An independent AI consultant or an agency? An honest Melbourne answer

This post answers one question: should you hire an independent AI consultant or an agency for your Melbourne business, and how do you actually tell which one fits your situation. Short version — it depends on whether you need one accountable person or a team with standing capacity, not on which option sounds more impressive. I’m the independent option here, so take that as the disclosure it is.

What’s the actual difference between a consultant and an agency?

An independent consultant is one person, doing the work themselves, accountable for the outcome with their own name attached. An agency is a business that assigns a team — sometimes a rotating one — and the accountability sits with the company rather than any single practitioner. Neither structure is inherently better. They’re built for different shapes of problem.

The practical tell isn’t the label on the invoice, it’s who you’d actually call if something breaks at 9pm. With a solo consultant, that’s one specific person who built the thing. With an agency, it’s whoever’s rostered onto your account that week — which can be a strength (someone’s always available) or a weakness (nobody has the full picture in their head).

When does an agency genuinely win?

When the work is a programme, not a project. If you need multiple workstreams running in parallel — a website build, an automation rollout, and ongoing content, all at once, with dedicated people on each — that’s more surface area than one person can cover well. Agencies also win when you need guaranteed ongoing capacity: a retainer where someone is always available regardless of any individual’s calendar, sick days, or holidays. And they win on breadth of specialism — if a project genuinely needs a designer, a backend engineer, and a strategist as three different skill sets working together, a team structure fits that better than one generalist stretching across all three.

If that’s your situation, an agency isn’t a compromise. It’s the right tool.

When does one accountable person win?

When you want a single point of accountability and no handoff loss. Everything the consultant knows about your business stays with the person doing the work — nothing gets lost in a briefing document passed to whoever’s free this sprint. There’s no account manager translating between you and the person actually building; you’re talking directly to them.

It also tends to win when the work benefits from continuity of judgement — the same person who scoped the problem is the one who built it and the one who’ll maintain it, so decisions made in week one are still visible to them in week twelve. And it wins on focus: a solo practitioner has one client’s problem in front of them at a time, not a portfolio of accounts split across a team’s attention.

Claim: for a single, well-scoped build or a focused advisory engagement, an independent consultant is usually the better fit than an agency. Evidence: the difference is structural, not a quality gap — a solo practitioner removes handoff and briefing loss but caps out on parallel capacity; an agency adds capacity and specialism breadth but adds a coordination layer between you and the person doing the work. Bottom line: match the structure to the shape of the problem. One clearly-scoped thing that one accountable person can own end to end favours a consultant. Several things running at once, or work that needs to keep running regardless of any one person’s availability, favours an agency.

Where do I actually fit?

I’m a pharmacist by registration and an AI consultant and builder by the work I do — the “AI Opportunity Audit” as a front door, then automations and agents, websites and products, or healthcare-specific AI work, depending on what the business needs. I do the work myself rather than assigning it to a team, which is the consultant end of the trade-off above: you’d be talking to and working with the person who actually builds the thing, not a rotating account team.

What that means in practice: I’m a better fit for a scoped audit, a specific automation, or a single product build than for a business that needs five workstreams running simultaneously with guaranteed always-on capacity. If your business is the latter, an agency is probably the more honest recommendation, even from someone selling the other option.

What about cost — is one cheaper than the other?

I’m not going to put a number on this. Rates across the Australian market vary widely enough by seniority, scope, and specialism that any figure I quoted here would be more misleading than useful, and I’d rather you get a real quote for your actual scope than anchor on someone else’s average. What I can say structurally: an agency’s rate typically has to cover more overhead — account management, bench time between projects, layers of review — while a solo consultant’s rate is closer to just the work itself. Whether that nets out cheaper depends entirely on the scope, so ask for a quote against your specific problem rather than trying to compare list prices.

How do you decide, concretely?

Write down what you actually need done, then count the workstreams. One clearly-scoped thing — an audit, a specific build, an advisory engagement — points toward a consultant. Several things running in parallel, or a need for capacity that doesn’t depend on one person’s calendar, points toward an agency. If you’re not sure which camp you’re in, that uncertainty is itself useful information to bring to whoever you talk to first, consultant or agency — a decent one will tell you honestly if you’re better served by the other option.

See the builds this has actually produced at /work.