From Wednesday, NSW small businesses can book free one-on-one sessions with a business adviser. The NSW Small Business Advisory programme replaces Business Connect, which was cut in last year’s budget, and it comes with $37 million over four years. It includes explicit support for businesses adopting AI. Free expert time is worth taking. It is also routinely wasted, and it gets wasted in a predictable way.
What the NSW Small Business Advisory actually offers
According to Service NSW’s own summary of the review, the programme runs through the Service NSW Business Bureau. The core offer is fee-free one-on-one advice, delivered in person, by phone or online. There are also networking events, webinars, and a library of self-service resources. Funding is $7 million this financial year, then $10 million a year for three years.
Two things are worth knowing before you get your hopes up. This is the initial phase, not the finished service — SmartCompany reports the full suite of activities arrives in February 2027, with the wider rollout to priority groups in the first quarter of that year. And it is NSW only. Core services are open to small businesses across the state, with priority support for construction, hospitality, Aboriginal-led, women-led, multicultural and regional businesses.
There is also an AI component: tailored support for businesses looking to adopt AI tools, plus an AI-powered Business Needs Assessment tool and a resource navigator.
The question that wastes the session
It is “how should we be using AI?”
It is a fair thing to wonder and a terrible thing to ask. An adviser who has never seen your business can only answer it in categories. Try a chatbot for customer enquiries. Look at automating your invoice entry. Consider a tool for drafting quotes.
All of that is true, none of it is wrong, and you could have read it anywhere. You have spent your free hour receiving an article. The same goes for “should we get a dashboard?” and “are we compliant?” — questions broad enough that the honest answer is always “it depends”.
What a better question looks like
A good question is one where the answer changes what you do on Monday. In practice that means it is specific, bounded, and carries its own context.
So instead of asking how to use AI, describe the job. You hand-key roughly two hundred supplier invoices a month, it takes a day and a half, and you want to know what automating the data entry would involve and where it usually goes wrong. That is answerable. An adviser can tell you what similar businesses tried, what it cost, and what broke.
Instead of asking whether you need a dashboard, describe the disagreement. Three people report monthly revenue and produce three different numbers, and you want to know how businesses in your sector settle the definition. Instead of asking whether you are compliant, name the obligation. From 10 December, businesses covered by the Privacy Act have to disclose significant automated decisions in their privacy policy, and you want to know whether the way you screen job applicants counts.
Bring the numbers with you. How many, how often, how long it takes now. The adviser cannot supply those, and without them every answer collapses back into general advice.
The part to be sceptical about
The AI-powered Business Needs Assessment tool will be a questionnaire that routes you to resources. That is triage, and triage is useful. It is not a diagnosis, and I would not treat its output as a finding about your business.
The deeper limit applies to the AI advice generally. An adviser can tell you which tools exist and what other businesses in your industry are doing. What they cannot tell you in an hour is whether your own records are consistent enough for those tools to return something you would trust. Whether your supplier list holds three spellings of the same company is not a question about AI. It is a question about your data, and it decides whether any of this works.
None of that is a criticism of the programme. It is just the difference between advice and diagnosis, and it is worth knowing which one you are booking.
What it is genuinely good for
Plenty, if you aim it properly:
- Operators in their first couple of years, who mostly do not know what they do not know yet. This is the highest-value case by a distance.
- Navigating grants and government programmes, which is genuinely hard to do alone and is exactly what a Service NSW adviser will know cold.
- Sector-specific questions, given construction and hospitality are named priorities.
- A free second opinion before you spend money on something. Even a sceptical half-hour is cheaper than the wrong purchase.
- Regional businesses without an accountant or adviser nearby.
Before you book
Four things, none of which take long:
- Write your question down in one sentence. If you cannot, that is the work to do first, and it is more valuable than the session.
- Collect the counts and timings that go with it.
- Ask the adviser directly what they cannot help with, and who does that. A good adviser answers this happily.
- Book early. The initial phase is narrower than what arrives in 2027.
If the honest answer to “what would I even ask?” is that you are not sure, the useful first step is working out what data you already hold and where it contradicts itself. That is what a data health check is for, and it turns a vague sense that something could be better into the specific question worth taking to a free hour with an expert.