
Map your AI cost per user interaction
You'll have a spreadsheet or simple model showing your AI's cost to serve one customer action, plus the threshold where free usage stops making business sense.
What you'll make
AI Unit Economics Calculator
- Cost Structure: your compute and overhead costs per interaction
- Usage Thresholds: where free becomes unsustainable
- Conversion Triggers: which behaviors predict payment
- Pricing Framework: your sustainable freemium boundaries
Write out your calculations and decision rules as a working spreadsheet description with actual numbers
The 3 parts
Each part ends with a check you can see.
Calculate True Serving Cost
Build a realistic model of what each user interaction actually costs you to serve
Done when
You have a cost-per-interaction figure that includes compute, infrastructure, and overhead — not just the API bill
Identify Sustainability Thresholds
Find the usage levels where free users become unprofitable and need to convert or churn
Done when
You have specific usage thresholds where the math breaks — queries per day, tokens per month, or whatever your key metric is
Build Sustainable Freemium Framework
Design freemium boundaries that gate usage intensity rather than core features
Done when
You have a specific freemium structure with clear boundaries that align user value with your unit economics
Capstone: you put together your AI Unit Economics Calculator.
Where this helps
An elective: a single lesson that sits outside rtwo's job certificates and doesn't count toward one.
About this lesson
What you'll be able to do
- Calculate your actual cost to serve one user interaction including compute overhead
- Identify the usage threshold where your free tier becomes unsustainable
- Have a tested pricing model that balances user acquisition with unit economics
- Know which user behaviors predict conversion vs which burn cash without paying
Who it's for
- AI product managers trying to figure out sustainable freemium models
- Founders whose AI product costs are scaling faster than revenue
- Anyone building AI features who needs to understand the real economics before launch
What it won't cover
- Complex enterprise contract negotiation
- Multi-model cost optimization across providers
- Venture fundraising strategy for AI companies
- Technical infrastructure cost modeling
Skills you'll practise
- Unit economics calculation for AI products
- Usage threshold analysis
- Freemium boundary design
- User behavior cost prediction