Many industry watchers have been speculating on the success of the previous two frameworks and what could be improved upon.

With the government’s announcement that it has launched the third round of procurement for its cloud initiative, G-Cloud iii, many industry watchers have been speculating on the success of the previous two frameworks and what could be improved upon to further break down those cumbersome, expensive, traditional IT contracts that the public sector is so used to dishing out.

No-one can deny that the G-Cloud has been a huge step in the right direction for the public sector’s use of cloud services. However, since the initiative launched less than a year ago, the government has spent just over £4 million trhough the G-Cloud (less than 0.5 percent of overall government IT expenditure). It is obvious that more work needs to be done to convince departments and local authorities that public cloud is the easiest, cheapest and most effective way to procure services.

One of the benefits of the G-Cloud is that the government runs a procurement every three to six months, which allows for feedback to be given by users, and for changes to be made accordingly next time round. So how is the G-Cloud team doing on its third time entering this phase? Despite being low on resources, Computerworld UK understands that the team has achieved a lot in 12 months, but there is still a significant amount of work to do.

A ghetto for SMEs?

Since its launch, those keeping an eye on the G-Cloud have been speculating whether or not Amazon and Google, two giants in the public cloud arena, will eventually be signed to the framework. Shortly after the first G-Cloud framework went live, director of the initiative, Denise McDonagh, said that she ‘fully expected’ Amazon to be on the framework in the near future. However, the second framework came and there was still no sign of either supplier.

Computerworld UK spoke to Joe Dignan, chief analyst at Ovum, who believes that it is essential to get the incumbent cloud providers on the G-Cloud.

“I think before this is going to be taken seriously, they at least need Amazon on there. I know the Cabinet Office is very keen to get both Amazon and Google on there, because what they don’t want is the G-Cloud to become an SME ghetto,” said Dignan.

“It has got to be seen as the primary procurement engine, and without those guys, it is becoming that SME ghetto.”

Dignan argues that the problem for these suppliers is that not enough sales are being made through G-Cloud to make it a credible option. Some might point to and argue that this isn’t the case; given it is a cloud heavyweight and it signed in the second round of procurement. However, Dignan argues that did this because it wants to be seen as being cooperative with the public sector, not because it expects much business through the G-Cloud itself.

“The issue for Amazon is that they have no interest in developing a public sector specific proposition – they see it as a commodity item. To make it work for them they need scale, and there just isn’t the scale through Cloudstore at the moment that allows them to use their low profit margin, commodity model,” said Dignan.

“They need to convince some large departments to purchase some big-time bundles of cloud and that will bring in the big boys.”

However, not everyone agrees. Simon O’Kane VP of Enterprise at British-born SME Huddle, which has arguably been the most successful supplier on the G-Cloud, believes that by including the likes of Amazon on the framework takes away from the government’s agenda of giving more IT work to SMEs and not traditional vendors. 

“There were a lot more companies on the catalogue in the second iteration (458, versus 257 first time round) and a lot of those were the more traditional IT vendors – the big SIs, software and hardware companies. The original vision for the G-Cloud framework was to promote genuine cloud services and value for government, as well as SME business,” said Kane.

When asked if he thought that by allowing Amazon and Google onto the framework this would isolate the SMEs, such as Huddle, Kane replied “yes”.

“I think it adds more confusion in terms of what message the government is trying to send out. If the message is that it wants to promote SME companies, especially UK-based ones, doing this will dilute it. Looking at the catalogue recently I saw a lot of the old names with multiple entries,” added Kane.

Sales and Procurement teams are scared

Ovum’s Dignan also believes that if the G-Cloud is to succeed, there needs to be further changes made in the internal government procurement teams and the vendor sales teams. Dignan has been made aware of some rather unsavoury practices being carried out in the public and private sector that could ultimately prove to be the nail in the G-Cloud’s coffin.

“I was talking to one major government department who said that they were about to buy services off the Cloudstore, but were then offered a better deal not to purchase through it. This is because the sales person wasn’t going to get a penny if the department bought if off the cloud framework,” said Dignan.

