Usage-based pricing tiers
Usage-based pricing lets small customers start almost free and lets big customers pay what the value is worth - without a sales call. This playbook shows you how to pick the right usage metric, design tiers that don't punish success, and migrate existing customers without a revolt.
- Effort
- Medium-high
- Cost
- $50 - $300/mo
- Time to first result
- 1-2 months
- Best for
- Expansion revenue
Save this strategy and send it to my inbox
Get the full playbook - scripts, tool stack and checklist - delivered as a single email you can act on today.
No spam. One email, and only what is worth reading.

Why usage-based pricing works
Flat subscriptions force every buyer into the same bet. Small teams feel overcharged and churn, while heavy users get enormous value for the same $49 you charge everyone else. Usage-based pricing removes both problems: the entry price drops so trials convert more easily, and revenue expands automatically as customers get more value.
It also changes what your growth depends on. With a usage model, revenue can grow from existing accounts even in a month where you acquire nobody new - the metric most investors call net revenue retention. The trade-off is predictability, which is why the best implementations combine a small platform fee with usage on top.
Three real examples
Stripe charges only when customers make money
Stripe's per-transaction model means a business with zero revenue pays zero. The barrier to starting is essentially removed, and Stripe's revenue scales in perfect lockstep with each customer's success - the cleanest possible alignment of price and value.
Source: Stripe pricingTwilio built a business on pay-per-message
Twilio's pricing is a direct function of usage, with automatic volume discounts as customers scale. Developers can test with a few dollars, then grow into six-figure accounts with no contract renegotiation at each step.
Source: Twilio pricingSnowflake separated storage from compute
Snowflake charges for compute and storage separately, so customers pay for what they actually run. It made switching easy for buyers used to huge upfront licences, and it produced exceptional expansion revenue as data volumes grew.
Source: Snowflake pricingPublish your overage rate openly. Hidden overage pricing is the number one reason buyers distrust usage models - transparency here is a competitive advantage.
The step-by-step playbook
- 1
Step 1: Find the metric that tracks value
The right metric goes up when the customer wins, is easy to explain in one sentence, and can't be gamed. Candidates: messages sent, contacts stored, API calls, tasks automated, GB processed, seats actively using it. Test it against your best customer - does the number match how much they get out of the product?
- 2
Step 2: Model the revenue impact before you build anything
Export 12 months of usage data and price every existing account under the new model. You'll usually find winners and losers. If more than a fifth of your revenue would drop sharply, adjust the platform fee or the included allowance until the model is at worst revenue-neutral.
- 3
Step 3: Add a small platform fee
Pure usage pricing makes revenue unpredictable and lets accounts drop to zero. A modest base fee that includes a generous allowance keeps a revenue floor, protects you from dormant accounts, and still lets customers start cheap.
- 4
Step 4: Design tiers, not a raw meter
Most buyers hate open-ended meters. Package usage into three or four clear tiers with an included allowance and a stated overage rate. People can then predict their bill, which is the single biggest objection to usage pricing.
- 5
Step 5: Build metering and in-app visibility
You need accurate event tracking, and users need a live usage bar showing where they are against their allowance. Send an alert at 80% and 100%. Surprise bills are the fastest route to chargebacks and churn.
- 6
Step 6: Grandfather existing customers
Never force a migration overnight. Keep existing customers on their current plan, show them what they'd pay on the new model, and let them switch when it's cheaper for them. Give at least 60-90 days' notice if you eventually sunset old plans.
- 7
Step 7: Rewrite the pricing page around a calculator
Add an interactive estimator: the buyer enters their expected volume and sees their monthly cost. Include an FAQ covering overages, downgrades, unused allowance and what happens at the limit.
The migration email (steal this)
Use this when you introduce usage tiers to existing customers. Lead with the guarantee that nothing changes for them today - that's what stops the panic.
Subject: New pricing options (your plan isn't changing)
Hi {first_name},
We're introducing usage-based plans for {product_name}. First things first: your current plan and price stay exactly as they are. Nothing changes for you unless you choose to switch.
Why we're doing it: customers told us they wanted to pay in line with how much they actually use {product_name} rather than a flat fee.
Based on your usage over the last 3 months, the new {plan_name} plan would cost you about {estimated_price}/month versus the {current_price} you pay now.
See the new plans and a calculator here: {pricing_link}
If you'd like to switch, just reply and we'll move you over. If you have questions about your specific usage, reply and I'll walk you through the numbers.
{your_name}Questions to pressure-test your usage metric
- Does this number go up only when the customer gets more value?
- Can a customer explain their next invoice without a spreadsheet?
- What stops someone gaming the metric to stay in a lower tier?
- What happens to revenue if our ten biggest accounts switch tomorrow?
- How do we handle a customer who spikes 10x for one month?
If the metric can go up while the customer's outcome stays flat, it will feel like a tax and it will cause churn.
Recommended tools
Stripe Billing
% of revenueMetered billing, tiered pricing and usage records - the fastest way to implement usage tiers without building billing yourself.
Baremetrics
Paid plansTrack MRR, expansion and churn as you migrate, so you can see immediately whether the new model is working.
June.so
Paid plansAnalyse which usage events correlate with retention - the metric you should be charging for.
Datafast
Paid plansSee how pricing page changes affect conversion after you launch the calculator and tiers.
Missive
Paid plansHandle the wave of pricing questions from existing customers as a team, without dropping replies.
Prices are indicative and change - check the vendor before buying.
Do it faster with AI Prompts
Choose the value metric
My product is {product_description}. Customers use it to {job_to_be_done}. Here are the actions users take: {list_of_actions}. Propose 5 candidate usage metrics for pricing, and for each explain how well it tracks customer value, how easy it is to explain, and how gameable it is.
Design the tiers
Design a 4-tier usage-based pricing structure for {product_name}. Metric: {metric}. Current flat price: {current_price}. Typical usage distribution: {distribution}. Give each tier a name, base fee, included allowance, overage rate, and the type of customer it targets.
Write the pricing page
Write pricing page copy for usage-based tiers of {product_name}. Include a headline, one line per tier explaining who it's for, and an FAQ of 8 questions covering overages, downgrades, unused allowance, annual billing and what happens at the limit.
Model the migration
I'll paste customer usage data: {data}. Under this proposed model - base fee {fee}, allowance {allowance}, overage {rate} - calculate each customer's new bill, total revenue change, and flag every account whose bill increases more than 20%.
Mistakes that kill this play
- Charging for a metric that rises without the customer getting more value.
- Pure metered pricing with no base fee, which makes revenue unpredictable.
- No in-app usage meter, so customers only discover overages on the invoice.
- Force-migrating existing customers instead of grandfathering them.
- Tiers so complex nobody can estimate their own bill.
- Launching without modelling the revenue impact on your current book.
Your first 5 days
- Day 1: List candidate usage metrics and score each on value alignment, clarity and gameability.
- Day 2: Export 12 months of usage data and model revenue under two or three candidate structures.
- Day 3: Lock the tiers: base fee, allowance and overage rate for each. Sanity-check against your top 20 accounts.
- Day 4: Spec the metering: which events you count, where the usage bar lives, and the 80% / 100% alerts.
- Day 5: Draft the new pricing page with calculator and FAQ, plus the grandfathering email to existing customers.
Save this strategy and send it to my inbox
Get the full playbook - scripts, tool stack and checklist - delivered as a single email you can act on today.
No spam. One email, and only what is worth reading.