Pricing Strategy
Pricing model design, competitive analysis, and pricing page optimization.
Install
git clone https://github.com/thatrebeccarae/claude-marketing.git && cp -r claude-marketing/skills/pricing-strategy ~/.claude/skills/
Pricing Models
| Model |
How It Works |
Best For |
Risk |
| Per-seat |
Price per user/month |
Collaboration tools |
Discourages adoption |
| Usage-based |
Pay for what you use |
APIs, infrastructure |
Revenue unpredictability |
| Tiered |
Feature-gated plans |
SaaS with clear feature tiers |
Complexity, feature gating frustration |
| Freemium |
Free tier + paid upgrades |
PLG, network effects |
Low conversion (2-5% typical) |
| Flat-rate |
One price for everything |
Simple products |
Leaves money on table |
| Hybrid |
Base + usage or seat + usage |
Complex products |
Pricing complexity |
| Per-unit |
Price per resource (project, site) |
Agencies, multi-property |
Usage ceiling anxiety |
Value Metric Selection
The value metric is what you charge for (seats, API calls, revenue managed, etc.).
Good value metric criteria:
- Scales with customer value — as they get more value, they naturally pay more
- Easy to understand — customer can predict their bill
- Measurable — you can track and bill accurately
- Controllable — customer can influence their usage
| Product Type |
Common Value Metrics |
| Email marketing |
Contacts, emails sent |
| Analytics |
Events tracked, monthly users |
| CRM |
Seats, contacts |
| Infrastructure |
Compute hours, storage, requests |
| E-commerce platform |
GMV, orders, SKUs |
| Design tools |
Seats, projects, exports |
Competitive Pricing Analysis
Analysis Framework
- Inventory competitors — list all direct and indirect alternatives
- Map pricing models — how does each charge?
- Normalize to common unit — price per seat/month at comparable usage
- Position on value map — plot price vs. feature completeness
- Identify gaps — underserved price points or segments
Positioning Strategies
| Strategy |
Description |
When to Use |
| Premium |
Price above market, justify with value |
Strong brand, unique features |
| Competitive |
Match market prices |
Commoditized market, feature parity |
| Penetration |
Price below market to gain share |
New entrant, land-and-expand model |
| Value |
Price based on ROI delivered |
Measurable customer outcomes |
Good/Better/Best Tier Design
Tier Structure
FREE (or Starter): Entry point — enough to get hooked, not enough for serious use
PRO (or Growth): Primary revenue tier — designed for most customers
ENTERPRISE: High-touch — custom pricing, premium support, compliance features
Tier Design Rules
- Most Popular = middle tier — anchor it visually and with a badge
- Free tier limits should push to paid naturally (not by being crippled)
- Feature fences should align with customer maturity (not arbitrary gates)
- Price anchoring — show enterprise price to make pro feel reasonable
- Annual discount — 15-20% for annual commitment (show monthly equivalent)
Pricing Page Optimization
Best Practices
- Default to annual billing toggle (pre-selected) with monthly option
- Show savings for annual ("Save 20%")
- Highlight recommended plan (visual emphasis + "Most Popular" badge)
- Feature comparison table with max 8-10 rows (not 40)
- CTA per plan with specific text ("Start Free Trial" vs "Contact Sales")
- Social proof per tier ("Used by 500+ teams")
- FAQ section addressing pricing objections
- Money-back guarantee or free trial prominently displayed
Price Testing Methods
| Method |
Sample |
What It Tests |
| Van Westendorp |
Survey 200+ |
Price sensitivity (too cheap, too expensive, bargain, deal-breaker thresholds) |
| Gabor-Granger |
Survey 300+ |
Demand curve at specific price points |
| A/B test |
Website traffic |
Actual conversion at different prices |
| Cohort pricing |
New signups |
Different prices for different cohorts over time |
Integration with Other Skills
- cro-auditor — Pricing page conversion optimization
- ab-testing-framework — Design price tests with statistical rigor
- landing-page-optimizer — Optimize the pricing page layout
- icp-research — Willingness to pay by segment