AI Is Creating More Founders. The Ecosystem Is Already Struggling.

The Wave

In the first ten weeks of 2026, around 4,450 tech jobs were cut at companies operating in Australia, with AI named as the reason in nearly every case (ACS, 2026). WiseTech announced about 2,000 cuts worldwide in February, to be implemented over the next two years. Atlassian announced 1,600 cuts globally on 11 March. Telstra cut 650 here in Australia. The Australian portion of the WiseTech and Atlassian numbers isn't fully broken out, so 4,450 is best read as a ceiling, not a confirmed local count.

A lot of these people will end up starting something. Not because they planned to, but because the bills don't stop. The redundancy payout runs out. The mortgage doesn't. The job market is tight. So they take the leap.

Researchers have a name for this. They call it necessity entrepreneurship: people who start a business because they have to, not because they want to. The Global Entrepreneurship Monitor's 2024-25 global report finds that fear of failure now deters 2 in 5 adults globally from acting on a good business opportunity. Necessity-driven founders carry that fear into the venture itself, with less margin for error and less time to course-correct (GEM, 2024-25).

And many won't call themselves founders at all. They'll hang out a shingle and call themselves a consultant, a contractor, a freelancer, a solo practice. The label doesn't change the work. Eric Ries, who wrote The Lean Startup, defines a startup as "a human institution designed to create a new product or service under conditions of extreme uncertainty" (Ries, 2011). A solo consultancy fits that definition exactly. Same questions: what are you selling, who's buying it, at what price, with what unit economics, is the model repeatable enough to keep paying you next year, and what happens when the first client referral runs dry. The failure rate doesn't care which word you use.

You won't find them in any official count yet. But anyone in a founder Slack group has already met them.

And these aren't 22-year-olds with a Y Combinator stipend and ten years of runway. They are mid-career professionals. Many have a mortgage. Many have young families. They have a redundancy payout that's already draining and no salary backstop. They have months to find direction, not years to experiment and pivot. Every wrong turn is paid for in runway they can't get back.

The Ecosystem Was Already Failing

Start with the path founders are told to take. Get into a top accelerator. The reality: Antler Australia, the country's largest residency program, runs about 1,000 applicants through its filter to land roughly 80 founders in the Pre-Launch program. Fewer than 1 in 10 get a foot in the door, and that's just into the residency, not into investment (Antler Australia). Y Combinator, the global benchmark every Australian founder still gets compared to, has seen its acceptance rate fall to a reported 0.6% in its Summer 2025 batch, a record low, with the cohort dominated by AI (YC, 2025; YC S25 Analysis, 2025).

And for the few who do get in, the price of entry is equity. Y Combinator takes 7% on a $125K cheque plus a $375K uncapped MFN SAFE that converts at the next round (YC, 2025). Antler Australia takes 12% on AU$225,000, plus a separate program fee (Startup Daily, 2023). That stake compounds with every later round. By Series B, three months of accelerator can cost the founder several times over in ownership, decision-making power, and the cash a future sale would have put back in their pocket. It also locks in the ecosystem's wrong measure: that VC outcomes are the only score that counts.

And the AI redundancy wave hasn't even hit the application queues yet. The door narrows further with every batch.

The funding behind those doors is concentrated. Australian startups raised a record $5.4 billion in 2025, but the top 20 deals captured 58% of it. 61% of capital flowed to AI companies. Two-thirds of deals included international investors. Median pre-seed rounds were $1 million; median seed rounds, $2.5 million. In Victoria, the country's largest startup state, early-stage capital fell year on year and is now similar to 2016 levels (Cut Through Venture, 2025; LaunchVic & Dealroom, 2025). The classic founder path of raising a small round, getting into a top accelerator, then scaling, is harder than it has been in years.

And the failure rate hasn't moved. For solo founders, and most of the new wave will start there, almost 6 in 10 new businesses fail within their first three years (ABS, 2024-25). Worldwide, about 9 in 10 startups eventually fail (Startup Genome, 2025). The reasons why have been the same for a decade. CB Insights' 2026 analysis of failed venture-backed companies finds 70% ran out of capital, 43% missed product-market fit, 29% were killed by bad timing, and 19% by unsustainable unit economics (CB Insights, 2026). The rest comes down to cofounder fights, scaling too fast, and gaps in how the business is actually run (Wasserman, 2012).

None of this is bad luck. These are patterns. They are predictable. They are preventable. The frameworks that prevent them have been written down for decades. Yet the rate hasn't moved.

Why Doing More Won't Fix It

Australia's response to startup failure has been the same for a decade. Build more programs. Stand up another accelerator. Open another incubator. Launch another grant. Cut another ribbon on another hub. Today there are more than 480 Australian programs, grants and networks for early-stage founders.

The failure rate hasn't moved.

