A founder recently replaced his entire customer service team with an AI coding agent. Not downsized. Not restructured. Replaced. The tool could write and deploy software on its own, and suddenly companies like Salesforce weren't the automatic default anymore.
That story, reported by TechCrunch this week, is one of dozens circulating through boardrooms and investor group chats right now. Software stocks shed nearly $1 trillion in value in early February alone. Analysts coined a new acronym (FOBO, fear of becoming obsolete) and the press landed on a catchier label: the SaaS-pocalypse.
It makes for great copy. But here's the thing: the diagnosis is wrong.
SaaS isn't dying. It's moulting. And if you understand what's underneath the old skin, the opportunity is massive.
The panic is real. The conclusion isn't.
Here's what spooked the market: AI agents got good. Really good. Good enough that a startup can now spin up in days what used to take a funded team months to build. The barriers to creating software have dropped so far, so fast, that the whole "build vs. buy" equation, the foundation the SaaS industry was built on, is tilting hard toward build.
That matters. But the hot takes leapt straight from "AI can replicate some of what SaaS does" to "SaaS is dead." And that's where the narrative goes off the rails.
AI isn't killing software. It's reframing where the value lives.
The old model had it coming
Let's be honest about what the traditional SaaS playbook looked like for most customers. Predictable recurring revenue, great for investors. Per-user pricing, great for sales teams. Feature bloat that justified more seats, more tiers, more line items on your invoice, great for nobody except the vendor's Q4 numbers.
Customers tolerated it because there was no alternative. Now there is.
When employees can simply ask their AI of choice to pull data from a system, or orchestrate outcomes across a dozen tools, the obvious question lands: why am I paying for all these apps individually? Suddenly the value isn't in the UI or the feature list. It's in the intelligence that drives decisions and gets work done.
That shift was overdue. The SaaS-pocalypse isn't a catastrophe. It's a correction.
Build vs. buy just flipped, and it changes everything
I remember the first time an AI tool generated a production-ready app for me, faster than a junior engineer could have coded it by hand. My first thought wasn't "this will replace developers." It was more like: "every buyer in every meeting is about to ask the same question: why are we paying a vendor for this when we could build it ourselves in a week?"
That's the build-vs-buy shift, and it's the part the SaaS-pocalypse headlines consistently underplay. It's not just that AI can do what SaaS products do. It's that customers now have the ultimate negotiation leverage: if they don't like a vendor's price, they can credibly threaten to build their own alternative. Klarna did exactly this in late 2024, ditching Salesforce's flagship CRM in favour of a homegrown AI system. What seemed like an outlier move then looks more like a preview of where the whole market is heading. Even when companies don't take the build route, the possibility creates downward pressure on every renewal contract in the pipeline.
For investors, this is existential. Public markets price SaaS companies by estimating future revenue. But when there's no telling whether customers will still be paying for those products in one year or five, the maths breaks. That's why every time a new AI capability launches; software stocks feel a tremor. It may be the first time in history that the terminal value of software is being fundamentally questioned.
It's reshaping what customers demand. Per seat, per year is giving way to usage-based billing that scales with actual consumption, outcome-based contracts where vendors share the risk, and products that embed into an AI-orchestrated stack rather than demanding their own login and learning curve.
In plain English: people don't want to pay for software anymore. They want to pay for results. And the companies that can't make that shift may find their customers building the alternative themselves.
Where SaaS survives, and where it doesn't
Not all SaaS is equally exposed. The venture investors I've been listening to are drawing some clear lines.
Horizontal point solutions are under real pressure
The tools that do one thing, one way, and charge per seat. These are the easiest for AI to replicate or displace entirely. If your product is a thin workflow layer, generic analytics, or surface-level project management, investors are already looking past you. And they're right to.
Vertical and domain-specific software still has moats
Healthcare, logistics, manufacturing, financial services: areas where complex regulations, specialised workflows, and decades of embedded domain knowledge create real defensibility. A general-purpose AI isn't replacing that overnight. But disruption is coming here too, just more slowly and more selectively.
AI-first platforms are the new frontier
This is the part that matters most. A wave of AI-native startups is rising at record pace, having completely redefined what it means to be a software company. The SaaS companies that thrive in the next decade won't be the ones that bolted a chatbot onto their sidebar. They'll be the ones that rearchitected around intelligence from the ground up, where the AI isn't a feature, it's the foundation.
Here's the counterpoint the doom-and-gloom crowd misses: traditional software firms that are actively integrating AI aren't just surviving, they're thriving. AI augments strong products and enables entirely new ones. It doesn't automatically replace the established platforms. But it absolutely punishes any coasting on lock-in and inertia.
What others frame as collapse, I see as evolution. The future winners will own unique data moats, price around outcomes instead of seats, and blend automation with human judgment in ways that neither can achieve alone.
So, is SaaS dead?
No. But SaaS as we knew it, the per-seat, per-feature era, is being forced to shed the parts that never delivered real business value and reinvent itself around intelligence and outcomes.
This is exactly what Ageiro is building for
At Ageiro, we're not reacting to this shift. We're architecting for it.
We believe humanity's last app isn't another per-seat tool bolted onto your existing stack. It's an AI-orchestrated layer that sits across your workflows, connecting the tools, data, and domain logic that already exist in your business, and letting humans and machines solve real problems together through the provision of software.
In practice, that means outcome-based value from day one. It means deep domain intelligence that a generic AI model can't replicate though Ageiro.ARK. And it means a platform designed so that the more your business evolves, the smarter it gets, not another subscription you're locked into whether it delivers or not.
The companies that win this next era will be the ones that stop selling seats and start delivering results. That's the bet we're making.
The future of software is about better outcomes.
Darren Edwards
Chief Product & Operations Officer, Ageiro
Darren leads product strategy and operations at Ageiro, bringing deep expertise in enterprise software, AI-driven platforms, and the future of autonomous business systems. He is passionate about building technology that delivers real outcomes, not just features.