Mini Series: A once in a generation opportunity: Saas-pocolypse
Right now, the market doesn’t know who the winners will be from AI… so it’s treating everyone like a loser.
This has been a classic shoot-first, ask-questions-later sell-off. It started in October and accelerated sharply in January and February after Anthropic previewed its Claude agent.
This wasn’t just another chatbot. It demonstrated autonomous AI agents capable of executing complex workflows with minimal human input — and that triggered a major shift in thinking.
If one AI agent can replace five to ten employees, what happens to the traditional SaaS model? Historically, software growth has been tied to headcount. Licenses were sold per employee, and valuations assumed hiring would continue.
So, if AI replaces people, that means:
Fewer seats
Lower revenue base
Compressed valuations
And that’s exactly why multiples have dropped so aggressively.
Even in Australia, the Aus Tech Index is down over 40% from its highs. In the U.S., the iShares Expanded Tech-Software ETF (IGV) is down around 30% from its peak.
Valuations have collapsed. Forward P/E is now roughly 22x, versus a long-term average closer to 33x — a two-standard deviation drop for companies still delivering strong earnings growth.
This is why people are calling it a SaaS-pocalypse.
But the Selling is Overdone
To understand what’s really happening, we need to understand agentic AI. Think of it as a cognitive layer added on top of existing software. It doesn’t replace software — it enhances it:
Summarises information
Generates insights
Executes workflows
Yet markets are pricing software as if AI is about to dismantle the entire sector. That disconnect is where the opportunity lies.
Even Jensen Huang has pushed back on this idea, calling the notion that AI will replace software “illogical.” AI is just another tool — and tools sit on top of software, they don’t replace it.
Anthropic itself is hiring to integrate its models into platforms like Salesforce — proof that AI is being built into ecosystems, not replacing them.
The Real Winners: Incumbents
The winners here aren’t the disruptors everyone expects — they’re the incumbents. Platforms that already control:
Data
Workflows
Customer relationships
Think Salesforce, Oracle, Microsoft, and ServiceNow.
Global enterprises aren’t moving their most valuable data to unproven vendors. They won’t take on new security risks, compliance headaches, or massive switching costs. The path of least resistance? Adopt AI inside platforms they already trust.
These platforms have deeply embedded workflows, massive data gravity, and strict security frameworks — making them sticky and perfectly positioned to monetize AI.
Monetizing AI at Scale
We’re already seeing results:
Salesforce Agentforce growing ~200% YoY, heading toward $1B in ARR
ServiceNow maintaining near 100% renewal rates while rolling out AI across its platform
Pricing models are evolving, too. We’re shifting from pure seat-based pricing to hybrid models: subscriptions plus usage-based pricing per AI task or automated workflow.
Even if AI replaces a human, it doesn’t operate for free. Compute and infrastructure costs get passed through, making usage-based models potentially more profitable at scale.
The Bottom Line
Software is being priced for disruption — but the reality is very different. AI isn’t replacing SaaS; it’s enhancing it.
With valuations at multi-year lows and early signs of capital flowing back into software, this could be one of the most mispriced opportunities in the market today.
Drop your thoughts in the comments below.