Small businesses carry the economy. The World Bank reports that small and medium enterprises represent around 90 percent of all businesses and account for more than half of global employment. In Australia the concentration is sharper still. They are the base layer, not a sideshow, and the figures below show by how much.
A small business is a firm with few employees. The Australian Bureau of Statistics defines a small business as one with fewer than 20 employees. "Small and medium enterprises", or SMEs, is the broader term used in most international data, usually capped near 250 staff. The thresholds differ by country, but the pattern they describe does not: a very large number of very small firms doing most of the work.
How many small businesses there are
Start with the count, because the count is the argument. In June 2025, 97.3 percent of all Australian businesses were small businesses. The same source puts the total number of active Australian Business Numbers in scope at 2,729,648 for 2024-25. So more than nineteen in every twenty trading entities in the country employ fewer than 20 people, and many employ none beyond the owner.
That ratio is not an Australian quirk. The World Bank's global figure of about 90 percent of all businesses tracks the same shape across developed and developing markets. When people picture "business", they tend to picture the listed companies in the news. The data points the other way. The typical business is a cafe, a trade, a clinic, a two-person agency. The large firm is the exception.
What small business contributes to the economy
A count alone could be dismissed as a long tail of tiny operators. The value figures close that argument off. Small business in Australia (firms employing fewer than 20 people) added over $590 billion of value in 2023-24, one third of Australia's total GDP. One in three dollars of national output comes from the smallest tier of firms.
Employment tells the same story at the global scale. The World Bank attributes more than half of jobs worldwide to SMEs. Put the two together and the phrase "economic backbone" stops being a slogan. A third of output and over half of jobs is structural load. If small firms stall, the whole economy feels it, because there is no other sector large enough to take up the slack.
There is a catch the same World Bank page names plainly. SMEs face a finance gap in the trillions of dollars across emerging market and developing economies. The firms doing most of the work are also the ones most starved of capital. That gap shapes which tools they can buy, which staff they can hire, and how fast they can adopt anything new.
Why a local dollar works harder: the multiplier
The local multiplier effect is the extra economic activity created when money is respent within the same community, and it is the mechanism that turns the counts above into something worth protecting. The figures say small business is big. The multiplier explains why its shape matters as much as its size.
Follow one dollar. Buy your coffee from an independent cafe and the dollar's next stops are mostly nearby: wages to staff who live in the suburb, an order to the local roaster, and the bookkeeper two streets over. Some of the dollar leaves at every step, but each round of local respending adds activity the town would not otherwise have. Buy the same coffee from a distant chain and a thinner slice stays behind: local wages and rent, while procurement, marketing, and profit route to a head office somewhere else. Neither purchase is wrong. They simply leave different amounts behind, and an economy built on 2.7 million mostly local firms leaves a great deal behind.
The second mechanism is speed. Small firms concentrate decision-making in the owner, which means they can change direction in days rather than quarters. A cafe can rewrite its menu on Tuesday because suppliers changed prices on Monday. That speed is a genuine edge against larger rivals, and it is one of the few advantages that scales down rather than up.
The third is distributed risk. Small firms carry bets that bigger organisations avoid: new methods, odd niches, untested products almost always start small. An economy with many independent firms has many independent experiments running at once, each failure small, each success available for others to copy. That diversity is what makes a region resilient. A town that leans on three large employers is fragile in a way that a town of three thousand small firms is not.
How small business is using AI
This is where the present moment turns, because a long-standing rule of technology adoption has quietly broken.
The assumption AI breaks
Every previous wave of business technology favoured the big end of town first. Mainframes, ERP systems, and early e-commerce each demanded capital, specialist staff, and long integration projects, so "wait for it to trickle down" was rational small business advice for decades. AI inverted that. The cheapest useful version is available to a sole trader on the same day it reaches a corporation, sold by subscription, with no hardware and no integration project required to start. For once, the assumption that new technology belongs to big firms is wrong. What has not changed is the know-how gap, and the adoption data shows it plainly.
Wide adoption, shallow use
In Australia, two-thirds of SMBs are using AI, according to Deloitte Access Economics research released on 25 November 2025. Usage is not the same as benefit. The same study found that just 5 percent of surveyed SMBs using the technology are fully enabled to realise its potential benefits. Most have opened a chatbot tab. Far fewer have wired AI into how the business actually runs: the quoting, the follow-ups, the customer records, the repetitive admin that eats an owner's week.
