This week I was pointed to a couple of articles on PublicTechnology.net, which they interview with Crown Commercial Service (CCS) technology director, Niall Quinn, about the future of tech procurement in Whitehall. The piece makes a fascinating read and has caused a bit of a stir (to put it kindly) amongst SME suppliers and digital government observers - with many left asking the question...WTF?
Now, before I proceed, it’s important to say that I haven’t interviewed Quinn myself (although I have put in a formal request), and so the following dissection of his comments are taken as face value off the back of PublicTechnology.net’s published pieces. That being said, I’ve spoken to contacts that have been dealing with CCS in recent months and a lot of the frustrations listed below are feedback from people that have worked in government or supplied to government over a number of years.
Simply put, there’s a great deal of frustration over Quinn’s comments on the future direction of the Digital Marketplace (which incorporates the highly successful G-Cloud and Digital Outcomes and Specialists frameworks) - namely that his plans won’t make things easier for SMEs wanting to supply to government and that they won’t force good behaviours from buyers in Whitehall and the broader public sector.
I’m assuming that most people reading this will have some background on tech procurement in Whitehall, so I won’t go into detail about the impact of the Digital Marketplace (you can read more here), but it’s worth reminding ourselves of two points. Firstly, that since 2012 the G-Cloud framework alone has seen £4.79 billion of spend go through it, £2.15 billion of which it is claimed has gone to SMEs. Something is working. Secondly, CCS recently took over responsibility for the Digital Marketplace, taking it away from its previous owners, the Government Digital Service.
Okay, so here we go, let’s take a look at exactly what Quinn said.
Where’s the focus on outcomes?
Quinn explains to PublicTechnology.net that the future of the Digital Marketplace - Digital Marketplace 2 - will be something new entirely. He compares the current version to Amazon, but claims the future should be more GoCompare. The key quote states:
If our buyers want to buy cloud services, we need a platform that can ask them: what’s important to you – price? Length of term? Flexibility? – and then take those responses and give you 10 suppliers in that space… then you need a bit of problem definition, so you can type your needs in and that goes out to those 10 suppliers and they get back to you. It is more of a comparison website as a front-end, rather than it being a point-and-click Amazon-type scenario.
Firstly, what should be most important to a government buyer should not necessarily be price, length of term, flexibility, etc. Where is the outcome? A platform can’t solve for complex user needs and then serve up a list of 10 suppliers to solve that problem. Secondly, focusing on these questions first and foremost will inevitably favour the large vendors that benefit from scale.
Secondly, I have *huge* concerns about a platform deciding which 10 vendors would be best for a government buyer’s needs. Who is controlling that? Will we have transparency into that algorithm? What mechanisms will be introduced to remove gaming/bias of the system? A comparison website for government suppliers, who are looking to solve complex outcomes, just seems too far fetched to me.
It may just about work for hosting. But, to be honest, the market is moving towards multi-cloud anyway and systems are being developed that solve this problem. It doesn’t really sound like Quinn is talking about a dynamic purchasing cloud for IaaS here though.
Quinn goes on to add:
[The platform] will know: you’re in a school in south-west London, you are an overworked admin assistant who wants to buy a new photocopier – here is the one everybody buys in your location
I understand the logic here. Common user needs serving up helpful information for buyers that may be confused. Sure. But again, a system serving up a limited amount of vendors for buyers that are all using the same thing? Sounds like the 90s to me.
My gut also tells me that these recommendations will more than likely favour the large and well known over an innovative SME.
Quinn also told PublicTechnology.net that in the future, G-Cloud - or more specifically, “a framework dedicated purely to cloud hosting and related services” - will be split into three sections covering hyperscale hosting (Microsoft, Google, AWS, etc), smaller hosting providers and “related services”. Not only this, it will likely feature 1,000 suppliers at the most, compared to the 4,200 suppliers on G-Cloud 11.
Let’s start with these points. Firstly, why are we splitting out hyperscale from other hosting providers? Some smaller UK hosting providers have been able to build a successful business because they’ve finally been listed alongside the likes of AWS and Microsoft. For a buyer, what is the need to split them out? Equally, how will this benefit SMEs?
Secondly, limiting the number of suppliers on any future cloud framework will make life incredibly difficult for SMEs. Yes, G-Cloud needs improvements around search, discovery and the ability to find what you need. But as noted above, the reason G-Cloud has been so successful is that it has been open to all and has presented SMEs alongside large suppliers, transparently. This goes against the founding principles of G-Cloud.
Quinn also told PublicTechnology.net that areas such as machine learning, artificial intelligence, analytics and RPA will also be broken out into smaller dedicated frameworks outside G-Cloud.
My question - why is more frameworks a good thing? I’d argue that more often than not, buyers don’t really buy this way. Or shouldn’t. Again, the lack of focus on outcomes is really worrying. Someone shouldn’t be sitting in a Whitehall department saying - ‘do you know what we need, AI!’. Instead, an outcome focused on user needs should be supported by a variety of technologies and suppliers available through a single framework. Doesn’t that just make more sense?
The end of two year contracts?
Finally, and what will likely be the nail in the coffin for long-time supporters of G-Cloud, Quinn said that he is looking at increasing the current two year contract term to five years. He said:
We may also bring in longer terms. So, rather than two years, plus one, plus one – if you’re putting records into the cloud, you don’t want to have to repeat that [procurement] all the time – it might be more appropriate to have a five-year term.
Five years in the new age of cloud and digital is bizarre. In fact, it’s damaging. Why would anyone want to tie themselves into a contract for five years when the pace of change is so intense? Or when contract prices are actually reducing every quarter? And for those who don’t already understand this, the two year contract terms are there to force suppliers to please their buyers and to force buyers to think about their needs. It doesn’t mean you can’t renew with your current provider, it just means you have to prove that the current arrangement is delivering value.
I thought this message had been delivered, but clearly not.
I hope I get to interview Quinn soon because I would love to understand his motivations. Some have suggested to me that CCS makes its money by taking a cut of the revenues going through the Digital Marketplace, so more frameworks = more ££. Longer contract terms could also be quite useful on that front, I suppose…
That being said, more than anything, this is all so worrying because it makes the changes that have been enforced over the past few years much harder. Buyers will slip into old buying habits. SMEs will find it much harder to supply to government.
Yes, the current frameworks and Marketplace aren’t perfect. Suppliers already find it difficult to bid for work and buyers sometimes are overwhelmed by the scale of the Digital Marketplace. But I’m not convinced Quinn’s proposed solutions are the answer. Disaggregation sounds like complexity to me. Longer contract terms and limiting suppliers sounds like a future of lock in to me. And little mention of user need or a focus on outcomes doesn’t exactly fill me with inspiration.
Let’s hope I’m wrong.