless access.
And then there are counter counter arguments.
Yeahhhh, not really....
4 - I did mix up SaS and wireless, more or less deliberately. There really isn't a difference as fas as the basis to my arguement against them goes. Can't reach the 'net? Screwed. Data invaded? Screwed. Provided outage? Screwed. Unless there is some way for ME to have final control over MY apps and MY data, cloud will never be a component of MY critical infrastructure.
That is because "cloud" is nothing new.
Pagan jim
Is that similar to Mind Control?
It's not about old habits, but about PROTECTION
The cloud promises much, but as we saw with Google, a fall -- and they have to happen -- means many people unable to access. A delay in email access is one thing; a delay in mission-critical information is quite another.
Then there's the problem of hacking. Granted, we all live with this daily, and many are the tools to obviate hacking. But hacks still occur, so the more dependence there is on the cloud, the more hacking can be devastating.
Same is true for anything in life. You BACK UP your computer, ideally in multiple sets. If it's really important, you also PRINT the material.
So cloud storage and computing offer an added layer of flexibility, and maybe even backup, but the risks must be recognized. For some of us, the risks are too great. I don't even store my email addresses in gmail. My clients have suffered email hacks at AOL, Comcast, Embarq, Yahoo, LinkedIn and Facebook (which hacks your email address book even after you quit). So what would they think if I stored any of their info in the cloud? I'd lose them as customers.
Pen drives are easy to carry. Yeah, the risk is you can lose them. I submit that risk is lower.