Outcome-Based SaaS Is Still SaaS

In early April 2026, HubSpot announced that two of its Breeze AI agents — the Customer Agent and the Prospecting Agent — would move to outcome-based billing. The Customer Agent now costs $0.50 per resolved conversation. The Prospecting Agent costs $1.00 per qualified lead surfaced. The seat fee is gone. You pay when the agent delivers.

The framing from HubSpot is progressive: align spend with value, reduce the barrier to deployment, tie the invoice to the result. That is a reasonable argument. It is also a retreat.

The seat licence model was always a proxy for something else. It was never billing for the software — it was billing for the coordination infrastructure underneath the software: the access management, the approval workflows, the dashboard reviews, the human handoffs that required named users to be logged in and operational. Per-user pricing was a tax on headcount. You paid for seats because you had people, and the people needed the tool, and the tool was priced to match the organisational structure that produced the demand for it. That argument is made precisely in Memo #14: The Death of the Seat Licence.

When agents enter that structure, the seat model begins to fail. An agent does not log in. It does not navigate a dashboard. It does not consume the interface layer the seat fee was built to support. The UI Tax — the cost premium embedded in SaaS products designed for human use, including the interface development, the permissions infrastructure, and the UX premium baked into subscription pricing — has no surface to attach to. HubSpot’s move to outcome-based pricing is an acknowledgement of exactly this. The vendor is responding to the same pressure the memo described: the seat model is structurally incompatible with agentic workflows.

But the position worth taking is not that HubSpot has solved the problem. It is that outcome-based SaaS is still SaaS.

Paying $0.50 per resolved conversation is a better deal than a flat seat fee. It is not the same as owning the resolution logic. An autonomous business that has built its own support infrastructure — a deterministic loop connected directly to its data layer — does not receive an invoice when a conversation closes. The compute cost of that resolution is a fraction of $0.50, it scales with actual volume, and no vendor repricing decision changes the unit economics next quarter. The logic is owned. The outcome is not billed. This is what Memo #03: Overhead Is a Design Choice makes precise: overhead is not reduced by repricing the vendor relationship. It is eliminated by removing the vendor from the architecture.

Outcome-based pricing still embeds a vendor’s margin, a vendor’s definition of “resolved,” and a vendor’s determination of what qualifies as a lead. You are not paying for a seat anymore. You are paying for someone else’s judgment about whether the output counts. The Coordination Tax — the cost of human-to-human alignment that Arco’s architecture eliminates — has a vendor-side equivalent: the cost of delegating the definition of success to a third party. Outcome-based SaaS does not eliminate it. It repackages it.

HubSpot reports that its Customer Agent resolves approximately 65% of conversations. The 35% it does not resolve are returned to a human. Under outcome-based pricing, those unconverted interactions carry no charge — but the 65% that do resolve compound at scale. At volume, the arithmetic still favours ownership over subscription. The billing model has become more honest. The underlying dynamic has not changed.

There is a reason incumbents like HubSpot can reprice but cannot rebuild: the structures that built them are the same structures preventing them from changing. That is Legacy Liability in precise terms. Outcome-based pricing is a billing layer adjustment. It does not alter the fact that HubSpot’s agents run on HubSpot’s infrastructure, governed by HubSpot’s definitions, logged in HubSpot’s CRM. The vendor relationship remains. The dependency remains. The Operational Drag generated by third-party definitions and third-party infrastructure is not removed by changing when the invoice arrives.

For a business that has already De-SaaS-ed — that pays for compute, not for coordination — HubSpot’s announcement is not a pricing innovation to evaluate. It is a market signal confirming what the architecture already assumed: the seat licence era is ending, and the vendors who built on it are working out how to survive what comes next.

The seat licence is dying. Outcome-based SaaS is its successor. Owning the logic is the exit from both.