---
name: pricing
description: Design SaaS pricing and monetization: value metric, tier structure, price points, freemium, price increases, pricing-page layout. Use when the user says "how much should I charge", "pricing tiers", "freemium", "value metric", "willingness to pay", or "my pricing is wrong". Not for offer construction (bonuses, guarantees, scarcity) on services, courses, or high-ticket B2B: use offers. Not for testing price changes: use ab-testing.
license: MIT
metadata:
  author: TechTide AI (Alex Cinovoj)
  provenance: rewritten from patterns in an open SaaS marketing skills pack (MIT)
  category: Marketing & Growth
---

# Pricing Strategy

Price on value delivered, not cost to serve. The ceiling is the customer's perceived value, the floor is their next best alternative, your price lives between them. Cost is a sanity check, never the basis.

## Workflow

1. **Load context.** Read `.agents/product-marketing.md` or `.claude/product-marketing.md` if present. Then confirm what is missing: product type, target market and GTM motion (self-serve, sales-led, hybrid), current pricing and its performance (conversion, ARPU, churn), competitor pricing, and the goal (growth, revenue, or profitability).
2. **Pick the value metric.** What you charge for must scale with the value received. Test: "as a customer uses more of [metric], do they get more value?" If no, price and value diverge and churn follows. Common metrics: per seat (collaboration), per usage (variable consumption), per contact or record (CRM/email), per transaction (payments), flat fee (simple tools). Good metrics are easy to understand, grow with the customer, and are hard to game.
3. **Structure tiers.** Default good-better-best: entry tier with core features and tight limits, recommended middle tier as the anchor, premium at 2-3x the middle. Differentiate on features, usage limits, support level, and access (API, SSO, custom branding). Detailed tier patterns and persona-based packaging: read references/tiers-and-research.md when designing the actual grid.
4. **Set price points from research, not guesses.** Van Westendorp for the acceptable range, MaxDiff for which features drive willingness to pay. Methods in references/tiers-and-research.md. No research and no time? Anchor on the next best alternative and the value gap, then commit to revisiting with data.
5. **Design the page.** Comparison table above the fold, recommended tier highlighted, monthly/annual toggle with 17-20% annual discount, "who this tier is for" lines, FAQ. Psychology, used honestly: anchor high first, make the middle tier the obvious best value, charm pricing ($49) for value positioning, round ($50) for premium.
6. **Plan increases when signals fire.** Signals: prospects do not flinch, "so cheap" feedback, conversion over 40%, monthly churn under 3%, major value shipped since last change. Mechanisms: grandfather existing customers, announce 3-6 months ahead, tie the raise to new value, or restructure plans entirely.

## Verification

Run the value-metric test on the chosen metric: pick three real or realistic customers of different sizes and compute what each pays per month under the new scheme. Expect the bigger-value customer to pay more, roughly in proportion to value received, and no absurd cliff (a customer 2x the size paying 8x). If the ratios are broken, the metric or the tier limits are wrong: adjust before any page work.

## Good vs Bad

**Bad:** "Competitors charge $30 so we will charge $25, three tiers at $15/$25/$45, features split evenly." Cost-plus-undercut, no value metric, tiers differentiated arbitrarily.

**Good:** "Customers get value per monitored project. Free: 1 project (acquisition loop). Pro $29: 10 projects plus alerting, the anchor. Business $79: unlimited plus SSO and priority support. Annual at 20% off. Van Westendorp on 40 users put acceptable Pro range at $24-38."

## Completion checklist

- [ ] Value metric chosen and passes the "more usage = more value" test
- [ ] Tier differentiation deliberate (features, limits, support, access), not arbitrary
- [ ] Price points anchored in research or explicitly flagged as provisional
- [ ] Three-customer proportionality check done
- [ ] Annual discount and page layout specified
- [ ] Increase strategy defined for existing customers if prices changed

Any box unchecked: not done. Fix or say so.

## Footguns

- **Charging on a metric that punishes success.** Per-seat pricing on a product whose value is automation (fewer humans needed) caps your revenue and fights the customer. Fix: charge on the unit that grows with their win.
- **Freemium as a default.** A free plan is an acquisition investment with real serving costs; it only works when free users convert or refer at meaningful rates. Fix: model conversion before launching free, else use a trial.
- **Silent price increases on existing customers.** Churn spike plus public backlash. Fix: grandfather or give 3-6 months notice tied to shipped value.
- **Copying a competitor's grid.** Their tiers encode their costs, segments, and strategy, not yours. Fix: start from your value metric and personas, then compare.