The Cloud First term first entered the public sector lexicon in the US, of course, in the early days of the first Obama administration.
What was interesting in the wake of that was the reaction to this mandate from Federal Government from state level CIOs and IT decision makers.
While the center said ‘this is a must do’, at local level, in all too many cases, the reaction was rather more: ‘Er, that’s not how we do things round here’.
Attitudes may be changing over time, but the US experience highlights one of the issues that such mandates run into, specifically that it’s one thing issuing directives to central government departments, quite another when it comes to getting a positive response at regional level.
For that reason, many other Cloud First countries have taken a more hands-off approach, including the likes of India and Australia. In the UK, for example, the official Public Cloud First policy applies to central government departments only, not to local government.
This week, an interesting row has sprung up in the UK in the wake of the results of a Freedom of Information (FoI) request issued by supplier Bull Information Systems.
This reinforced the suspicion that to all intents and purposes local government in the UK has resisted the allure of commodity cloud service procurement via the G-Cloud framework, despite it offering cost and time benefits of up to 50%.
Now, local government’s lack of take-up so far isn’t exactly a secret.
G-Cloud has the potential to reach an estimated 30,000 buyers across the public sector. Yet research carried out by the 6 Degree Group suggests that nearly 90% of local authorities have not heard of G-Cloud.
But Bull’s request for information quantified the problem in cash terms by revealing that of the £440 million ($740 million) spend on IT during 2012-13, less than one percent – £385,000 ($647,262) – was spent via the G-Cloud framework.
FoI responses from 26 country councils found that just eight bought services through the G-Cloud’s Cloudstore last year, with the 12 purchases ranging from £1,511 to £154,911. Only 31 councils in the UK (out of 486) have spent more than £50,000 in total via G-Cloud.)
This, at a time when all local authorities are complaining of the need to make cutbacks and slash budgets. So why are they turning their backs on a more cost-effective way of delivering IT-enabled citizen services?
The two examples picked out in Bull’s findings were the two largest local authority IT spenders during the period in question, Kent County Council and Hampshire County Council. They spent £38.5 million and £38 million on IT respectively.
Of that, Kent spent a mere £94,750 via the G-Cloud.
Mind you, that’s £94,750 more than Hampshire which spent absolutely nothing!
Now there are extenuating circumstances of course to be applied. For example, Hampshire County Council’s CIO Jos Creese told Government Computing that his authority had made one of the highest local government G-Cloud spends the previous year and expected to:
see its use increase as it offers new and more flexible ways of contracting for IT goods and services.
What’s more alarming is the comments issued by Kent County Council that G-Cloud is:
not currently able to offer the time savings, quality assurances and consistency necessary to make it effective. We have raised these issues with the Cabinet Office and believe that until these are addressed, the G-Cloud does not offer the best route for sourcing software for local authorities.
Now there are of course exceptions to prove the rule, such as Hounslow Council, which has put £2.4 million of spend through the G-Cloud framework to date, or Bristol City Council, which has spent £1.5 million through the CloudStore online shopfront.
Nonetheless, the overall picture is downbeat, or as Andrew Carr, CEO at Bull, says:
hugely disappointing and quite shocking. By sharing infrastructure costs and moving to the cloud, county councils could take 20% to 25% out of total IT costs – they’re wasting millions not doing this.
That’s a comment that local government has taken umbrage over with local authority IT association Socitm issuing a rebuttal of that conclusion on the grounds that what is referred to as ‘the G-Cloud’ is a procurement framework only, not a cloud offering in its own right.
Socitm argues that as a result, it is not fair to suggest that low use of G-Cloud correlates with low use of cloud services. In other words, just because local government bodies aren’t using the G-Cloud framework, doesn’t mean they’re not interested in using cloud services.
It goes on to argue that another issue is that there aren’t enough local government line of business offerings available in the Cloudstore and that G-Cloud’s origins stem from central government needs.
Martin Ferguson, Head of Policy at Socitm, states:
G-Cloud is already a useful procurement framework. However it is still in its relative infancy, as is cost effective public cloud provision for use in councils more generally.
What will make the G-Cloud increasingly attractive will be the flexibility to use it in ways which deliver best value and sustainable IT architectures fit for the future, especially where these impact on councils’ increasing need to join-up and deliver services with partners in Health, Police, Voluntary and other sectors.
It is also the case that the biggest beneficiaries of cloud computing, and G-Cloud as a procurement vehicle, are likely to be the smaller public service organisations which were not covered in the FoI research carried out by Bull.
The resistance at local level to a central government cloud mandate is scarcely new. We’ve seen it in the US; we’re seeing it in the UK; we’ll see it in other countries as well.
Frankly there’s a large pinch of an older problem here though. Public sector IT procurement at local government level has always been troubled by a ‘not invented here’ attitude, with too many bodies deciding that their needs as an authority are different to others. See shared services, for example.
I’m not suggesting that there’s an easy answer here. The prospect of a Local Cloud First policy isn’t realistic or practical. So there needs to be a lot of carrot and not the central government stick in play.
The point is well made by Socitm that not using the G-Cloud framework doesn’t mean a rejection of the basic idea of cloud computing by local government. But it’s equally clear that the merits of the G-Cloud framework need to be sold and resold to a skeptical audience.
That’s going to mean more evangelism from the centre in the form of the Government Digital Service team, as well as strong leadership and guidance from Socitm members who’ve acted as pathfinders to the cloud.
State-level resistance to the federal Cloud First mandate in the US is eroding bit by bit, but it’s taking time. The same may well be true in the UK and elsewhere.
Final point: let’s try to avoid exchanges such as this one between a supplier and a leading local authority CIO:
As Ciuffini has been one of the more enthusiastic advocates of cloud computing in local government in recent times, I’d say we can file that particular sales pitch under ‘unhelpful’.