AI companies should align their pricing with what matters to their customers after demonstrating value, rather than relying solely on fixed or consumption-based pricing models.

When introducing a new product category, initial pricing often reflects the status quo, but as competition increases, prices tend to align with what competitors charge.

Pricing strategy should focus on delivering value to customers, pushing the company to improve continuously rather than adopting a low-cost model that may compromise long-term viability.

Charging by activity and changing pricing based on workflows are effective strategies for AI companies, moving towards value-based pricing rather than cost-based pricing.

Expect your pricing strategy to evolve over time, typically every year or so, as you gather more data about your customers and market.

Project-based pricing is common in consultancy and professional services, but founders should transition to more scalable pricing models as soon as possible.

Pricing should be iterative; continuously adapt and change it based on market feedback and data until you find the right price point.

Three key parts to pricing are: 1) Understand the value equation and price at about a third of the value delivered, 2) Consider your costs to avoid pricing below them, and 3) Differentiate your product to avoid price wars with competitors.

Test your pricing to find out how much customers are willing to pay; as your product matures, your ideal customer profile may change, and you should adapt your pricing accordingly.