Pricing Is Your Most Important Product Decision
Most SaaS founders treat pricing as an afterthought. They launch with "$29/month because Notion costs that" and wonder why they can't close enterprise deals or why churn is high. Pricing is a product decision, a market positioning decision, and a growth strategy decision — all at once.
The Three AI SaaS Pricing Models
1. Seat-Based (Per User)
Charge per user per month. Simple to understand, easy to bill. Works when: value is tied to individual users (productivity tools, communication tools). Problem for AI SaaS: a company with 100 employees might have 5 heavy AI users — seat pricing doesn't capture value correctly.
2. Usage-Based (Consumption)
Charge per AI call, per document processed, per token generated. Aligns price with value delivered. Customers love the low start, grow into it. Examples: OpenAI, Anthropic, SendGrid. Best for: AI SaaS where usage varies significantly by customer.
3. Outcome-Based (Value Pricing)
Charge based on results: leads generated, cost saved, revenue influenced. Requires strong attribution and trust. Commands premium pricing when proven. Used by Regie.ai ($ per booked meeting), legal AI tools ($ per contract reviewed).
Choosing Your Value Metric
The value metric is what you bill on — it should: grow with customer value (more use = more value = more revenue), be easy for customers to understand, and be measurable by both you and the customer. Bad metric: number of integrations. Good metric: documents processed, API calls, seats with active usage.
Pricing Page Optimization
- Show 3 tiers maximum (too many options = analysis paralysis)
- Highlight the "most popular" plan (anchors perception)
- Annual discount prominently featured (reduces churn by 30–50%)
- Show savings in absolute dollars, not percentages
- Include ROI calculator or "saves you $X/month" claims with proof
- List competitor pricing to establish favorable comparison