Business Counts Are Rising. That Does Not Make a Young Firm Safe

A country can add thousands of businesses and still be a difficult place to keep one alive.

Australia offers the clearest current example. The Australian Bureau of Statistics counted 2,814,778 actively trading businesses at 30 June 2026, an increase of 85,130 over the year. Beneath that net gain sat much larger movements: 460,461 entries and 375,331 exits. The business population grew by 3.1%, but roughly four enterprises left the count for every five that entered.

The United Kingdom's latest complete annual demography tells a related story at a different speed. The Office for National Statistics recorded about 317,000 business births and 280,000 deaths in 2024. The headline balance was positive. The harder statistic appears later in the same release: only 38.4% of UK businesses born in 2019 were still active five years later.

These numbers are not a league table

It would be tempting to put the Australian entry rate of 16.9%, the UK birth rate of 11.1% and a US startup measure in adjacent columns, then declare a winner. That would be poor analysis. Each statistical agency defines its unit, register and time period differently, and the words business, firm, enterprise, establishment, birth and entry are not interchangeable.

The UK's register is built mainly from businesses appearing in tax and employment systems. It defines a business as an enterprise with decision-making autonomy and counts activity through turnover or employment. Australian counts cover actively trading businesses on the national business register but distinguish those from the larger population merely holding an Australian Business Number. US Bureau of Labor Statistics survival data follow private-sector establishments—individual operating locations—not necessarily entire legal companies.

The correct comparison is therefore directional. All three systems show that business populations are dynamic, closures are normal and the passage of time removes a substantial share of each starting cohort. None provides a personal probability calculator for a particular bakery, software company or manufacturer.

Australia's growth came with heavy turnover

Australia's 2025–26 entry rate rose to 16.9%, while its exit rate was 13.8%. Entries exceeded exits by 85,130. Only 996,203 of the 2.81 million actively trading businesses were employing, which is a useful reminder that a rising business count does not translate directly into the same number of new employers or jobs.

The industry detail also resists a single national story. Health care and social assistance recorded the largest percentage increase in business numbers, at 6.7%. Transport, postal and warehousing rose 4.9%, and financial and insurance services rose 4.3%. Agriculture, forestry and fishing was the only broad industry with a decline, down 0.5%. Manufacturing grew just 0.2%.

For an owner, those figures describe the environment rather than the business model. Rapid growth in the number of firms may signal demand, but it can also bring more competitors. A flat industry count can conceal productive expansion by established firms. Headcount, revenue, margins and productivity are absent from a simple population total.

The UK's five-year filter is severe

UK business deaths fell to 280,000 in 2024, the lowest death rate since 2016. Births exceeded deaths by 37,000, and the gap between the two rates widened to 1.3 percentage points. Yet the 2019 cohort's five-year survival rate was 38.4%. Both statements can be true: annual conditions improved in 2024, while most firms from an earlier starting group did not remain active for five years.

Industry matters. Transport and storage had the UK's highest birth rate in 2024, at 15.6%, but also its highest death rate, at 16.5%. High entry did not imply easy endurance. By contrast, the ONS counted 14,330 high-growth businesses among enterprises with at least ten employees, equal to 4.9% of that eligible population. Information and communication had the highest high-growth share, at 9.2%, followed by finance and insurance at 7.3%.

The high-growth definition is demanding: average employment growth above 20% a year over three years, among firms beginning with at least ten employees. It excludes microbusinesses by design. That makes it useful for tracking scale-ups, not for judging whether a self-employed owner or a three-person shop is performing well.

US evidence shows what the long horizon does

The US Bureau of Labor Statistics reported that 34.7% of private-sector establishments born in March 2013 were still operating in March 2023. The ten-year figure is not a forecast for firms opening now, and it spans a cohort that experienced both the pandemic and very different industry conditions. Its value is simpler: survival continues to fall long after the first anniversary.

BLS one-year data also show that survival changes with geography, industry and the business cycle. A new establishment born near a recession does not face the same environment as one launched during expansion. Meanwhile, the US Census Bureau's Business Formation Statistics track applications and project likely employer startups. For July 2026, Census projected 29,959 employer startups would form within four quarters from that month's applications. An application is a signal of intention; it is not yet an operating employer and should not be reported as one.

What the comparison says about the first five years

The first lesson is to separate launch activity from operating strength. Registrations, applications and new tax accounts measure movement into business. They do not show whether customers repeat, invoices are collected, labour becomes productive or the owner earns an adequate return.

The second is that survival is a cash and customer problem before it becomes a national statistic. A firm can have positive accounting profit and still fail when stock, tax or late invoices absorb its cash. It can grow sales and lose money on every order. A useful operating dashboard therefore follows gross margin by product or service, cash runway, overdue receivables, customer concentration, repeat purchase, inventory days and capacity—not registrations or social-media reach alone.

Third, sector churn should change planning. The UK's transport figures show an industry where both entry and exit are unusually high. A founder in such a market should test utilisation, maintenance, insurance, fuel, platform fees and price competition under a downside case. In a professional service, the constraints may instead be staff time, client concentration and delayed payment. National averages cannot replace the unit economics of the chosen sector.

Survival is not the only definition of success

A closed registration can represent insolvency, but it can also follow a sale, merger, retirement, restructuring or a deliberate return to employment. An active business may survive without paying its owner well. Statistical survival is therefore necessary context, not a complete verdict on entrepreneurial success.

Growth requires a separate test. The UK figures show that rapid employment expansion occurs in a minority even among businesses already large enough to qualify for measurement. Owners and lenders should distinguish a sound small firm from a scale-up. Both can be valuable; they need different capital, controls and expectations.

A better way to use national business data

An entrepreneur can use these releases to challenge assumptions rather than predict fate. Start with the official definition. Compare the chosen sector with the national total. Check whether the figure describes applications, businesses, employers or locations. Note the cohort year and any provisional data. Then return to the company's own evidence: customers, price, margin, cash, staff capacity and compliance.

The international picture is neither a slogan about inevitable failure nor proof of an entrepreneurial boom. Australia added businesses while processing hundreds of thousands of exits. The UK improved its annual birth–death balance while its five-year cohort remained thin. US data show further attrition across a decade. The honest conclusion is less dramatic and more useful: entry is common, endurance is difficult, and the work that improves survival begins inside the operating model.