Notion
An $8 AI add-on worth about a dollar a seat — withdrawn, and replaced with a $20 tier boundary worth an estimated $120–165M.
Published 2026-09-19
Notion sold AI three ways in three years and only one of them made money. The $8 add-on had to be bought for every seat in a workspace, including the seats that would never use it — so it under-attached, capped revenue per seat no matter how hard a team used it, and left the most expensive capability in the product sitting on the free tier. Withdrawing it in May 2025 and folding AI into the $20 Business plan changed the question from one a CFO can decline — is AI worth $8 a seat — into one only the team can answer about itself. The headline 33% increase was the smaller half. The migration, at +150% per converted seat, was the event.
Force-ranked
- 1The re-fence made the money. The headline increase was the smaller half of it.
- 2An add-on billed across every seat under-attaches by construction — it taxes the people who will never open it.
- 3A 33% rise with no visible churn is not vindication. It is evidence the rise was too small.
- Outside-in only — public price lists, release notes and reported figures. No first-party data, no mandate.
- The revenue decomposition is a model, not a disclosure. Notion publishes ARR milestones and one attach statistic; the seat counts, mix and revenue per seat behind our number are our own estimates.
- No claim about Notion's margin, cost of inference, or internal reasoning.
- The date Plus moved from $8 to $10 is not firmly sourced. That it moved is certain; when is not.
This is a public-source analysis, published so you can judge the method before you buy it.
We'd rather do this on your pricing than Notion's.
Same method, your own data behind it, and a proposal you approve or don't.