The AI pricing playbook for product leaders
Ship AI features that retain customers, protect margins, and survive the shift to usage-based pricing
Summary
Most AI features don't fail because of pricing. They fail because the AI doesn't add enough unique value to the job it's meant to do. Usage grows, costs scale with it, and margins erode before anyone has a real conversation about packaging.
We sat down with Kyle Poyar (Founder, Growth Unhinged) and Aakash Gupta (Founder, Product Growth), who work daily with teams shipping AI inside Atlassian, Bolt, Salesforce, Cursor, HubSpot, and dozens of other products. They unpack the patterns they keep seeing: the 10% of power users driving 80% of cost, the AI tourists who churn before paying back, the per-seat models that quietly subsidize your best customers.
This guide gives you sharper language to talk about AI margins, a clearer framework for sequencing value before pricing, and concrete tactics to change pricing on live products without breaking customer trust.
What's inside 👀
Why retention beats pricing - and how the power user cost curve quietly destroys per-seat margins
4 pricing models compared, plus the hybrid credit pattern now used by Atlassian, Adobe, Bolt, Clay, Cursor, HubSpot, and Lovable
The anti-patterns to avoid, from AI tourist churn to Cursor-style pricing surprises to outcome-based billing disputes
Practical tactics from Kyle Poyar and Aakash Gupta for making the business case, modeling scenarios, and phasing pricing changes