← back
Notion's Token Town — Sarah Sachs, Notion
Takeaway
Sustainable AI products need deliberate model economics and durable product value beyond reselling tokens.
Summary
- Notion frames AI adoption as progressing from thought partner to assistant, teammate, and interconnected business systems.
- A durable shared system of record is presented as essential for collaboration among humans and agents.
- Model upgrades can raise actual costs through longer outputs or higher prices even when feature improvements appear attractive.
- Sachs advocates provider flexibility and differentiation through customer understanding, product interfaces, orchestration, and task-specific capability-cost-latency choices.
token-economicsnotionmodel-selection
Original description
Every month it is the same trap. A reasoning model gets upgraded at the same price per token, then quietly burns three times the output tokens. Or the new version costs 40% more and deprecates its predecessor in four months. Are you growing 40%? Making three times the revenue? Sarah Sachs, who leads AI engineering at Notion and negotiates its model contracts, says no, and that is why she treats every vendor as a competitor. Buy tokens from a lab that also sells the first party product and you are paying a markup on a markup for something you cannot defend, with no exit if you lock yourself to one provider. Her answer is to stop winning on token economics and win on product: data flywheels, orchestration, and staying model agnostic so optionality stays your leverage. Notion's auto model quietly routes about 75% of traffic and swaps providers underneath, her AI Switzerland approach; triaging an email inbox on Opus, she says, rips off the customer and Notion both. Route by cost per capability per second, not per token; use open weight models for the moderate middle; reach for CPUs over GPUs, because you do not need an LLM to turn a CSV into a PDF. The final stretch turns to the lethal trifecta and a live demo of Notion agents scoping a task, tagging in teammates, and opening a PR. Speaker info: https://x.com/sarahmsachs / sarahmsachs Timestamps: 0:00 - Welcome to Token Town: AI poor vs AI rich 1:25 - Negotiating AI contracts as Notion's AI Anna Wintour 2:30 - The AI transformation journey and the system of record 4:46 - Why cost is the structural barrier 5:34 - The monthly model pricing traps 7:28 - Your supplier is your competitor 8:44 - Win on product, not token economics 9:50 - Not all traffic belongs on the frontier model 12:10 - Optionality is your leverage 13:25 - The auto model and the model agnostic playbook 15:04 - Open weight models for the middle 17:00 - CPUs over GPUs and governance 17:53 - Security: the lethal trifecta 19:46 - Live demo: orchestrating Notion agents into a PR