SaaS Isn’t Dead. Old SaaS Is Dead: Sierra Adds $150M ARR in One Quarter While ServiceNow Drops 27% in a Day

We have been hearing that SaaS is dead for nearly two years. The markets are flat, the growth is hard to find, and the multiples have shrunk. But the reality isn't that software as a service is dying; it is that the old way of building and selling it is becoming obsolete. We just saw two massive signals that confirm the industry is splitting into two distinct eras.

The Tale of Two Quarters

In a single 24-hour window, the market sent a message that should keep every founder awake at night. ServiceNow, a titan of the enterprise workflow space, saw its stock price crater by nearly 27% in a single day following its quarterly report. This wasn't a general market dip—the S&P 500 barely moved. This was a targeted rejection of the legacy model.

Meanwhile, on the other side of the fence, Sierra—the AI agent startup co-founded by Bret Taylor—reportedly added $150 million in annual recurring revenue (ARR) in just one quarter. To put that in perspective, most elite SaaS companies take a decade to reach that number. Sierra did it in ninety days. One company is a legacy giant struggling to justify its weight, while the other is an AI-native rocket ship that is consuming the market's remaining oxygen.

The Efficiency Trap

For a long time, the promise of SaaS was efficiency. You bought a tool to help your employees work faster. You paid per seat, and as your company grew, the software company’s revenue grew. This created a symbiotic relationship where more headcount equaled more software spend. For builders, this was the golden goose. You didn't have to prove the software did the work; you just had to prove it made the person doing the work slightly more organized.

AI-native companies like Sierra aren't selling seats. They are selling outcomes. They aren't providing a dashboard for a customer support agent to use; they are providing the agent. When you move from selling tools to selling work, the revenue scales at a rate that traditional SaaS cannot match. This is why ServiceNow is feeling the heat. If an AI agent can handle 80% of the workflows that used to require a complex enterprise suite, the value of that suite drops toward zero.

What This Means for Builders

If you are building in the crypto or AI space right now, you have to realize that the "workflow tool" category is a graveyard. If your pitch deck includes the words 'streamline,' 'optimize,' or 'manage,' you are fighting for scraps from the legacy era. The new era is about execution. The market is no longer interested in paying for a better hammer; it wants a house that builds itself.

The Decline of the Seat-Based Model

The 27% drop in ServiceNow is a warning shot for anyone still relying on seat-based pricing. In an AI-first world, headcount is a cost to be minimized, not a metric for software licensing. As companies automate more of their internal processes, the number of 'seats' they need will naturally decline. If your business model depends on your customers hiring more people, you are bettting against the primary trend of the next decade.

• Focus on outcomes: Don't measure your success by daily active users; measure it by the volume of tasks your system completes autonomously.
• Ignore the noise: The 'SaaS is dead' crowd is half right. The delivery model is fine, but the value proposition has shifted from enablement to achievement.
• Build for the lean enterprise: The winners will be tools that allow a 10-person team to do the work of a 1,000-person company.

The Crypto Connection

There is a lesson here for the Web3 crowd as well. We spent years building infrastructure that was 'faster' or 'more secure,' but we forgot to build things that actually do the work. The reason AI is currently eating the world while crypto languishes in 'infrastructure phase' is because AI provides immediate utility. Sierra's $150M quarter happened because they solved a problem that companies were willing to pay for today. They didn't ask users to wait for a roadmap or a token launch; they just fixed the customer support queue.

The Founder Perspective

I’ve talked to enough founders to know that many are still trying to 'AI-ify' their existing products. They are adding a chatbot to a sidebar or a summary feature to a document editor. This is the equivalent of putting a motor on a horse-drawn carriage. It’s better, sure, but it’s not a car.

Sierra is the car. They didn't start with a legacy codebase and try to sprinkle some LLM magic on top. They started with the premise that the work could be handled by the machine from day one. That is why they can add $150M in ARR while the giants are shrinking. They aren't competing for the same budget; they are creating an entirely new category of spend.

The shift we are seeing is not cyclical; it is structural. We are moving from a world where software helps people work to a world where software is the work.

The Takeaway

The lesson from this week is simple: the gap between 'Old SaaS' and 'New AI' is widening into a chasm. ServiceNow’s bad day isn't a fluke; it's a symptom of a market that is tired of paying for complexity. Sierra’s massive growth isn't a bubble; it's a demonstration of what happens when you solve a high-value problem with 100% automation.

As a builder, you have to choose which side of the chasm you want to be on. You can try to defend the old model of seat-based workflow, or you can build the autonomous agents that make those seats unnecessary. The market has already decided which one it values more. Don't be the person building a better dashboard when the world is looking for a driver.

Read the original at SaaStr →