It's Wednesday, 4:47pm. Maria runs a hair salon in Newtown. She has 23 unread WhatsApp messages from clients. Three invoices to chase. A BAS deadline in 11 days. She's about to spend two hours writing follow-up emails to four clients who haven't rebooked.
She's done this every week for six years. It eats two evenings.
(Maria is a composite drawn from conversations with Sydney-area salon and small-business owners, not a real client. The work she describes is real. Most owners reading this do some version of it.)
This is the work AI can actually help with. Not the cool stuff. Not the demos. The boring stuff.
And new data from MYOB shows that small businesses doing exactly this — using AI for the boring stuff — are growing 2.8 times faster than businesses that aren't.
The number that should make every owner pay attention
On 22 April 2026, MYOB released the latest results from its Bi-Annual Business Monitor. The survey covered 1,087 Australian SMEs. Alongside the survey, MYOB analysed anonymised income data from hundreds of thousands of small businesses across Australia and New Zealand over time.
The headline finding: Australian SMEs using AI in their day-to-day operations are growing 2.8 times faster than those that aren't.
This is not a vendor pitch about a single product. It is a comparison of median income growth across two groups of real businesses. One group uses AI tools. The other doesn't. The first group is pulling away.
40% of Australian SMEs are now in the first group. The other 60% are still deciding.
What "using AI" actually looks like
When MYOB asked the AI users what they were doing with it, the answers were not about chatbots replacing humans, or autonomous agents running operations. They were about time.
54% of AI users say AI saves them time. 34% say it improves their productivity.
The most common applications are mundane:
- Invoice reminders that send themselves on a schedule
- BAS prep that gets pre-populated and flagged for review
- Bank reconciliation that auto-matches transactions
- Customer follow-ups drafted automatically, edited manually, sent in batches
- Cash flow projections that update without a spreadsheet rebuild
That is the boring AI. It is not the AI you read about in headlines. It is the AI that quietly removes three to five hours of admin from your week, every week. Across a year, that is the difference between working until 7pm on Wednesdays and going home for dinner.
The Australian sectors leading adoption are predictable: retail and trade, professional services, healthcare, hospitality. The sectors that historically run on phone calls, paper, and "I'll get to it tomorrow."
Why almost half of small businesses are standing still
The same MYOB data shows the other side. 46% of Australian SMEs say they are not using AI and have no plans to over the next 12 months.
Not "considering it." Not "researching it." Actively choosing not to.
There is no judgement in this. The reasons are practical and most owners give some version of these:
- "I don't know where to start." (The first post in this series covered this — 1 in 3 SMEs say the same.)
- "I tried ChatGPT once and it made up a fact about my business."
- "I think you need a developer to set anything up."
- "I'm too busy. The day already runs me, not the other way round."
- "It feels risky. I'm not sure what could go wrong."
These are honest answers. None of them are wrong on their face. But the 2.8x growth gap is the early shape of a divide MYOB CEO Paul Robson described as widening. The longer half of the market sits still, the further behind it falls.
The training nobody is doing
Here is the part that should worry anyone planning ahead.
Two-thirds of Australian SMEs are not actively seeking AI experience when hiring. 72% have no plans to provide AI training to their team.
Read that again. 72%.
So the businesses standing still are not just choosing not to use AI today. They are not preparing to use it next year either. The team they hired in 2024 is the same team they will have in 2027. The competitor down the road, who is hiring this month and asks one extra question — "have you used a chat tool to draft customer emails?" — just gained a small but compounding advantage.
This is not about replacing people. The MYOB data is clear: AI is being used to remove boring admin, not headcount. But the salons, clinics, and trade businesses that quietly hire AI-literate staff in 2026 are the ones that will look 2.8x ahead in 2028.
What Maria did
Maria, the Newtown salon owner, started small. One task. Drafting follow-up emails to clients who hadn't rebooked.
She used a free AI tool. She gave it the client name, the last service they booked, and the date. The tool drafted a friendly, three-sentence email. Maria edited the tone in the first few drafts. After a week, the drafts were good enough that she could send them with one or two small tweaks.
The task that used to take 90 minutes once a week now takes 15. That is 65 minutes back. Five and a half hours a month. 66 hours a year. Almost two working weeks.
She is doing the same with quote follow-ups now. Then BAS prep. Then social media captions.
This is what the 40% are doing. Not magic. Not transformation. Just one boring task at a time.
What you can do this week
The 2.8x number does not come from a single big AI deployment. It comes from many small ones. So the answer to "where do I start" does not have to be ambitious.
Pick one task that:
- You do at least three times a week
- Takes between 15 and 60 minutes
- Mostly involves writing, summarising, or rewriting
Examples that come up repeatedly with Australian small-business owners:
- Drafting WhatsApp responses to customer enquiries
- Writing follow-up emails after a quote
- Summarising weekly numbers from your accounting software
- Translating customer reviews into a short reply
- Cleaning up a meeting note into a list of next steps
Try one task. One week. Honestly assess whether it saved you time. If yes, do another. If no, drop it and try a different one.
This is the boring, unglamorous, deeply practical reason 40% of Australian small businesses are growing 2.8 times faster than the rest. They started with one task.
Coming next
The next post in this series is about the opposite end of the AI conversation: the agencies and freelancers selling miracle AI builds at high speed. Recent research from Georgia Tech on what is known as "vibe coding" — AI-generated code shipped without review — shows the number of vulnerabilities in AI-built software jumped from 6 to 35 critical security flaws between January and March 2026 alone.
Before you sign anything with an AI agency, you need to know what to ask them. That is Post 4.