Here's why. Most programs don't have a diagnostic at all, just an intake form. Most accelerators run a single curriculum, in a single order, regardless of where each founder actually is. Wharton's analysis of recent accelerator research (Cohen, Hallen and Bingham, HBR 2024) puts it plainly: "there isn't a one-size-fits-all approach when it comes to accelerator programs" (Knowledge at Wharton, 2024). And yet that's exactly what most cohorts deliver, which is why the criticism every accelerator alumni Slack carries is the same: it was cookie-cutter. Workshops cover one domain, sometimes two, almost never the full picture. Mentors are matched by what's available, not by what the founder needs. Support depends on the cohort, the day, the schedule. Every consultant, in turn, fixes one functional area while the other eight quietly fall apart. A founder asks a book one thing, gets a different answer from a mentor, gets a third from an advisor, and a fourth from the next podcast. Nobody shows them how it all connects. Nobody tells them that a gap in Governance cascades into Finance, which cascades into Strategy, which cascades into Marketing.

Adding more programs to that doesn't reduce the noise. It is the noise. The fragmentation isn't a side effect of the system. It is the failure mode. A wave of new founders walking into more of the same will give the same result. Or worse.

The Answer Was Always There

Here's the bit of good news hiding in the failure numbers. Startups don't fail in mysterious ways. They fail in patterns. And patterns can be taught.

The frameworks that prevent these patterns have existed for decades. Big organisations have been using them since the 1980s. Hospitals codified theirs into protocols. Aviation codified theirs into checklists. Engineering codified theirs into standards. Baldrige (1987). EFQM (1988). ISO 9001. McKinsey 7S. The Business Model Canvas. None of it is new. All of it is published, peer-reviewed, and proven.

What hasn't existed, until now, is a way to put all of that into the hands of one founder, on one screen, in plain language, calibrated to where they actually are.

That gap (between what we already know about why startups fail, and what a single founder can actually use) is what Sova was built to close.

The Layer Underneath

Sova isn't another accelerator. It isn't another VC. It isn't another incubator. It's the layer underneath all of them. The starting point every founder should have before they spend a single dollar on anything else.

And it is built around the constraints the new wave actually has.

• Available immediately. No application. No acceptance gauntlet. No equity. No waiting for the next cohort. The nine-element check is free. Paid plans start at $99 for three months. Fifteen minutes from start to first direction. For a founder whose runway just started ticking, that isn't a feature. It's the only viable on-ramp.
• Holistic. All nine parts of the business at once: Governance, Purpose, Strategy, Performance, Finance, Marketing, People, Process, Technology. Most accelerators teach one or two. Most consultants fix one while the other eight quietly fail. Sova maps the lot, and shows how a gap in one cascades into the others. The fragmentation that drives the failure rate stops at the door.
• Expert. Built on more than fifteen years of transformation work inside Australian hospitals, law firms and professional services. The kind of structural diagnostic that usually lives inside firms billing six figures, now in the hands of one founder for less than the price of a single consulting hour.
• Evidence-based. More than 350 research findings from 80+ published sources: Harvard Business Review, McKinsey, CB Insights, Startup Genome, the Global Entrepreneurship Monitor. Every recommendation cited. Not an opinion. Not a vibe. Not a hot take from the same podcast circuit founders are already drowning in.
• Best-practice. Grounded in the same frameworks Fortune 500 organisations and governments have used for decades. Baldrige (1987). EFQM (1988). ISO 9001. McKinsey 7S. None of it new. All of it proven. What hasn't existed, until now, is a translation layer between those frameworks and a one-person business.
• Personalised. Tells the founder exactly where their business is exposed, what to fix first, why it matters, and which of the 480+ Australian programs is the right next step for them. Not generic. Not one-size-fits-all. The fragmentation problem solved at the founder's level, instead of by adding another program to the pile.
• Always on. An integrated AI coach that already knows their results, their stage, their industry. Available 24/7. Answers in plain language. Not a generic chatbot guessing from public training data. A coach with the founder's actual context.

Take what usually costs tens of thousands of dollars in consulting hours, and put it in the hands of the founder whose runway just started ticking. Who can't afford to spend three months hunting for the right consultant. Who can't afford to apply to four accelerators and wait six months for the next intake. Who can't afford a wrong turn this quarter, let alone three of them.

The Choice

So we have two paths.

Keep doing what we've always done. Add another accelerator. Open another hub. Stand up another grant. Watch the failure rate stay where it has been for the last decade. Watch a wave of new founders, with mortgages and limited runway and no time to spare, fail in the same patterns the founders before them did.

Or finally give every Australian founder the same starting point. The same playbook. The same evidence base. The same way to find the right help. The same way to know if it worked. Built on the best of what Australia already has, organised properly, in their hands, on day one.

That doesn't need more capital. It doesn't need more programs. It needs the resources we already have, finally working together, in every founder's hands.

That is what Sova is. Better businesses build a better world. Every founder deserves the tools to build one.