Town's Pooled Credit Pricing Model

Town is pricing around the real cost driver, which is machine work, not seats. Every routine can trigger model calls, searches, app actions, and follow ups, so pooled credits let Town charge more when a workflow runs more often or gets more complex. That is a cleaner fit for an assistant that actually does work than fixed unlimited plans, which become less profitable as autonomous usage rises.

• Town already sells this way in practice. Its pricing page highlights pooled credits on team plans and lower per credit cost at higher tiers, while its docs show the assistant working across messages, search, documents, and team Squares, all of which can add variable inference and tool costs per job.
  1 town 2 town 3 town
• The packaging also matches how work is distributed inside companies. The Business tier uses a large shared credit bucket instead of forcing every occasional user onto a full seat, which fits organizations where many employees submit requests but a smaller group runs the heaviest routines.
  1 town 6 sacra
• This is the opposite of products that bundle broad AI usage into fixed subscriptions. Notion moved core AI into higher tier plans with unlimited usage, while Zapier emphasizes pay per task billing and pay as you go options. Town sits closer to the Zapier logic, charging against execution volume rather than assuming average usage stays tame.
  4 zapier 5 zapier 7 sacra

As agent products move from drafting to repeated execution, pricing is likely to converge on credits, tasks, or other metered units tied to actual machine work. That favors Town if it can keep credits easy to understand while expanding routines deeper into company workflows, because more automated work will then expand revenue instead of only expanding compute expense.