For several years, Key Group ran what I would describe as a pick-and-mix technology strategy. A large number of suppliers, a wide spread of platforms and technologies, and the significant overhead that comes with managing all of it across a fast-growing financial services business. It worked, but by the time we decided to change it, two problems had become impossible to ignore.
The first was cost. Our technology bills were genuinely difficult to model. Pricing a new service or product proposition was slow and approximate, and getting it wrong had real consequences for a VC-backed business that needs to make investment decisions at pace. The second problem was speed. Provisioning a new environment was taking weeks, sometimes longer, at a point when our product teams needed to move far faster than that.
Underpinning both was a strategic decision we had already made as a business: to consolidate fully onto the Microsoft ecosystem and build everything we wanted to do next, including an increasingly important AI programme, on top of that platform. What we needed was a partner who had been there before us, not one who would be figuring it out alongside us.
The evaluation
We ran a structured RFI process and looked at the market properly. A few things separated Version 1 from the rest fairly quickly.
The ASPIRE Managed Services Framework stood out immediately. What surprised me was how many providers we spoke to that did not actually have a documented delivery methodology. They had plenty to say about how they work, but when it came to showing us a run book or an established operating model, most fell short. Version 1 had it baked in. That gave me confidence before we had even got into the commercial conversation.
On the Microsoft side, we asked specifically about the number of certified individuals across the estate. Version 1 had the highest of any provider we evaluated, and they were across the newer accreditations too, the ones that matter if AI is genuinely part of your roadmap rather than just a slide in a deck. They also identified licensing and pricing optimisation opportunities during the RFI itself, before any contract existed. That kind of good faith matters when you are trying to work out who you actually want to spend five years with.
The transition
My one goal for the transition was simple: that nobody inside Key Group would know it had happened. No downtime, no disruption, no incidents that would undermine confidence in a new partner before the relationship had a chance to prove itself.
We achieved it. The transition was seamless. And within the same 12-month period, we completed a second transition, bringing our Azure Virtual Desktop managed service across to Version 1 with the same outcome. Two transitions, both invisible to the wider business. I do not want to take that for granted because it is genuinely hard to do, and getting it wrong has a long tail.
What I did not fully appreciate before we started was how much useful information comes out of the transition process itself. Version 1 did a well-architected review of our entire estate as part of the handover, and the Continual Service Improvement log they generated from that became the engine for everything that happened next. The incident reduction we experienced in the first six months was not coincidence. It was the direct result of a systematic process of identifying and removing root causes, and it started at transition.
What actually changed
Before Version 1, Key Group had spent the better part of three years running an IT-change-focused programme: technical debt, data centre migration, infrastructure upgrades. It was necessary work, but it was consuming the team. After the transition, the infrastructure stopped being a problem that needed managing. Incidents that were running in double digits per month dropped to low single digits within six months, and in Q1 of this year our total incident count for the entire quarter was under ten.
That shift in operating tempo, from firefighting to building, is the outcome that everything else depends on. My teams now focus on building experiences and customer outcomes. That is where they should be spending their time, and it is where the real competitive advantage in financial services gets built.
The commercial change has been just as significant. Every bill since go-live has been exactly what was quoted. For someone who spent years dealing with technology costs that were difficult to model and harder to predict, that level of transparency has fundamentally changed how quickly we can make investment decisions. It sounds like a small thing. It is not. The reduction in supplier complexity has been equally tangible. Managing multiple vendors means managing multiple accountability gaps, multiple commercial relationships, and an innovation agenda that nobody owns end to end. Moving to a single strategic partner removed all of that.
Time to market has improved substantially too. We provisioned a complex runtime environment of around 40 virtual machines in approximately four weeks earlier this year. Previously that would have taken three or four months. The difference is not just operational; it changes what is commercially possible.
Beyond managed services
What I want to be clear about is that the managed service is the foundation, not the full story. Over the past year, Version 1 has supported us with Copilot enablement workshops for our end users, a significant mortgage origination platform upgrade, and a structured AI opportunity assessment run through Version 1’s AI Co-Creation approach that has already identified two near-term initiatives we are moving forward with. What I previously thought would take months of scoping and significant budget can now be stress-tested and validated in a single session. They have also provided technology market analysis that has helped us pressure-test our roadmap at leadership level and demonstrate to the executive team what is actually achievable on AI, and at what speed and cost.
One of the challenges in financial services right now is that AI ambition often runs ahead of what people believe is commercially viable. Version 1 helped us show that some things we thought would take years and significant investment can be delivered in a four to six week timeframe. That changes the internal conversation completely.
What comes next
Key Group’s direction is to consolidate fully onto the Microsoft ecosystem, get all of our data into a single governed place that is accessible to the Microsoft AI suite, and build customer and broker experiences on top of that platform. Version 1 is central to all of it, across infrastructure, data, cloud and AI. The managed service gives us the stable, optimised platform that everything else depends on.
If I were to distil it to one thing for other IT leaders in financial services who are considering a similar move: the question is not just whether a partner can manage your infrastructure. The question is whether they can support the business you are trying to build on top of it. Those are different criteria, and the provider who is right for the first is not always right for the second.
For Key Group, Version 1 has been both. Twelve months in, the infrastructure just works. And that has freed us to focus entirely on building the experiences and AI capabilities that will define how Key Group competes over the next five years.
Jack Neary is Group Technology Director at Key Group. Version 1 and Key Group have been in partnership since 2024.
