Before hiring a tech manager, 11-50 staff businesses should ask whether a whole AI team on Units costs less than one mid-level hire.
You’re finally ready to move forward with the new project, and you’re about to post the job ad. The role is something like “Operations and Technology Manager” or maybe just “someone who can handle the tech stuff.” Your business has 30 team members. The number of subscriptions have started to pile up. Your technology MSP doesn’t really know anything and is just keeping the lights on. Nobody owns the roadmap, and every time something breaks or a new tool gets proposed, it lands on whichever manager is nearest.
So you write the job description, set a salary band around what a capable mid-level person costs, and prepare to hire.
Before you post it, one question is worth sitting with: Did you know that for roughly the same monthly cost as that one hire, you could have a whole team?
What one hire actually covers
A mid-level technology manager at a 30-person business does a lot of things adequately, but very few things well. They manage the vendor pile. They field the IT questions. If they’re good, they start to document how the business runs and where the gaps are. That’s genuinely useful, and most businesses at this size have needed it for a while.
What they rarely do is build. Most mid-level hires aren’t AI engineers. They’re not data architects. They’re not the person who maps your quoting process end to end and then builds an agent to handle the first pass. That work either doesn’t happen, or you hire again.
The nature of technical work has shifted. Senior engineers now use AI to solve problems that used to require a team, which means the old job description, the one you’re writing now, is already out of date. You’d be hiring for a version of the role that AI is changing underneath you. The hire you make today may cover less ground in two years than you’re expecting it to cover now, because the work is moving faster than a single person can follow.
There’s also the time it takes to get a new hire to full productivity. You write the ad, screen candidates, interview, negotiate, onboard, and wait while they learn how your business runs. That process takes months, and the technology problems don’t pause while it happens.
What a team on Units covers
An AI Partner starts where the hire would (or in a lot of cases, should): understanding how the business runs. That’s People Units, business analysts mapping your processes and documenting the work, solution architects working out where AI belongs and where it doesn’t. Most engagements start here, and it’s the part that makes everything after it work. You’re not handed a plan and left to execute it. The team owns the execution too.
From there, Compute Units keep the infrastructure steady and current. Token Units are what the AI agents consume as they take on the work your team mapped out. The mix shifts month to month as the engagement matures, so you’re not paying for people when you need compute, or paying for compute when agents are doing the lifting.
The cost anchor is the same as your hire. One mid-level salary each month, against a team that covers strategy, build, and the running of it, on the current best AI model, without locking your business to any vendor’s mould. And because the mix of Units adjusts as the work changes, you’re not carrying a fixed headcount through the months when the work is lighter.
What you don’t get with a hire
A single hire brings their own experience and stops there. They don’t bring cross-industry expertise from dozens of prior engagements. They don’t come with agentic harnesses already built and tested. They don’t know what worked for a wholesale distributor in Brisbane last quarter, or what failed for a professional services firm in Melbourne the quarter before. That kind of pattern recognition takes years to build and usually lives in a team, not a person.
They also leave. And when they do, the knowledge walks out with them. The documentation, if it exists, is a snapshot of what one person understood on the day they wrote it. Rebuilding that context is slow and expensive, and at a 30-person business, there’s rarely anyone left who knows enough to pick it up quickly.
An AI Partner owns the roadmap, the build, and the record. Your data, your prompts, and your workflows sit apart from any model, so you’re not starting over if the relationship changes or a better model comes along. The business keeps what was built, and the next step is always clear.
What this means for you
This isn’t an argument against hiring. Some businesses at 30 to 50 staff genuinely need a person in the building, and an AI Partner can work alongside them. The argument is narrower: if you’re about to hire because nobody owns the technology function and AI keeps coming up in management meetings, posting the job ad might not be the fastest or cheapest way to fix that.
The maths changes when a whole team costs what one person costs. Take the AI Roadmap Interview and talk through where your business is stuck. You’ll get a custom plan for where to start, and a clearer picture of what the right move actually is.
