What is Tiered Usage-based Pricing?
Written by Arnon Shimoni
✓ Expert
Last updated on:
What is tiered usage-based pricing?
Tiered usage-based pricing charges customers per unit of consumption, with the per-unit rate changing as usage crosses predefined bands. The first block of usage is priced at one rate, the next block at a lower rate, and so on. It combines the metering of usage-based pricing with the band structure of tiered pricing.
This is the dominant structure for API products, data platforms, and communication services, anywhere consumption is measurable and marginal cost falls with scale.
How does the math work?
The standard implementation is graduated: each unit is billed at the rate of the tier it falls into, not the rate of the highest tier reached.
An API product prices calls like this:
Tier | Usage band | Rate per 1,000 calls |
|---|---|---|
1 | First 1M calls | $1.00 |
2 | Next 4M calls | $0.80 |
3 | Above 5M calls | $0.60 |
A customer making 6M calls in a month pays:
Band | Calls | Rate | Charge |
|---|---|---|---|
Tier 1 | 1,000,000 | $1.00 / 1k | $1,000 |
Tier 2 | 4,000,000 | $0.80 / 1k | $3,200 |
Tier 3 | 1,000,000 | $0.60 / 1k | $600 |
Total | 6,000,000 | $4,800 |
At a flat $1.00 per 1,000, the same customer would pay $6,000. The graduated structure gives them an effective rate of $0.80, and the discount deepens as they grow. That's the commercial point: the customer's unit economics improve with scale, which removes their main argument for negotiating a custom deal or leaving.
Graduated, volume, or stairstep?
The same 3-tier table can be billed 3 different ways, and the invoice differs each time:
Method | Logic | 6M-call customer pays |
|---|---|---|
Graduated | Each band at its own rate | $4,800 |
Volume | Everything at the reached tier's rate | 6,000 × $0.60 = $3,600 |
Stairstep | Flat fee per band, no per-unit math | The tier 3 flat fee |
Volume pricing is more aggressive and creates boundary cliffs (crossing into tier 3 suddenly reprices all 6M calls). Stairstep pricing trades precision for predictability. Graduated is the middle path, and the most common.
For a full treatment of designing the bands, rates, and commitments around this model, see the guide to tiered usage-based pricing on the blog.
Why do companies choose tiered usage-based pricing?
For customers, it produces a predictable cost curve. They can model exactly what growth costs, and the effective rate only improves. For vendors, it captures revenue proportional to delivered value while building in the volume incentive that enterprise buyers expect anyway.
It also pairs naturally with commitments. Many contracts combine graduated tiers with a minimum commit: the customer commits to $50,000 a year, consumption draws down against it at graduated rates, and overage charges apply beyond it. At that point you're running a hybrid model, which is where most usage-based companies end up.
What does it demand from billing infrastructure?
Three things, and they're all harder than they look:
Requirement | Why it's hard |
|---|---|
Accurate metering | Graduated math on wrong usage data produces confidently wrong invoices. See usage metering. |
Band math at invoice time | Tier boundaries, mid-cycle plan changes, and multiple meters per customer interact |
Rate changes over time | Repricing tiers for new customers while grandfathering old ones means versioned price structures, not a single rate card |
Companies that model tiers in spreadsheets or hand-rolled scripts tend to discover errors during their first enterprise audit, not before. Solvimon treats graduated, volume, and stairstep tiers as first-class price configurations on top of native metering. Flexible pricing covers the details.
FAQ
Is tiered usage-based pricing the same as graduated pricing?
Graduated pricing is the most common way to implement it. Tiered usage-based pricing is the model; graduated, volume, and stairstep are the billing methods.
How do free tiers fit in?
A free allowance is just tier 1 at a rate of $0. This is how most API products structure their free usage: first 10,000 calls free, graduated rates beyond.
What unit should the tiers be defined in?
The unit customers can predict and attribute value to. API calls, gigabytes, messages, tokens. If customers can't forecast the unit, they'll experience the model as random, whatever the rates are.
How does this apply to AI products?
Token-based models are usually tiered usage-based pricing with tokens as the unit. See AI token pricing for how model providers structure it.
Related
The guide to tiered usage-based pricing: the full design walkthrough
Usage-based pricing: the hub term
Tiered pricing: the band structure without metering
Stairstep pricing: flat fees per band
Ready for billing v2?
Solvimon is monetization infrastructure for companies that have outgrown billing v1. One system, entire lifecycle, built by the team that did this at Adyen.
