SaaS Is Dead?

"SaaS is dead" spreads because it is half right. Software is not dying. The way we have been paying for it might be. Here is the argument going around, where it holds, where it breaks, and how I read it.

The argument going around

I have been working through a framework called Smart SaaS, and it stuck with me. The short version: the next software companies will not sell seats. They will run a three-layer play.

1. The free wedge. The commoditized layer - the CRM, the dashboard, the booking system - given away at no charge. AI can generate a working version of a $15-50 per seat tool fast, so the interface stops being the product.
2. The moat. The free tool is designed to capture workflow data and build network effects, so every new user makes the system more useful for the people already in it.
3. The monetization layer. Charge for what AI agents actually do - the outreach sent, the books reconciled, the workflow executed - and wrap premium human service around it for the top tier.

What holds up

• The interface really is cheap now. If a product is a thin UI over a database, someone can rebuild it over a weekend. Defending $30 a seat for that gets harder every quarter.
• Outcome pricing lines up with value. Paying when the agent books the meeting is easier to justify than paying for seats nobody opened.
• Data and workflow are the real asset. The tool is replaceable. The record of how a business runs is not.

Where I push back

• Free is not free. Servers, support, and trust all cost money. A free wedge with no path to the third layer is a bonfire, not a strategy.
• Captured data is not automatically a moat. Data you cannot use is a storage bill.
• Seats survive where the product is the system of record. Nobody rips out payroll or accounting because a dashboard got cheaper to build. The boring essentials keep earning their subscription - nobody brags about them, and everyone notices the day they lapse.

How I think about it

Build companies. Not complexity.

That line sits on my wallpaper, and it is the filter I read this framework through. A new tool or a new business has to do one of two things: generate cash, or remove me from routine work. Ideally both.

The three-layer model interests me because it points the same direction - software that does the work instead of software that tracks the work. An agent you pay per outcome is a worker. A dashboard you pay per seat is a chore with a login.

More AI tools and more planning are not the goal. Outcomes are. The measure of any system I build is whether it runs without me watching it, because the point was never the automation. The point is the time it buys back.

The better question

"Is SaaS dead?" is the wrong question. Which layer are you standing on?

If you sell software: what happens to your business the day your interface is free? If you buy software: fund the tools that do the work before the tools that organize the work.

SaaS is not dead. Lazy SaaS is - the kind that charged rent for a login and called it value. What replaces it looks less like a subscription and more like a worker.