When it comes to AI adoption, Australian small businesses appear divided into two distinct camps: the curious and the cautious. Some have dived straight in while others have held back, waiting to see whether the technology would become a workplace fad or feature.
As an AI consultant and author working with SMEs, Tracy Sheen has watched the split emerge. “I’m starting to see a bit of a chasm opening up,” she says. She meets with some business owners who were early adopters and others who fear they’ve missed the boat. But she offers reassurance. “This happens with any kind of tech leap: those getting on board versus a lot of nervousness.” In her book, AI & U: Reimagine Business, she stresses it’s not too late to become a late adopter, providing practical advice to help owners tailor AI to their businesses.
Barriers To AI Adoption Are Mental And Physical
There’s a distinct trend in the rapidly-changing digital landscape: the Federal Government’s AI Adoption Tracker last June put the SME uptake rate at 41 per cent, while Deloitte in November suggested two-thirds were using AI but only 5 per cent had full implementation. It’s not as embedded as some may think.
Tracy Sheen notes parallels with the introduction of social media. “We had a lot of SMEs that said, ‘I don’t like social media. I don’t like Facebook. I’m not on Instagram,’” she recalls. “They didn’t understand it, so it took them a little bit longer to get the gist of why they needed to start getting involved in that revolution. And it’s the same now.”
She says hesitation is generally the result of overwhelm. “It’s one more thing I’ve got to learn,” owners tell her. “I’m already spinning all of these plates and now you’re telling me I’ve got to get my head around this stuff as well?”
Others face physical blocks. Sheen highlights connectivity barriers for regional businesses. “Some people are leaping forward and going, ‘We’re all in with AI,’ while others are still just getting decent internet speeds.”
Despite Confusion, It Will Pay To Learn The Lingo
Jargon is another barrier to entry. Sheen says talk of LLMs, GPTs and vibe coding can feel understandably foreign.. She cites the example of customised AI assistants, which can be programmed for specific tone or brand voice: they are called GPTs by Open AI and Gems by Google but are functionally the same.
Despite the overwhelm, she says even busy business owners must learn the ABCs of AI. “Much like you need to know how to read a P&L or a bank statement or how to do inventory – you need to know the basics of these glossary terms in AI.” She says AI assistants like ChatGPT, Google Gemini or Claude make it easy for users to start from scratch since they use natural language processing. “If you’re just opening up an AI account today, you say, ‘Hey, this is my first conversation. I have absolutely no idea what I’m doing here. Help.’ Start prompting it to ask you questions.”
While leaders learn about AI, Sheen says they must also teach their AI assistants about their businesses, likening the experience to onboarding a new employee. “You need to make sure that you’re educating your Large Language Model around who you are, who the company is, what your role is, what your expectations are, what their guidelines are, what you want help with. Because the more information that you can give it, the better the responses are going to be and they won’t sound beige.”
Have An AI Policy That Puts Humans In Charge
SMEs that are in the AI race often haven’t made it far out of the blocks. Decidr’s AI Readiness Index found 76 per cent of Australian SMEs had no formal strategy or roadmap, even though 83 per cent believed the technology would significantly impact their operation. Sheen believes a lack of AI governance is a major problem. “One of the things I’m very passionate about is that AI and technology should only support humans. Only augment, not replace,” Sheen insists. “We need governance in place so everybody knows where the line in the sand is: what is acceptable, what isn’t.”
Sheen says the fact AI adoption has often been employee-led, rather than the result of a business-wide strategy, has meant the technology has been used without guardrails and with unexpected risks. “An employee might set up something, but maybe it’s in their own private ChatPT account or private Microsoft account. Well, then, what data have they been sharing? What are those kinds of breaches and protocols?”
She recommends SME owners involve staff in strategic conversations from the beginning to ensure protocols are enshrined that protect client and proprietary data. Sheen also points out if they don’t wish to hire a consultant, businesses can use AI assistants, perhaps ironically, to prepare an AI policy. She laments the fact the absence of broader regulation by tech companies or governments means SMEs have to bear much of the burden themselves. “If you’re a business owner, that’s a scary place to be, going, ‘What’s my exposure like here? What do I need to be worried about?’”
Think Critically To Find The Problems You Want AI To Solve
For businesses with ad hoc AI adoption, Sheen recommends a refresh to ensure workflows that emerged organically are actually best practice. “A lot of the conversation now is to go, ‘Okay, let’s just pause. Let’s go back and have a look at how you set this stuff up. And let’s have a talk about how you’re interacting with these systems.’” She says the power of prompt engineering – writing instructions to elicit the best outcomes – cannot be overstated.
Businesses should also ensure AI tools are being devoted to solving the problems that most need solving. Says Sheen: “It’s really about getting everybody together and just finding out what’s causing them grief. We all have those things in our roles that perhaps take too much time. They’re monotonous, repetitive tasks that are easily replicable and probably better handled by the AI.” She gives examples in different fields: slashing the time taken to do rosters, optimising SEO and blogs for more hits, or updating policy documents before they’re out of date.
Product Manager, AI at Employment Hero Russell Dias suggests broadening horizons beyond ChatGPT. “When I say AI, most small business owners picture a chatbot. That’s about 1 per cent of what AI does. Right now, AI screens job applicants, schedules your team, follows up with customers who’ve gone quiet and catches compliance issues you’d otherwise miss.” He says it’s fine to start small. “My advice: see what other businesses in your space are doing – webinars and industry forums are a good start. Then pick one small problem and test a tool on it. One real experiment beats months of research, and over time you build a toolkit that actually fits your business.”
Tracy Sheen believes AI plans should be written in pencil and subject to continual evolution. Despite the inevitable strain on businesses in the early stages, she says owners will know they’ve succeeded in AI adoption when they have fewer plates spinning. “I always say, if you’ve found time back in your week, if you’ve streamlined an activity, or if you’ve become more efficient at something, that’s what good looks like. Get back and do the stuff you love and why you got into business in the first place.”
