The Death of the Seat: Why Anthropic’s Growth is a Warning for B2B SaaS

The End of the Per-Seat Era

The numbers coming out of Anthropic are hard to process. Reaching a $47 billion run-rate just three years after earning its first dollar is not just a success story—it is a structural anomaly in the history of software. To put that in perspective, Anthropic has effectively surpassed the revenue run-rate of Salesforce, the company that defined the modern SaaS category. While it is easy to dismiss this as "AI hype" or a massive spend on compute, the deeper story is about how software is being sold and consumed.

For two decades, the "seat-based" model was the gold standard for B2B SaaS. You built a tool, you counted the heads using it, and you sent a monthly bill. But Anthropic isn't selling seats; it’s selling tokens. It is selling work. This shift from charging for access to charging for output is perhaps the most significant change for SaaS founders since the transition from on-premise to the cloud. When the fastest-scaling software company in history doesn't sell a single seat, the rest of the industry needs to pay attention.

Monetizing Work, Not Logins

The fundamental problem with seat-based pricing in an AI-driven world is that it rewards inefficiency. If a product management tool charges per user, its revenue is capped by the headcount of the customer. However, as AI agents begin to take over workflows, the number of "users" in a company may actually decrease while the volume of work increases. If your software makes a team five times more productive, but that team shrinks from ten people to two, a seat-based model would see your revenue plummet exactly when you are providing the most value.

Anthropic’s growth is driven by consumption. Developers using tools like Claude Code are running agents against entire codebases, consuming tokens at a rate that no human-typing-in-a-browser could ever match. In several internal tools we've built at Solviba, we've seen that shifting to a usage-based architecture allows for a much tighter alignment between the value provided and the revenue captured. When the software is doing the "heavy lifting," the customer is much more willing to pay for the outcome than for the license to log in.

The Efficiency Gap and the "Workflow" Trap

Many incumbent SaaS companies are currently trapped. Their entire valuation is based on the number of "knowledge workers" they can seat. As AI automates the very tasks these platforms were built to manage—data entry, status updates, basic reporting—the value of the "interface" drops. If a customer can get the same result by pinging an API rather than logging into a dashboard, the traditional workflow tool becomes a commodity.

When we consult with early-stage SaaS teams at Solviba, we often find that the biggest hurdle to adopting AI features isn't the technology—it’s the business model. Founders are afraid that by making their users too efficient, they will cannibalize their own seat-based revenue. Anthropic's trajectory proves that the market for output is significantly larger than the market for seats. The goal shouldn't be to keep users in your app longer; it should be to process as much of their work as possible, as fast as possible.

How Product Teams Should Pivot

If you are building a digital product today, you need to ask whether your pricing model survives a world where the "user" is often an agent. This doesn't mean every startup needs to switch to pure consumption-based pricing tomorrow—that can be a nightmare for financial predictability. However, it does mean that your roadmap should prioritize "high-leverage" features that produce measurable output rather than just better "collaboration" tools.

One approach we often recommend at Solviba is the hybrid model: a base subscription for the platform (the "system of record") combined with a usage-based kicker for AI-driven actions. This protects your baseline revenue while allowing you to capture the massive upside of agentic workflows. As Anthropic continues to outpace the growth of the giants, the companies that thrive will be those that realize software is no longer a place where people work—it is the thing that does the work for them.

If you're exploring how to transition your product to a consumption-based model or trying to figure out how to build agentic features that actually drive revenue, the Solviba team often helps startups think through these decisions and build the first versions of their systems. Feel free to reach out if you'd like to discuss your project.