Creating G-Cloud

I created G-Cloud. With 4,200 suppliers, 30,000 services, and over £5bn, it’s the largest and most successful IT procurement framework in UK government. G-Cloud has opened up government to SMEs, delivered tens of millions of pounds in savings, revolutionised the way UK government buys services, and has been copied by governments around the world.

If you are a public sector body or a supplier and you want to know how to use G-Cloud effectively, contact me today.

The Problem

When Francis Maude (Baron Maude of Horsham) was Minister for the Cabinet Office, he said that he wanted the government to use cloud services, and he wanted the process to be as quick and cheap as an app store.

The Approach

When the idea of a government cloud was first mooted, I was the commercial lead (later director) of the government’s largest internal hosting service, The Club. The Club was reaching the end of its strategic supply contracts so was looking to not only renew but fundamentally change the way it provided hosting to the main central government departments.

The Club’s strategy and the government cloud strategy as it was then termed merged and I joined, then led, the cross-government commercial strategy team that was charged with thinking the unthinkable and deliver a cloud app store for the government that was fully compatible with procurement law.

The Club’s strategy and the government cloud strategy as it was then termed merged, and I joined, then led, the cross-government commercial strategy team that was charged with thinking the unthinkable and delivering a cloud app store for the government that was fully compatible with procurement law.

This was of course considered impossible. Government procurement was too broken and the law was seen as a straitjacket to innovation. Other issues included:

  • The way government buys most services was, and largely still is, slow and expensive
  • The procurement process itself creates structural inequalities between large organisations who have bid teams and SMEs who don’t
  • The process excludes start-ups as they fail the usual risk and financial tests that are applied during the buying process
  • Typically, one can only buy services you know about, we wanted to set up a market of things we’d never heard of
  • If one buys a ‘commodity’ then it can’t change – but cloud services do change

But we were told to think the unthinkable. So we did.

The first breakthrough was to realise that as services changed rapidly all we had to do was keep buying them as they changed, so we proposed that we procure every 6 months. This was seen by some as madness, because procurements typically take 9 months so one would be re-buying services that that one has not yet bought.

For us not to spiral into insanity we had to re-run the procurement process every three months. This gave rise to the second breakthrough: while elements of a cloud service may change, 90% of each procurement is exactly the same, so we could almost automate the process through the use of online forms.

The next issue was ensuring that we guaranteed that buyers got quality services from suitable and robust companies. We looked at various ways of assessing quality and risk, and learned that every option added time and cost to the process. Then I realised that we’d missed the point, which was that we were establishing a marketplace for a wholly new, agile way of working. While some parts of the government needed the cloud for huge strategic services, others just wanted new tools to experiment with.

I therefore proposed that we apply effectively no quality or financial tests to suppliers. We simply required them to be existing cloud services that fully met the US governments’ definition of cloud – again, why invent a definition when one already existed. It was a tough sell, but when people realised they could just go onto the internet and see all the services in action and even test the lives services for little or no cost, the argument for setting up a quality industry fell away.

This brought us to the elephants in the room: government is special. It has unique needs, it makes suppliers sign up to contractual terms it dictates, and creates bespoke services that only government can and will use.

Fortunately, this is almost entirely untrue:

What’s unique about government is not the services it uses, but the way it uses those services.

This realisation opened up a whole set of possibilities and implications for the way that services are bought and sold. I started with the assumption that everything from the technical service itself to the supplier’s contract was a commodity that the government would not change. This allowed us to create a marketplace where fixed commodity products could be seen by everyone: buyers, the public, and other suppliers.

We termed it radical transparency and many suppliers did not like it. We published all supplier prices and told suppliers that they could change them at any time just so long as:
(1) they only lowered prices
(2) the prices were lowered for everyone

This meant that suppliers could no longer divide and conquer government departments – essentially before G-Cloud, what the taxpayer ended up paying was based on the buying and negotiating skill or sales skill of an individual department and supplier.

To make this radical transparency work I created a wrap-around contract that sat on top of the supplier’s contract and trumped it, but crucially only in very special key legal areas. The government wrap-around cloud store contact contained no SLAs, no IP clauses, and nothing that changed the commodity services provided by the supplier.

Bringing all these factors together we created the commercial structure of G-Cloud, and the team developed the original digital marketplace as an online store where people could buy services based on their service definitions and terms.

The last problem was people. We’d worked out a way of buying technology but if the government had no one with the skills to manage it and if the process of obtaining those skills was slow and cumbersome, what was the point of it all?

So I invented what I called ‘people as a service’ (now specialist cloud services). Here I applied exactly the same logic to professional services as we had for technical services. The supplier defined their service up front and people can buy it at a fixed price. This has risk, as there’s no custom contract which buyers are accustomed to seeing, but professional services and outputs can be defined just as well as technical ones if one thinks of them as commodities.

The Result

G-Cloud was born!

Not only did we run all these ideas past senior government lawyers but we talked to the bill team who wrote the procurement law – it all checked out. To get suppliers on board, we tweeted, we live-streamed, we put up videos on VIMEO, I wore skinny jeans and talked at start-up events in Shoreditch.

It worked. So far there’s been over £5bn in sales through G-Cloud. SMEs who find out my role in G-Cloud still thank me and tell me it’s opened up government to them and has created a revolution that hasn’t stopped.

In fact, now I’m no longer involved in G-Cloud you can buy my companies services from G-Cloud it right here.