The size of that gap is also the size of the opportunity, and Deloitte put numbers on the rungs:
| Move up the ladder | What it looks like | Linked payoff (Deloitte) |
|---|---|---|
| Basic to intermediate | From an occasional chat tab to AI wired into a few routine processes | Around a 45 percent lift in profitability |
| Intermediate to fully enabled | AI integrated across operations, with staff skilled in using it | Around a 111 percent lift |
Those are not rounding errors. They are the difference between a business that survives and one that compounds.
For a practical starting point, the highest-return uses tend to be unglamorous. Drafting and triaging customer email. Answering common questions through a chatbot or automation so the owner is not the only support channel. Keeping the customer database clean and acted on. Summarising documents and meetings. None of these need a data-science team. They need someone to set them up correctly once. We cover the real boundaries of the technology in What AI Can't Do, because adopting it well also means knowing where it fails.
The risk hidden in the averages
A national average can disguise a split. "Two-thirds are using AI" reads like broad progress. Read against "5 percent are fully enabled" and a different picture appears: a small group pulling ahead fast, and a large group stuck at the surface. The barrier is not cost or scepticism. Deloitte found the top obstacle is not knowing where to start, a lack of awareness of what AI can do for a specific business.
That is a solvable problem, and solving it is worth real money to the country as well as the firm. Deloitte estimates that if just one in ten SMBs advanced one rung on the AI adoption ladder, $44 billion could be added to GDP annually. The upside is huge, and it is concentrated in firms that get guidance. Left alone, AI risks widening the gap between large and small. Directed well, it does the opposite. The deciding factor is access to help, not access to the tools, which are already cheap and everywhere.
Finding your first rung: a one-week exercise
The gap between using AI and benefiting from it closes one process at a time. This is the sequence we run in discovery, and you can run a version of it yourself this week.
- 1.Audit one week of repeated work. Your sent folder, your invoicing system, and your calendar are the record. Write down every task you performed more than twice: quoting, chasing invoices, appointment reminders, retyping details from one system into another.
- 2.Place yourself on the ladder honestly. AI use that lives in a browser tab you sometimes open is basic. AI wired into even one running process is intermediate. Be strict, because the payoff numbers above assume the move is real.
- 3.Pick one candidate process. The best first target has volume, repetition, and low judgement. Replies to common customer questions and invoice follow-ups are classic first wins. Pricing decisions, disputes, and anything judgement-heavy are the wrong place to start.
- 4.Set the measure before you build. Write down the hours that task takes now. Without the before number, you will never know whether the change paid.
- 5.Wire it in, or get help to. Setup is a once-off cost and the return repeats every week. This single rung is the move Deloitte's modelling links to that 45 percent profitability lift.
In our discovery work the list in step one almost always lands between three and five tasks, and owners are usually surprised by which one is largest. The surprise is the value. You cannot fix a cost you have never written down.
The Australian picture
Australia is a small business country by the numbers, more so than the global average. With 97.3 percent of businesses classed as small and that tier producing one third of GDP on over $590 billion of value added, the health of the small end is close to the health of the economy itself.
It is also a market where the AI question is live and quantified. The $44 billion figure is an Australian estimate about Australian firms, and the adoption split it describes is the local reality right now. The firms best placed to act are the ones with the structural advantages small business already holds: speed, close customer relationships, and deep focus on a niche. Pair those with even intermediate AI use and a two-person operation can match the output of a much larger team. We work through how to build a business that runs on systems rather than on the owner's constant attention in Systems Thinking for Small Business, which is the foundation any AI sits on top of.
The Enki approach
We work with small and medium businesses because that is where the impact is, and the figures back that choice. A third of national output and most of its jobs run through firms that have never had a real technology partner. The barrier was never the tools. It is knowing which tool, in what order, wired into which part of the business.
Our work starts with understanding before building, the same way a useful diagnosis comes before a prescription. We map how a business actually makes money, find the few processes where AI pays off first, and set them up so the owner is not left maintaining something they cannot see inside. We say no when AI is the wrong answer. If you want the groundwork before any of that, start with the value of discovery and audit. Small business is the backbone. Our job is making sure it has the tools to stay strong.