No IT person, no clear starting point, no time to work it out. Here's why that's not a gap you fill by hiring, and what to do instead.
You’ve played with ChatGPT. Maybe tinkered with Claude a bit too. You can see they’re impressive, but when you try to picture how any of it connects to your quoting process, your client files, or the way your team actually works, the picture goes blank. So the tab gets closed, the week moves on, and AI stays in the “we’ll do that in the future” basket.
That scenario, not a staffing gap, not a lack of motivation, is what’s stopping most small and medium sized business owners. The question isn’t whether you have an IT person. The question is whether anyone in your business knows how to take a general-purpose AI tool and make it do something useful for the specific way you work. Almost nobody does when they start, and that includes most IT people.
An IT hire won’t solve this
It’s worth being straight forward about this, because the instinct to hire your way out of the problem is understandable. You see a capability gap, you think about who could fill it, and an IT person seems like the obvious answer. But general IT skills and AI implementation are different things, very different things. Someone who manages your network, keeps your laptops running, and handles your software licences has a completely different set of skills from someone who can map a business process, connect data sources, build agentic harnesses that carry out real work, and keep the whole thing current as the models improve. Hiring the first person to do the second job means paying for the wrong thing, and finding out months later.
The practitioners who have worked through AI rollouts across many industries are consistent on this point. The critical ingredient isn’t technical ownership inside the business, it’s a leader willing to back the direction, paired with a partner who owns the technical execution. You need to know what you want the business to do better. The how is someone else’s job.
What “connecting AI to your business” actually means
Tools like ChatGPT and Claude are general-purpose. Out of the box, they don’t know anything about your clients, your processes, or the way your team works. They can answer questions and draft text, but they can’t file a job, follow up a quote, or pull a report from your systems, because nothing connects them to any of it (and using their connection methods can be risky). Making AI genuinely useful for a specific business means building what surrounds the model. That includes process mapping, so the AI is pointed at the right work in the right order. It includes data pipelines that move and transform your business’s information to where the AI can use it. It includes agentic harnesses, the custom scaffolding that lets AI agents carry out real tasks reliably rather than just generating text for someone to act on. And it includes prompt optimisation, tuning how the AI is asked so it gives the right answer more often and costs less to run.
None of that is something a general-purpose tool does on its own. None of it requires your team to become technical. It requires someone who has built it before.
What a Technology Partner actually does
We work frequently with a five-person project management agency. They use three main tools that don’t talk to each other. Job details live in one place, client comms in another, and invoicing in a spreadsheet that one person maintains and everyone else is afraid to touch. The owner knows AI could help but has no idea where to point it, and nobody on the team has time to work it out.
A Technology Partner comes in and starts with the work that always comes first: process mapping. A business analyst spends the first weeks documenting how the business actually runs, step by step, because you can’t build AI into a process you haven’t mapped. That work is People-heavy early on, and it’s paid for through People Units in the monthly mix.
From there, the data gets connected. The tools that weren’t talking to each other start sharing information through data pipelines, so there’s one source the AI can work from. Then the first agents go in, pointed at the highest-volume task, which for this consultancy turns out to be the back-and-forth of scoping new work. The AI drafts the scope, flags the gaps, and routes it for approval. Nobody re-keys the same information across three tools anymore. As the agents take on more of the routine work, the monthly mix shifts. People Units give way to Token and Compute Units, because more of the work is being done by the AI rather than by a person setting it up. The owner didn’t need to understand any of the technical build. They needed to know their business, back the direction, and say yes to the process.
The person closest to the tech shouldn’t have to own this
In most small and medium sized businesses, there’s someone who ends up handling technology by proximity rather than by skill. They’re the one who calls the software provider when something breaks, sets up the new laptop, and gets asked to “have a look at” whatever AI tool someone read about on the weekend. That person is usually good at their actual job, and technology has become a second one they never asked for.
The Technology Partner model exists to take that weight off. Not by replacing the person, but by giving the business a dedicated partner who owns the technology direction, the AI build, and the continual optimisation and integration as the tools and the business both change. The person who was handling tech by default gets to go back to the work they’re actually there to do.
That’s the practical answer to “we don’t have anyone who owns this.” You don’t need to hire someone. You need a partner who already does it, for businesses like yours, every day.
What this means for you
If AI has been sitting in the too-hard basket because no one internally knows where to begin, that’s not a reason to wait. It’s the exact reason the Technology Partner model exists. The business doesn’t need to arrive with a plan or a brief. Working out where to start is part of the work.
The AI Roadmap Interview is where that starts. You talk through your team’s goals, the tasks that eat the most time, and where things keep falling through the cracks. From that conversation, you get a custom roadmap for where and how to begin.
