How Many Businesses In My Industry Are Failing? One Day In August Moved The Answer More Than The Year

Fewer than the published figure suggests. ASIC published its weekly insolvency series on 2 November, with data to 18 October. Across every industry, 4,817 companies entered external administration for the first time this financial year, against 4,405 in the same window last year. A rise of 9.4 per cent. Twelve of ASIC’s nineteen industry divisions recorded fewer failures than last year, and one day in August accounts for more than the whole national rise.

Outstaffer sells employment, so it has an interest in how you decide to fill a seat. Read on with that declared.

Business failures by industry, as the data actually falls

Counted like for like, 1 July to 18 October, construction moves from 1,071 companies to 1,557, up 45.4 per cent. That is 486 companies against a national rise of 412, so construction is more than the whole increase by itself. Health care and social assistance moves from 139 to 195, up 40.3 per cent. Professional, scientific and technical services rises 7.1 per cent, to 318.

Three other sectors went the other way entirely. Retail trade fell 9.1 per cent, financial and insurance services 6.6, and information media and telecommunications 22.1.

What does ASIC’s Series 1 actually count?

Companies, not businesses. ASIC’s Information Sheet 80 defines Series 1 as the first time a company enters external administration or has a controller appointed. One trading operation can sit under more than one registered company, each counting separately.

One appointment day beats the entire year’s increase

Filter the workbook’s own data set to 25 August 2026. There are 638 first-time appointments on that date, 587 of them voluntary administrations. One external administrator is named on 540, every one of them a construction company in New South Wales, and a second on 44, all of them health care in Queensland. That leaves 54 outside both blocks.

The national increase for the year to date is 412 companies. Those two blocks are 584 appointments between them. One day, recorded against two practitioners, exceeds the entire rise it is being read as evidence of.

Take the 584 out. The national count becomes 4,233 against 4,405, a fall of 3.9 per cent. Construction becomes 1,017 against 1,071, a fall of 5 per cent. Health care becomes 151 against 139, up 8.6 per cent rather than 40.3.

Which number belongs in a hiring decision?

The one your sector produces once a single day’s concentrated appointments are stripped out. Anyone who read the construction line in October and held a role was responding to 540 registrations in a single state, on a single day.

The honest counter is that 584 companies really did fail

They did. People in them lost work and trade creditors lost money. ASIC’s data set does not record group relationships, so this post claims none. What it does record is that all 584 are voluntary administrations, filed on one date, against two practitioners, across two states and two industry divisions.

The counter has a limit. Hundreds of appointments filed on a single date tells you about that date, not about demand in your sector next quarter.

Related reading

• What an hour of work produced in your industry this year
• What an Australian tech team costs per head

What I would do before holding a hire in a failing industry

Count decisions rather than registrations, because nobody publishes a usable benchmark for your size of firm and the public series is coarser than it looks. Check the appointee column before quoting a monthly figure. Then split whether a role is worth filling from how long you are committed once you fill it. Those two get collapsed, and the second is the adjustable half.

Sector risk is a reason to shape a commitment carefully and a poor reason to leave work undone. How long you are tied to a seat is settled at the point of hiring, and it is the half nobody prices. The harder case runs the other way, where your company fails and your team keeps an employer.

So the answer to how many businesses in your industry are failing is probably fewer than you read, and the figure deserves one filter before it changes anything. To run it on your own numbers, sign up and price a role properly before leaving the seat empty.

Which industry figure have you acted on this year without opening the file behind it?