It’s great that there’s a lively discussion at the moment about the technology approaches local government might consider. It’s important given how fundamental it is for driving change.
As a sector, we need to liberate ourselves, as quickly as we can, from rigid and costly IT systems that tie us to 20th century business models and keep us in institutionally defined silos. Councils simply have to find ways to create new digital services for the internet era, services that are less bureaucratic, far more human and above all local.
API strategies (making systems more easily interoperable) are vital in helping organisations cope better with very persistent legacy environments; but because there is so much inefficiency and machine-like process baked inside these ageing and pre-internet systems, we have to get into them, free people up, by designing and developing anew. A prison break if you will.
SME innovation is a critical ingredient, because we need new minds and perspectives on our challenges, and the GovTech Catalyst is an exciting development here, supporting innovation by a new breed of companies in the public sector space. This kind of approach needs accelerating given the incredible pace of technology developments, especially in new areas where there is no capacity at all within councils (AI, machine learning etc).
But .. having said this .. there is a huge space and need for digitally enabled local change that these approaches won’t help us tackle.
Building in open source, the ideal set out by many, looks to me like a very challenging strategic choice for councils trying to solve myriad issues with very limited budgets. I’m really not sure about pace, programme capacity or cost, regardless of whether councils go for in-house capacity or commission work from digital agencies or a blend. It’s a time x day rate calculation.
This is a marathon not a sprint for councils, and it’s really important — existentially important — to look at how internal IT teams can change themselves to best effect .. to become better able to meet the demands of a wide variety of stakeholders needing change right now, in an affordable and sustainable way. Commissioning work out to great agencies is fantastic (it really is), but it’s really important to build internal capacity too, and I doubt most councils could afford a big enough, skilled enough team to build in open source. It would be great to hear from those making progress in this direction ..
Certainly there is no way my councils could have either built an open source dev team, or contracted everything out. We reviewed the open source approach back in 2015 (my philosophical preference) and concluded it was simply not feasible for us with the resources and money we had. We found a council that had tried it and it had failed quite badly.
After careful consideration, we went for what I think was a good, pragmatic compromise. Our chosen open standards platform (this is a must), providing a “low code” development environment, has a fixed enterprise licence fee that means we can not only build unlimited apps for ourselves, but can build apps for any public sector body operating in our geography at no additional cost. Development time is much faster than it would otherwise be, and the skills required are significant, but lower than other development environments.
On the cost side, we have recently avoided buying a housing repairs system by designing and building on our platform. That one system would have cost us more than our entire enterprise licence, so the financials are utterly compelling given we also have several other major apps running already.
Tip: negotiating a fixed enterprise licence is critical to being able to drive value, rather than build up cost.
I’ve noticed in the low code debate that there’s a tendency to attach bad practices to the approach which aren’t actually integral to it, so let me try and summarise again the approach to low code at Adur & Worthing:
- We use dedicated in-house (and sometimes external) developers to build in low code, with strong skills and right mindset. We don’t think anyone can do it, but we do think there is more developer potential out there in services to be found over time. This is a really good thing in time if managed well.
- It might take minutes to build a simple app, but enterprise grade end-to-end apps take 4–6 months. Good, holistic service design is a universal requirement, and of course we must always build what’s actually needed. I do get that there could be a temptation to build, build, build, but that’s about having a good method and control. There’s nothing inherent in the tool that forces you to build for the sake of it!
- You do need to hire interaction designers (we have a list of great freelancers and small agencies we use), because it’s not magic, and good UX design is as important in low code as anywhere else. There are sometimes functionality compromises in low code — there has to be a down side- but we have been ok, and these platforms are evolving quickly and there is only one way this will go. Our platform has come on leaps and bounds, and we’re excited about getting closer and closer to high quality products.
- It’s true that vendor lock-in is a question, and there is an “all eggs in one basket” risk as we work fast and keep costs down. However, the mitigations are there, and we will soon have auto-generated data architecture diagrams and workflows which is good enough insurance, along with ESCROW. Migration pathway challenges are ubiquitous, and we keep a close eye on the fortunes of our supplier. We’re proud it’s a UK SME who we can talk to and work with. The other main low code vendors are off-shore.
- We are also keeping our eyes open for other products, such as LiveCode, and are keen to see open source low code platforms emerge in a sufficiently mature state. That would make everyone happy I guess! We will diversify our technologies over time, but for now we’re busy making progress.
We think it’s obvious that more and more code will go ‘below the line’ over time. It’s a tech evolution which is giving us the capacity for true service design that we so desperately need. That is also something that a council should have ambitions to creating for itself.
As councils truly become service designers and digital developers, they will become smart commissioners of those agencies who do such fantastic work in those spaces. And the ecosystem will thrive.