This is the year most of our portfolio stops being a promise and starts being a product on shelves, in apps, and in the hands of customers we have spent two or three years quietly building for.
Dear readers of divergent, the last two or three years of this business have been spent building the expensive, quiet parts of our companies. Core platforms. OCPP integrations. Mobile network agreements. Regulatory approvals that took eighteen months and reached a total audience of six people. The work that doesn't show up in a blog post, doesn't photograph well, and is invisible to anyone who doesn't sign the timesheets.
2026 is the year that work shows up in public. Every company in the group has a commercial milestone this year, and several of them come to market for the first time after years of being asked, politely, when something is actually going to happen. I want to tell you what we are watching for , and what I hope you will watch for with us.
What I can tell you now is that the run-up has involved a lot more listening than building. The partners and customers we worked with through 2024 and 2025, the installers, the charities, the single-site operators, the independent groomers, the IT managers running phones for a hundred staff , are the reason the launches this year look the way they do. Almost none of the product we are putting on shelves was our first idea about what to build. It is our fifth or sixth, and it is better for it.
The bottleneck of 2026 is not technology, and it is not capital. It is trust. The companies ready this year are the ones that spent the last three earning it from the customers nobody else was listening to.
Not a roadmap. A short account, one paragraph each, of what you will see from our portfolio over the course of 2026, and what it took to get there.
After two years of quiet build alongside installers, charge-point operators and a very patient set of OCPP partners, Inflo opens its platform to the wider EU market this year. The public release closes the commercial gap in EV charging software: enterprise-grade back office, no subscription tax, no per-charger monthly fee. 2026 is the year Inflo stops being a thesis and starts being the default answer.
Neo spent 2023 and 2024 getting three separate products ready: Wander (eSIM for Travellers), Mobile (Business Mobile without the contract rises) and Voice (VOIP for businesses of all sizes). This year, they arrive in market as one platform. Expect 2026 to be the year a UK business can credibly retire the three-vendor setup that has defined small-business telecoms for two decades.
Reservate has been live since 2023 as Reserve, but 2026 is the year its pay-per-use pet care model stops being the curious experiment and starts being the norm. With 99.4% customer satisfaction across 16,000 end users, and a roadmap that adds two more pet care verticals before the end of the year.
Connect's Wholesale+ rate card has been quietly proving itself for over a year, most visibly with 20+ charities who use us as their only messaging provider. In 2026 we take that rate card wider: more direct international SMS routes, a robust email deliverability engine, and the same refusal to bend pricing to customer size.
Three ventures still in the build phase will begin to matter this year. I am not ready to name them yet, but the pattern is the same as everything else we have shipped: small customers first, honest rate cards, and technology that quietly becomes useful before anyone calls it interesting.
The second thing worth mentioning is AI, because pretending 2026 did not already change because of it would be silly. The tools available to anyone building software now are genuinely different from the ones available two years ago. The pace a well-scoped team can work at has quietly compounded, and we feel it every week. What it has not done, for anyone I trust, is replace the knowledge underneath the work.
AI gives leverage to engineers who already know what they are building and why. It exposes, harshly, the gap where the underlying understanding is not there. Used well, it lets someone with taste and judgement move much faster through the shape of a problem. Used as a substitute for taste and judgement, the output is technically present and practically worthless. The difference is not the tool. It is the person using it.
This is why our core delivery team stays small, on purpose. Not cost discipline, not resistance to hiring. The people we have already know the products, the customers, and the reasons a particular decision was made three versions ago. When they reach for AI as a tool, not a replacement, the leverage is real and the work gets better.
The same philosophy runs the other way in our products. Tris, our AI agent inside Reservate, is not there to tell a pet care operator how to run their business. It takes what the operator already knows about their opening times, their seasonal exceptions, their quirks, and turns that knowledge into a working configuration faster than clicking through a form ever could. AI in service of someone who already has the knowledge, whether that person sits on our team or uses our software. That is the version we build with, and the version we ship.
I should say one thing plainly. We are not trying to have a big 2026. We are trying to have a useful one, for the founders we back, for the customers they are finally reaching, and for the people those customers represent. The most reliable way to have a memorable 2027 is to get the quiet work of 2026 right.
Thank you, as always, for reading. You know where to find me.