A price experiment that led to a 3x conversion rate.

Problem

An AI productivity assistant on flat per-seat pricing. Heavy users were massively underpriced, and the flat fee was the number customers noticed and resisted.

Method

Designed a controlled usage-based experiment on a small cohort, with a comms split to isolate the reaction to the change.

Solution

Usage tiers layered onto seats. Org-level billing chosen over user-level at a fraction of the build effort.

Impact

15% more revenue in the test cohort. Conversion tripled relative to baseline — roughly 10% to a third of the cohort. Mix shifted toward annual plans.