For the best part of two decades, the software-as-a-service model has been the closest thing venture capital had to a sure bet. Build a product once, sell it as a subscription, watch the recurring revenue compound, and keep customers locked in because switching was more trouble than it was worth. That predictability underwrote thousands of startups and the funds that backed them. It is now, according to one of Australia’s most experienced technology investors, coming apart.
In a pointed opinion piece for Forbes Australia, veteran investor Daniel Petre argues that artificial intelligence is tearing up the SaaS rule book and, in the process, reshaping how venture capital itself works. He calls the moment the “SaaSpocalypse”, and his central claim is uncomfortable for anyone whose business or portfolio rests on the old assumptions. The advantages that made software companies so valuable, the sticky subscriptions and the high cost of switching, are being eroded by AI tools that can replicate features in weeks rather than years.
Why the old model is under pressure
The logic is straightforward once you sit with it. Much of the value in a SaaS business came from the difficulty of building the software in the first place, and the inertia that kept customers paying long after the initial sale. Generative AI compresses both. If a capable engineering team, or increasingly an AI coding assistant, can reproduce a competitor’s core functionality quickly and cheaply, then the moat that justified a lofty valuation starts to look shallow. Customers who once tolerated a mediocre product because rebuilding it was unthinkable now have credible, cheaper alternatives appearing constantly.
For Petre, a co-founder of AirTree Ventures and a figure who has watched several technology cycles pass through the Australian market, the consequence is a bifurcation in how money gets deployed. He describes venture capital shifting towards a barbell shape. At one end sit the mega-bets, the enormous cheques written for the handful of companies building foundational AI models and the infrastructure beneath them, where the capital requirements run into the billions. At the other end sit tiny cheques, small amounts invested in lean teams that can build and ship remarkably fast with a fraction of the headcount that a comparable startup needed a decade ago. The comfortable middle, the Series A and Series B rounds that funded steady SaaS scale-ups, is where the squeeze is sharpest.
Two ways to read the shift
There is an optimistic reading of all this, and Petre’s argument leaves room for it. If small teams can build valuable products cheaply, then the barrier to founding a company falls dramatically. A founder no longer needs to raise tens of millions to assemble a large engineering team before they have something worth selling. That should, in theory, unleash a wave of new companies and give capital-efficient founders real leverage in negotiations with investors. The best businesses might reach profitability on far less money, returning more of the upside to the people who built them rather than the funds that backed them.
The pessimistic reading is harder to dismiss. If moats evaporate, then so does the durability that made software such an attractive asset class. A product that can be copied in weeks can also be undercut in weeks, which puts relentless downward pressure on prices and margins. Investors who paid SaaS-era multiples for growth may find those valuations difficult to defend. And the barbell itself carries a warning: if capital concentrates at the extremes, plenty of otherwise decent companies in the middle could find themselves unable to raise the next round, not because they are failing, but because the funding structure that once supported them has quietly moved on.
What it means for Australia
This is where the Australian stakes come into focus, and why Petre frames the piece as a warning to a local audience rather than a detached observation about Silicon Valley. Australia does not build foundational AI models at the scale of the American labs, and it is not going to start writing billion-dollar infrastructure cheques from a standing start. That leaves the local ecosystem heavily weighted towards exactly the kind of software companies most exposed to the SaaSpocalypse: applied SaaS businesses selling into enterprise and government, many of them funded on the assumption that recurring revenue would keep compounding.
A great deal of Australian venture capital has flowed into that category over the past decade, and the country has produced genuine success stories on the back of it. If the moats those companies rely on are thinner than they appeared, then some of that capital is mispriced, and some of those businesses will need to prove they can defend their position against AI-enabled newcomers. The flip side is that Australia has a deep bench of technical talent and a track record of building capital-efficient companies out of necessity, given how far local founders have traditionally sat from the largest pools of money. In a world that rewards small teams shipping fast, that constraint could turn into an advantage.
There is also a policy dimension that sits alongside the investment one. Governments across the country, from the Commonwealth down to the states, have been leaning into artificial intelligence as an economic opportunity, and much of that enthusiasm assumes a healthy pipeline of local software firms creating jobs and export revenue. If the economics of that sector are being rewritten, the assumptions baked into industry strategies and procurement decisions deserve a fresh look too.
What comes next
Petre’s piece is an opinion, not a forecast with a date attached, and reasonable people in the local industry will disagree about how far and how fast the SaaSpocalypse actually plays out. Plenty of software businesses have moats that go beyond code, in the shape of proprietary data, regulatory approvals, entrenched integrations and customer relationships that AI cannot simply clone. Those companies may weather the shift comfortably. The ones that should be nervous are those whose defensibility rested mainly on being first to build a feature.
For founders raising now, the practical takeaway is to think hard about what actually protects the business once the code becomes easy to replicate. For investors, it is to question whether the middle of the market they have historically funded still offers the returns it once did. And for a country that has bet a good deal of its technology ambition on the SaaS model, the argument is worth taking seriously precisely because it is inconvenient. The old rule book was reliable right up until it wasn’t, and the moment to notice is before the next round, not after it.
Sources: Forbes Australia
