Pricing and packaging

Pricing is part of the product. It decides who you are for, which features matter, how unit economics work, and whether expansion revenue appears without a hero sales motion. Underprice and you starve the business and attract the wrong customers. Overprice without packaging the value and you stall self-serve conversion.

Good pricing captures a fair share of the value you create — not "cost plus a bit," and not "whatever the competitor's website shows." When price and value move together, upgrades feel fair.

Treat pricing reviews as regularly as roadmap reviews, not as a once-at-launch ceremony. Engineers touch this when they build entitlements, paywalls, metering, billing webhooks, and plan-gated features. Prefer centralized checks (can(workspace, 'sso')) driven by billing state. A wrong invoice is a churn email waiting to happen.

A workable picture: charge for a unit that grows with value and with your cost to serve (steps executed, not just seats), bundle it into three tiers so buyers self-segment, let the low tier reach the aha, warn before overage, and grandfather existing customers when you raise prices. Seat-only pricing on an AI or infra product subsidizes power users until margins break. Hidden overage and surprise invoices are retention events of the worst kind — surface projected spend in the product before the bill arrives. Metering events must be idempotent so a retry does not double-charge.

Concepts in this section

Step	Concept	What it answers
1	Value metrics	What unit do you charge for, and does it track value and COGS?
2	Good-better-best	How do tiers let buyers self-segment without gating the aha?
3	Freemium vs trials	Free forever, time-boxed, or reverse trial?
4	Usage-based pricing	How do you meter fairly without bill shock?
5	Willingness to pay	How do you research, test, grandfather and raise prices?

Interview angle: pick a value metric, sketch three tiers, explain what you gate and why, and how you would test willingness to pay without torching trust. Mention grandfathering and how you would migrate entitlements in the data model.

Common mistakes

Seat-only when costs are usage, especially for AI features.
Gating the aha on the free or trial plan.
Too many SKUs and scattered if plan === 'pro' checks.
Hidden overage and surprise bills.
A/B'ing prices on the same plan name for existing customers.

Key takeaways

Price to capture value; pick a value metric that tracks customer value and your costs.
Package good-better-best so buyers self-segment; gate admin/scale, not the aha.
Freemium needs low COGS or strong loops; trials fit expensive-to-serve products.
Usage and hybrid pricing need metering UX that prevents bill shock.
Research WTP, test carefully, and grandfather with a deliberate migration plan.