The Startup Equation is a simple inequality that must hold for any viable business:
Where:
- Value - the value to the customer (how much they are willing to pay)
- Price - the price you set
- Cost - your costs to create/deliver the product
If Value ≤ Price - customers will not buy. If Price ≤ Cost - the business loses money. Both conditions must hold simultaneously.
Two Key Gaps
Value - Price = Willingness to Buy
This is the "consumer surplus" - how much value the customer gets beyond what they paid. The larger this gap, the more eagerly customers buy and recommend.
If a tool saves a company $10,000 per month and costs $1,000 - the consumer surplus is $9,000. The purchase decision is obvious. If the tool cost $9,500, the decision would be much harder.
Price - Cost = Margin (Motivation to Sell)
The higher the margin, the more you can invest in marketing, sales, and product development. Low margin = a tough business.
- SaaS: 70-90% margin (low variable costs)
- E-commerce: 20-40% margin
- Services: 30-50% margin (limited by time)
- Marketplace: 5-20% take rate
How to Increase Value
Value is subjective. It depends on how the customer perceives the product.
- Solve a painful problem - the bigger the pain, the higher the value of the solution
- Save time or money - measurable benefit
- Create emotional value - status, confidence, joy
- Demonstrate ROI - help the customer see the value in numbers
"Our tool saves 5 hours per week on reports" → "At an analyst salary of $50/hour, you save $1,000 per month" → Much easier to justify a price of $200/month.
Pricing Specifics for AI Products
AI products have unique characteristics for the formula:
Cost: Variable Inference Expenses
Unlike traditional SaaS, AI products have significant variable costs:
- API call costs (OpenAI, Anthropic)
- Compute for self-hosted models
- Data and embedding storage
If your AI product uses GPT-4, every active user generates $5-50/month in API costs alone. This must be factored into your pricing.
Value: AI as a "Superpower"
AI can create enormous value when positioned correctly:
- Automation: does the work for a human
- Augmentation: helps a human work more efficiently
- New capabilities: things that were previously impossible
Pricing Models for AI
- Per user - simple, but does not account for usage
- Usage-based - fair, but unpredictable for the customer
- Outcome-based - pay for results (hard to measure)
- Tiered with limits - balance of predictability and fairness
Practical Checklist
Check your product against these points:
- Value: Can you calculate customer value in money or time? If not - how can you prove it?
- Price: Are you leaving enough "consumer surplus"? The price should be clearly advantageous for the customer.
- Cost: Do you know your variable costs per customer? Have you accounted for API costs if you use AI?
- Margin: Is it sufficient for investing in growth? For SaaS, aim for 70%+.
- Customers say "too expensive" - value is not communicated or is insufficient
- Low conversion of free users - free tier provides too much value
- Negative unit economics - Cost > Price, need to change the model
This article is a brief overview of a concept from the "AI Founder" course. In the course, we analyze AI product unit economics in detail and help you calculate optimal pricing for your project.