On public sector cloud adoption

The Government Digital Service (GDS) has been the subject of numerous Computer Weekly stories of late, as Whitehall sources claim some senior civil servants want to break it up and return IT to its previous departmental model.

For GDS, power struggles with other parts of Whitehall are a feature of normal operation. Given its mission to wage war on the old Whitehall CIOs and their big suppliers, it’s hardly surprising its demise has been predicted every year since its foundation.

But government’s take on digital has reached a turning point, and it looks likely the balance is tilting back to Whitehall departments.

I’d like to argue the shift we’re witnessing isn’t about personalities but technology, and a maturing market where government departments must take the lead in developing new skills and capabilities.

Government cloud: The first generation

Cloud has been one of the bright spots in an otherwise static public sector IT market in recent times. Kable’s data indicates public sector spend on cloud services (Iaas, PaaS and SaaS) has been growing at 45% per year.

GDS deserves credit. It didn’t invent cloud, but it put off-premise IT services on the public sector IT road map, thanks in part to the introduction of the UK government’s ‘Cloud First’ policy in 2013.

Cloud First was greeted with enthusiasm by new digital teams who were committed to bringing agile working practices to Whitehall for the first time, and found commodity cloud services quicker to procure and easier than their own IT departments to work with. And the first generation of digital services were well suited: greenfield, handling little sensitive data, and with light back-end integration requirements.

GDS nurtured and grew G-Cloud, which has been an excellent vehicle for early cloud projects: not just commodity IaaS, but also PaaS, and SaaS tools supporting new collaborative ways of working.

It offers buyers a simple and quick route to access a wide range of suppliers, especially SMEs who were often shut out of the public sector market. On average, the public sector is now spending over £50m per month via G-Cloud (although some 80% of that has been invested in professional services).

Reaching a plateau

This all begs the question: if public sector cloud use is growing, and digital teams are getting what they need,  why is GDS coming under fire?

A Kable analysis of G-Cloud sales data, supported by anecdotal evidence from suppliers, suggests sales growth through the procurement framework has reached a plateau, as many of the quick-win cloud projects that are typically funnelled through G-Cloud have been done.

The next wave of cloud activity – call it public sector cloud 2.0 – is not about G-Cloud or even Cloud First. It’s being driven by the digital transformation of existing services, rather than just new digital projects.

Opportunities will be open up as legacy IT outsourcing contracts come to an end, and organisations look to shift workloads to the cloud.

Make no mistake, this is where the big wins are. Cloud progress from here on in will dwarf anything achieved in the last few years. But this is complex work.

Next-generation public sector cloud

As my colleague Gary Barnett has argued, people sometimes talk about cloud as if it has magical, unicorn-like properties. The nitty gritty reality of cloud migration is ignored, and buyer enthusiasm is not always backed up by expertise.

We see evidence of this in digital services not being built for cloud, crashing under predictable demand spikes: the digital car tax and voter registration systems are good examples of this.

Another factor of note, interesting in the wake of the Brexit vote, is that the UK government’s stated preference for public cloud is very much out of step with its European counterparts, who are largely investing in secure, government-only private clouds.

Another instance of plucky British exceptionalism, or a sign that Cloud First is fundamentally misguided? I’d like to suggest, diplomatically, that hybrid cloud is the destination to which most departments are headed.

Making a success of cloud 2.0 is not just about swapping out expensive hardware for cheap public cloud infrastructure, and it’s not about G-Cloud either. It’s not even about trucking old servers down the M4 to the Crown Hosting datacentre.

It’s about enterprise architecture; data quality projects; new development and deployment processes; and governance models, security policies and service management. In short, it’s about hybrid cloud and orchestration.

GDS: Where does the buck stop?

It is not, however, in GDS’ gift to deliver this kind of complex departmental transformation.

Perversely, too much GDS involvement in departmental digital projects can actually reduce accountability when something goes wrong.

For instance, GDS played an important part in the fateful technology choices made for Rural Payments Agency’s (RPA) digital service, resulting in senior RPA and GDS staff blaming each other for a major project failure that cost taxpayers dearly and caused misery for farmers.

Responsibility – and accountability – should be assumed by the party with the most skin in the game. This is not to say that GDS has no role to play, simply that it would benefit from handing the reins back to departments in some areas, and allowing them more say over shared infrastructure that affects their client groups, such as Government-as-a-Platform.

Much has been said about GDS stopping departments making bad decisions. But in the more complex cloud 2.0 world, there’s an increasing chance GDS will inadvertently stop them making good decisions that are appropriate to their unique circumstances.

Where next for GDS in a cloud 2.0 world?

Arguably, the need for high-profile interventions and vetoes is decreasing anyway, partly thanks to GDS’ own efforts to increase departmental capability and to establish common standards.

To my mind, GDS deserves huge credit for the creation of a new digital movement – culture, values and ways of working – which has influenced practice beyond Whitehall and even beyond the UK.

The Digital-by-default Service Standard; the Service Design Manual; the Digital, Data and Technology Profession which has replaced the moribund Civil Service IT profession.

The leaders networks, the blogs, conferences and culture of openness. Hiring new senior people and digital teams into departments. These are GDS’ real levers of power.

GDS’ proudest achievement is not the vanquishing of its political enemies, but its winning of hearts and minds. It’s by the continuing projection of that soft power that it can best support the whole of government to move forward with cloud.

(Originally published on Computer Weekly)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s