If the government ditches its cloud first policy, it’s making a huge mistake

Profile picture for user ddpreez By Derek du Preez May 14, 2019
It has been revealed that the Crown Commercial Service and the Government Digital Service are planning to shift the cloud first policy to one that better reflects hybrid cloud environments.


If the Government Digital Service (GDS) and the Crown Commercial Service (CCS) follow through on plans to repeal the government’s forward-thinking Cloud First first policy, which has been in place since 2013, then we can expect digital change across Whitehall to happen at an even slower pace.

Why? Because culture change is hard and ways of operating are deeply embedded across the system, meaning that strict controls need to be in place at the centre to force people to think differently. Soften those controls and human behaviour will naturally gravitate towards what is easiest and what is familiar (not necessarily with any ill intent).

For those unaware, last week it was revealed that CCS and GDS are reviewing the long-standing Cloud First policy, indicating that they perhaps favour an approach that accommodates a hybrid cloud model.

In a statement to Computer Weekly, which broke the story, Neil Quinn, director of the technology pillar at CCS, said:

As we have worked though the digital transformation journey with many central government departments and wider public sector organisations, it has become apparent that one size does not fit all, and organisations should make sure they understand what the journey to ‘cloud’ is and means for them in terms of costs, risks, skills and timescales,” said Niall Quinn, director of the technology pillar at CCS, in a statement.

We are seeing more and more customers land on a hybrid solution and therefore ‘cloud first’ may not be right for everyone.

This follows some guidance that was recently issued by the National Audit Office, which states that “cloud first may not be right for everyone”. Worryingly it adds that “cloud solutions may not always save money” (who said they did?!).

Why this is misguided

Is it surprising that departments are landing on hybrid cloud as an approach? No. This broadly reflects what’s happening in the ‘real world’, where organisations are transitioning towards the public cloud, but it’s either not cost effective or operationally possible to migrate everything.

The question CCS and GDS should be asking themselves is, would government departments have progressed as far as they have without the cloud first policy in place? I highly doubt it.

Central controls have been absolutely critical in driving change across Whitehall. We have seen this diluted in recent months, particularly with central functions being moved out to less powerful departments - we have warned about the impact of this previously.

As noted above, change is really, really hard. To make sure it happens, there needs to be a delicate balance of stick and carrot. The ‘stick’ here being spend controls and a policy that dictates that cloud is assessed first and foremost.

The ‘carrot’ being improved procurement mechanisms that make purchasing cloud solutions easier than the alternative (thank you, G-Cloud).

I’d be less concerned about the policy being removed - or softened to reflect reality (as CCS and GDS will likely argue) - if I thought that the modern ways of operating had been embedded across government. But they haven’t. Yes, things have changed, but there is still a lot of change that needs to happen and the Oligopoly is lurking in the wings, ready to offer their version of the ‘carrot’.


Also, let’s be clear, the cloud first policy does not dictate that cloud has to be procured. It just has to be considered first and foremost. In fact, it states:

When procuring new or existing services, public sector organisations should consider and fully evaluate potential cloud solutions first before considering any other option.

It then adds:

Departments remain free to choose an alternative to the cloud but will need to demonstrate that it offers better value for money. HM Treasury’s ‘Managing Public Money’ (annex 4.6) defines ‘value for money’ as ‘securing the best mix of quality and effectiveness for the least outlay over the period of the use of the goods or services bought’.

In other words, if you want to remain on-premise for a service - go for it! Just prove it is more effective and provides better value for money. I’d argue that most of the time this wouldn’t be possible, but it’s still an option.

The same confusion lies in the government’s two year contract terms for cloud, which again, are reportedly going to be softened by CCS and GDS. Departments have bemoaned the two year terms since they were introduced, as they claim that it’s not long enough to deliver the value needed from procuring a new service.

Again, this is confused. Firstly, in the ‘real world’ pricing of cloud contracts falls year on year thanks to economies of scale - tying yourself into a 5 year contract doesn’t make any sense. Secondly, departments are wrong in thinking that the two year contract term means that they have to change provider every two years - they just have to prove that that provider is still the most effective. There’s nothing to stop the department sticking with that cloud provider for decades, as long as the cloud provider keeps up its end of the bargain and remains competitive.

It’s another central control that has been put in place to ensure behaviour changes are being forced and buyers are thinking about how they operate, without becoming complacent and palming off responsibility to a third party outsourcer for decades on end.

My take

If CCS and GDS follow through on this, it’s a huge mistake. Old behaviours will creep in and departments will be able to point to the softened policy and say - “Look, we are doing the right thing!”.

The reason this is likely being softened is because it’s hard and requires a lot of work from all those involved. It requires a lot of work on the part of CCS and GDS, and it requires a lot of work by departments. Will it be easier to shift the policy to a softer approach? Sure.

Butwhen did anything good come from taking the easy route?