“It is entirely up to the vendors themselves to get their act together to make sure these guys are incentivised for sales that come off the Cloudstore. If this happened they would then be happy to push it. The Cabinet Office should also be encouraging the vendors to make these changes to their internal processes.”

He added: “It’s a threat they need to sort out pretty damn quickly.”

However, the vendors aren’t the only ones to blame, and the government also needs to take a look at its own internal procurement teams, who are likely to view the G-Cloud’s simplicity as a direct threat to their jobs.

“This has always been a problem for old government frameworks – the people who run the procurement processes aren’t stupid. They know that if procurement processes are made simple their jobs will go – they fight tooth and nail to keep it as complex as possible,” said Dignan.

“One of the things they need to do is go back to their internal rules, particularly at a local authority level, and they need to take them out. There needs to be a blanket ruling from the Chief Procurement Officer that says that the G-Cloud is legal, reasonable and right.”

Andrea Di Maio, Gartner analyst, also argues that departments and local authorities should be given more responsibility when it comes to the buying of G-Cloud services – specifically with regard to accreditation. Currently the government accredits each service offered only once and then any government body can reuse that service without going through the accreditation process again.

However, this is all done centrally, something which Di Maio believes creates a bottleneck and slows down getting cloud services in use.

“The UK has the same problem with its certification and accreditation process. The US has a Fed-ramp process, which allows each agency to re-use an accreditation – very similar to the G-Cloud. The US has only just announced its first accredited service, but it has been available on their website for almost two years,” said Di Maio.

The UK has accredited a number of services, but overall has a lot more services available to purchase than the US.

“The problem with the UK is the bottleneck, as you have so many vendors signed up waiting to get accredited. The government has been over-promising – it looks great that there is a huge catalogue of services to choose from, but then you look at the bottleneck. If you are planning to put in a new service, but have to rely on a centralised process to get it accredited, this could screw up your timeframes completely,” he said.

“The process is too centralised, there should be some level of delegation. So if a department wants a service, it should be up to them to get it accredited. Think of it as accreditation-as-a-service.”

Cloud First

When the US embarked on its public cloud strategy, President Barack Obama mandated that government agencies would have to look to the cloud first for services, and if these weren’t appropriate, only then could it consider on-premise solutions. However, this has been avoided thus far in the UK, where key government officials involved in the initiative disagree whether it is appropriate or not, Computerworld UK understands.

But, to get some proper traction in the G-Cloud, does the public sector need a ‘Cloud First’ mandate? Huddle certainly thinks so.

“[A cloud first policy] would send out a great message across the UK government. If it doesn’t do this, the G-Cloud has the potential to run out of steam,” said Huddle’s Kane.

“There is always a risk that if you don’t get fast enough adoption and drive a lot of change, it will fizzle out.”

However, the analysts aren’t convinced that this is the way to go just yet. Ovum’s Dignan believes that the G-Cloud model still needs to mature in government before the team starts mandating its use across the public sector.

“I don’t think they are ready for this yet. I think at some point in the future they might be, but I definitely don’t think it will be this year. They are still learning what works and what doesn’t when buying off the Cloudstore – they mandate too soon it would be a recipe for disaster,” said Dignan.

“Let the model mature and let them learn what does and doesn’t work. There are going to be a couple of screw-ups, and they need to learn from that. If they tried to do it now I think it would be a catastrophe.”

Gartner’s Di Maio agrees:

“I think it’s a pretty bad idea. I think if you look at what has been done in the US, it has been taken by most agencies as a compliance issue – they take the simplest, most innocuous thing and put it in the cloud,” he said.

“Cloud can be useful in some cases, but can be a mistake in others, so you need to support whatever kind of sourcing option a department might have. What worries me is the mindset; departments would do it because they would have to, not because they need to.”

G-Cloud iii launched this week and plans to build on the success of the initiative. It also plans to increase representation on the framework in new areas, including identity services, service integration and management, software support and business process automation.

The value of the framework has increased by £100 million, with potential spend now reaching £200 million. The framework will last for 12 months, as with the previous G-Clouds, but there is potential for buyers to sign contracts up to 24 months in length.