1 in 5 Australian Small Businesses Say AI Is Transformational. What About the Other 4?

Four in five Australian small businesses using AI aren’t seeing transformational results. That’s the number buried in the headline of NAB’s new research, and it matters more than the one NAB led with.

The bank surveyed Australian SMEs on their AI adoption and found that 40% are actively using AI, with a further 13% planning to. Among those using it, 22% report transformational improvements. One in five. Good number. Except that means 78% of the businesses putting AI to work are getting something less: incremental gains, modest time savings, or nothing they’d call transformational at all.

So the question worth asking isn’t “why is AI transformational for some businesses?” It’s “what’s stopping the other four?”

What the research actually found

NAB’s research (released mid-2026) shows the benefits cluster around productivity (58% of AI users cite this), marketing (46%), and customer service (36%). Revenue and profitability sit well back, at 7% and 9% respectively.

That gap is important. AI is changing how work gets done before it changes what ends up in the bank account. NAB Head of Behavioural and Industry Economics Dean Pearson put it plainly: “For many SMEs, change begins inside the business, in faster workflows, better use of information and improved customer engagement, before it shows up in financial performance.”

Business Services firms are the furthest along. NAB found 34% of firms in that sector reporting strong improvements, well ahead of other industries. Trades, hospitality, retail? Still working it out.

The real barrier isn’t the tools

Here’s where the data gets uncomfortable. SMEs rate their own workforce’s readiness for AI at 3.8 out of 10. Only 17% say their team is highly prepared. And when asked how confident they are that AI will create new jobs in their industry, the average score is 2.7 out of 10.

Think about that second number. Businesses are adopting a technology while being deeply unsure it will benefit their people. That’s not a recipe for change. That’s cautious tinkering with one eye on the exit.

The 3.8/10 readiness score is the more useful one. AI tools are widely available and often cheap. ChatGPT, Copilot, Claude: you can be up and running in an afternoon. But knowing what to use it for, how to build it into existing workflows, how to train staff to get consistent results, that’s where most businesses stall. Access to AI was never the problem. Knowing what to do with it is.

NAB Executive for Small Business Krissie Jones described businesses that have crossed this gap: a Queensland manufacturer cut quote turnaround from a week to 10 minutes; a Sydney mechanic built a voice AI to handle missed calls. Both found AI through a specific operational problem, not a general push to “use more AI.”

What the 22% are doing differently

The businesses reporting transformational results share a pattern. They started with productivity, not revenue. They picked one broken process, found a tool that fixed it, and built from there.

Peter Leondis of Leondis Automotive (Sydney, three sites) built a voice AI called “Zoe” to handle missed calls and follow up incomplete bookings. He wasn’t chasing a growth strategy. He was solving a specific, costly problem: enquiries that fell through the cracks as the business expanded. Once Zoe was working, the downstream effects showed up in lead conversion and customer acquisition costs.

That’s the pattern. Start with a pain point. Build a narrow solution. Measure what changes. The 78% who aren’t seeing transformational results are often doing the opposite. They’ve signed up for a tool, given it a vague brief, and wondered why the revenue didn’t follow.

What this means right now

If you’re in the 78%, the gap probably isn’t your tools. It’s one of two things: you haven’t found the right process to fix yet, or your team doesn’t know how to use AI consistently enough to make it stick.

The workforce readiness score of 3.8 is worth sitting with. Before buying another subscription, ask: does my team know what we’re using AI for? Do they have scripts, examples, and clear outcomes to aim at? If the answer is no, the technology isn’t the constraint. The capability is.

NAB’s research is a useful benchmark. But the benchmark that matters more is your own: what’s the one process in your business that costs the most time, loses the most leads, or produces the most inconsistent results? That’s where AI should start. Not with change. With that one thing.

Sources

• NAB SME AI Adoption Research (2026): AI changing how Australian SMEs get work done
• NAB Research PDF: Benefits from AI (PDF)

📖 More deep dives: News Deep Dives

📖 More deep dives: News Deep Dives

📊 Compare AI tools side by side | 💼 Free resources & AI prompt packs

More hot takes: Hot Takes →