Pricing Tests
Test pricing changes with cohorts to optimize revenue without hurting conversion.
npx gtm-skills add product/upsell/pricing-experimentOutcome
1 pricing variant tested with directional revenue-per-user signal from 50+ users
Leading Indicators
- Revenue per user (treatment vs. control)
- 30-day churn rate (treatment vs. control)
- Checkout conversion rate
Instructions
-
Run the
usage-pricing-model-analysisdrill to identify the value metric, map usage distribution, compute churn by usage band, and model 3 pricing scenarios. -
Produce a structured hypothesis: "If we change from [current] to [variant], then revenue per user will [change] by [%] because [reasoning], and churn will not increase by more than [limit]."
Human action required: Review and approve the hypothesis before proceeding.
- Run the
pricing-experiment-runnerdrill to create variant Stripe prices, set up a PostHog feature flag (50/50 split, 50+ users), build the n8n migration workflow, and define guardrail metrics.
Human action required: Review Intercom message copy before launch.
-
Monitor for 1 billing cycle. The pricing-experiment-runner weekly monitoring checks guardrails automatically.
-
Run the
threshold-enginedrill to evaluate: did 50+ users complete a billing cycle? Is there a directional signal on revenue per user? Did churn stay within guardrails? -
If PASS, document and proceed to Baseline. If FAIL, diagnose and iterate.
Recommendations
Time
8 hours over 2 weeks
Play-specific cost
Free