usage-based-pricing

Usage-Based Pricing

Detailed Guide

Read [the detailed guide](references/detailed-guide.md) before executing this skill. It retains the complete procedure and reference material. Treat its safety, prerequisites, and validation requirements as mandatory. For focused work, load the relevant sections; for end-to-end work, read the guide completely.

When to Use

Use this skill when you need design pricing models that developers understand, accept, and can predict. Trigger phrases: usage-based pricing, API pricing, metered billing, developer pricing, pricing page, cost calculator, pay as you go, pricing transparency, competitive pricing, developer billing.

Design pricing models that developers understand, accept, and can predict—without surprise bills or confusing metrics.

Usage Metrics Developers Accept

Good Metrics: Direct Value Correlation

API calls/requests

• Developers understand what triggers a call
• Easy to monitor and predict
• Scales with actual usage
• Example: Stripe charges per transaction, Twilio per message

Compute time

• Clear relationship to server costs
• Predictable for consistent workloads
• Fair for variable workloads
• Example: AWS Lambda per GB-second, Vercel build minutes

Storage

• Simple to understand
• Easy to predict growth
• Clear cost driver
• Example: S3 per GB stored, databases per GB

Bandwidth/data transfer

• Makes sense for CDN and hosting
• Can be surprising if not monitored
• Example: Cloudflare per GB, Vercel bandwidth

Active users (MAU)

• Works for auth and user-facing tools
• Aligns with customer's growth
• Example: Auth0, Firebase Auth

Problematic Metrics

"Compute units" or proprietary measures

Bad: "1 CU = 0.25 CPU seconds at 1.5GHz equivalent with 256MB memory allocation"
Developers can't estimate usage.

Compound metrics

Bad: "Charged per operation, where operation = read OR write OR delete,
     multiplied by document size factor"
Too complex to predict.

Metrics that punish success

Bad: Per-user pricing that penalizes viral growth
Developer's successful launch becomes a cost crisis.

Metrics with hidden multipliers

Bad: "Per request, but each retry counts, and warming requests count,
     and health checks count"
Actual usage is unpredictable.

Metric Selection Framework

| Metric | When It Works | When It Fails |

|--------|---------------|---------------|

| API calls | Discrete operations | Streaming, persistent connections |

| Compute time | Variable workloads | Idle resources still cost |

| Storage | Data products | Temporary/cache data |

| Bandwidth | CDN, media | Retry-heavy protocols |

| MAU | User-facing apps | Machine-to-machine |

| Seats | Collaboration tools | Individual developers |

Examples: Pricing That Works

Stripe

• Per-transaction percentage (2.9% + 30¢)
• Aligns with customer revenue
• Predictable and simple
• Volume discounts for scale

Twilio

• Per-message/per-minute pricing
• Clear unit costs
• Usage dashboard and alerts
• Prepaid credits for discount

Vercel

• Clear tier structure
• Generous free tier
• Usage-based for bandwidth/builds
• Team pricing separate

DigitalOcean

• Predictable monthly pricing
• Clear size/price relationship
• Hourly billing option
• Bandwidth included in pricing

Examples: Pricing Problems

Confusing Unit Pricing

Some cloud providers:

• Per "compute unit" (undefined)
• Multiple meters per service
• Different rates for different operations
• Bill requires expert interpretation

Enterprise Tax

Some companies:

• SSO requires enterprise tier
• SSO tier is 10x team tier
• No intermediate option
• Punishes security-conscious teams

Punishing Success

Some user-based pricing:

• Free tier: 100 users
• Paid tier: $0.10/user
• Viral success = immediate $$$
• Discourages growth

Limitations

• Use this skill only when the task clearly matches its upstream source and local project context.
• Verify commands, generated code, dependencies, credentials, and external service behavior before applying changes.
• Do not treat examples as a substitute for environment-specific tests, security review, or user approval for destructive or costly actions.