The Seat Did Not Die. It Got a Meter on Top

Published

The Seat Did Not Die. It Got a Meter on Top

If you follow enterprise software, you may have seen a familiar story play out recently: AI agents are replacing employees, employees are being replaced by seats, and seats are becoming obsolete.

That story is compelling. It is also incomplete.

Between January and February 2026, the market appeared to reach a verdict on per-seat software pricing. The trigger was Anthropic’s release of Claude Cowork, an agent capable of reading and writing files on users’ desktops. Anthropic followed with eleven open-source plugins covering legal, sales, finance, marketing, and customer support.

The market reacted quickly. Jefferies analyst Jeffrey Favuzza coined the term “SaaSpocalypse,” leading to downgrades for companies such as Workday and DocuSign. The S&P Software & Services Index lost 25% of its value between January 12 and February 23. In one particularly sharp session, the index fell 4%, wiping out more than $285 billion in market capitalization. JPMorgan’s software index dropped 7%, LegalZoom fell nearly 20%, and Thomson Reuters dropped 16%.

The argument spreading across the industry was straightforward:

One human is replaced by an AI agent.
That agent replaces several SaaS licenses.
Those licenses represent recurring revenue.
Therefore, per-seat pricing is dead.

Vendors began experimenting with alternatives based on outcomes: one price per resolved ticket, conversation, action, or completed task. Outcome pricing was quickly declared the successor to the software seat.

But when you examine the actual pricing pages, a different picture emerges. The seat has not disappeared. It has acquired a meter.

There is still a base fee underneath the new pricing models, often tied to users, and an outcome-based charge layered on top. The seller also gets to define what counts as the outcome. That makes the model more flexible than it is revolutionary.

The seat is not dead. In enterprise AI, it remains one of the most important pricing mechanisms because it is still how much of the recurring revenue arrives.

The supposed successor is the smallest category

The narrative and the data have come apart.

According to Bain’s analysis of AI pricing, only about 10% of vendors introducing AI pricing rely primarily on outcome-based meters.

A much larger group prices based on effort, such as tokens, compute consumption, or agent hours. Another group charges for outputs, including summaries produced, records updated, or leads recommended.

More importantly, roughly one in five AI-native software companies still relies mostly on per-seat licensing, usually with usage entitlements layered on top.

There is another important detail: approximately four in every five vendors introducing AI pricing have adopted capacity-based models. These include prepaid credit pools and committed-spend plans, often without rollover.

Maxio also reviewed 107 vendor pricing pages. Pure outcome pricing appeared on fewer than one in five.

Even companies selling agent products have not abandoned the seat. Harvey, Gong, and Fathom are all marketed as agent products, yet each retains a per-user, per-month floor with usage or outcome charges layered on top.

That means the central axis of the argument was wrong.

Outcome pricing did not win. Capacity pricing did.

Prepaid credit systems such as Workday’s Flex Credits, Salesforce’s Flex Credits, and monday.com’s credit bundles became the more practical successor. They give customers a sense of predictability while allowing vendors to price consumption as it happens.

The seat did not vanish beneath this model. It became the foundation beneath a new metering layer.

This was first published on Sol AI — https://thesolai.github.io