Should a small business hire an AI consultant?

Small businesses are better positioned for this than large ones and they consistently believe the opposite.

The advantage is decision speed. A company where the owner can change a process on Tuesday does not need a steering committee, a change programme, or eighteen months of stakeholder alignment. Large organizations spend most of their AI budget on coordination. You can spend yours on the actual thing.

The disadvantage is that you cannot absorb a wasted quarter. So the sequencing matters more for you than for anyone.

What small companies should buy, in order

First, nothing. Spend a month using the general-purpose tools yourself, personally, on real work. This costs a subscription and it does two things: it builds judgment you cannot outsource, and it makes you much harder to mislead. Where do-it-yourself genuinely wins.

Second, a diagnosis, but only if you cannot rank your own opportunities. If you can already name the three processes you would fix in order, skip this and go to the third item. If you cannot, this is the cheapest useful purchase available and it is a natural stopping point.

Third, one build, deliberately small. Something at the edge of your operation, high frequency, low risk if it goes wrong. The point of the first build is learning how your business reacts, and that is cheaper to learn small.

Fourth, decide who runs it. Small companies rarely have a spare technical person, which makes this the item most likely to be skipped and most likely to matter in a year. What happens after.

The trap

Buying a strategy when you needed a builder.

Small companies are the most susceptible to this because the pitch is reassuring: you are busy, you do not have a technical team, so somebody should tell you what to do. But if you already know what to fix, a diagnostic engagement will hand your own list back to you in better formatting, and you will have spent the build budget on the analysis.

The test is one sentence: can you name your top three, in order, with a rough sense of what each is worth? If yes, hire construction. The full readiness test.

Match the firm to your size

A large consultancy engaged by a small business produces a junior team working from a template, at a price that assumes an enterprise budget. That is not dishonesty, it is what their model does.

What you want is somebody small enough that a senior person is actually on your project. Ask who specifically does the work, and get names in the agreement.

The corresponding risk with a small firm is capacity: no bench, and a problem if someone is unavailable. Ask what happens then. We are small, so that question applies to us.

If you are a mission-driven organization rather than a business, the arithmetic is different in kind rather than degree, and that version is here.

The one thing worth spending real money on

Not the build. The judgment about what to build.

A wrong first project costs you the fee, the internal credibility, and usually a year before anybody is willing to try again. That last cost is the one that hurts, because the conclusion drawn inside the company becomes "we tried AI and it did not work for us", and it is very hard to reopen.

Which means the conversation where somebody tells you not to do something is worth more than the build itself.

What to refuse

Long commitments. Nothing about your situation requires a year-long term, and a supplier who needs one is managing their revenue rather than your risk.

Anything running on their infrastructure under their accounts. This matters more for small companies because you have less leverage in the renegotiation you will eventually have. How to avoid it.

A pilot with no defined success condition. If nobody will say what result would mean stopping, the pilot cannot end. How that pattern works.

The honest case for waiting

Some small businesses should not do this yet, and the signs are specific.

If your process is unclear or disputed, fix that first on paper. If nobody senior can give the project real attention for a few weeks, wait until they can, because an engagement run around an absent owner produces a document nobody acts on. And if your constraint is demand rather than operations, automation will not create customers. Which problem class you are in.

Our own work is aimed at exactly this size of company, which is why the roadmap is priced as a fixed fee with a hard stop and no obligation to build with us: how it works. If the answer after one conversation is "you already know what to do, go build it", that is the answer you will get.