Blog
Tiered pricing explained: how to design SaaS pricing tiers
Package tiers, graduated pricing and volume pricing: what each means, worked examples across seven usage levels, and how to design tiers customers choose
Published:
·
14 min read
“Tiered pricing” sounds like one idea, but it is used for three different things. Some companies mean a set of packages at different prices. Some mean a price per unit that changes in bands, where each unit is charged at the band it falls into. Others mean a price per unit where the whole quantity is charged at the rate of the band it reaches. The three give different bills for the same customer, and the pages that rank for this keyword rarely separate them.
This guide does. It defines the three structures precisely, runs the same usage through each so you can see the difference in dollars, explains the price cliffs that one of them creates, and shows how to design package tiers that customers choose. It also covers what tiers do to your billing: mid-period upgrades, downgrades and proration. Where we rely on a rule or a statistic, it names the page it comes from. Where common advice online has no traceable source, we say so.
What is tiered pricing?
Tiered pricing means that the price depends on which band or package a customer falls into, rather than being one fixed amount per unit or per customer. The bands can be defined by features, by limits or by volume. That one sentence hides three different structures:
Package tiers: a few plans (for example Starter, Growth and Scale) at different prices, each with more features or higher limits.
Graduated pricing: the price per unit changes in bands, and each unit is charged at the rate of the band it falls in.
Volume pricing: the price per unit changes in bands, and the whole quantity is charged at the rate of the band it reaches.
Payment and billing documentation uses these words inconsistently. Stripe’s documentation uses “tiered pricing” as the umbrella and then separates “volume” and “graduated” modes. Other billing vendors use “tiered” to mean graduated and “volume” to mean all units at one rate, and at least one vendor’s documentation uses “volume tiers” to mean a flat fee per tier. A vendor’s label tells you little. The reliable way to tell which one you are looking at is the arithmetic.
The one-line test. If a customer crosses a threshold, does the rate change for the units they already used? If yes, it is volume pricing. If no, it is graduated pricing.
Tiered pricing vs volume pricing: the three meanings side by side
Structure | What the tier decides | Crossing a threshold | Total bill can fall as usage rises? |
|---|---|---|---|
Package tiers | Which plan the customer buys | The customer moves to another plan | Only if the higher plan is cheaper, which it should not be |
Graduated | The rate for units inside each band | Only the extra units use the new rate | No |
Volume | The rate for every unit, set by the band reached | All units are re-priced | Yes |
Worked example 1: Stripe’s own numbers
Stripe’s documentation uses a simple three-band example: units 1 to 5 at $7 each, units 6 to 10 at $6.50 each, and 11 or more at $6 each. The same quantity produces different totals depending on the mode:
Quantity | Volume total | Graduated total |
|---|---|---|
5 | $35 | $35 |
6 | $39 | $41.50 |
20 | $120 | $127.50 |
25 | $150 | $157.50 |
At 6 units, volume pricing charges all six units at $6.50, which is $39. Graduated pricing charges five units at $7 and one unit at $6.50, which is $41.50. The gap between the two modes does not shrink as quantity grows. Here it settles at $7.50.
Worked example 2: our numbers across seven usage levels
Here is a second example with a larger spread, so the effects are easier to see. All figures are illustrative.
The bands: $0.10 per unit for units 1 to 1,000, $0.08 per unit for units 1,001 to 5,000, and $0.06 per unit for 5,001 and above.
Units used | Volume pricing | Graduated pricing | Gap |
|---|---|---|---|
500 | $50.00 | $50.00 | $0.00 |
1,000 | $100.00 | $100.00 | $0.00 |
1,001 | $80.08 | $100.08 | $20.00 |
3,000 | $240.00 | $260.00 | $20.00 |
5,000 | $400.00 | $420.00 | $20.00 |
5,001 | $300.06 | $420.06 | $120.00 |
10,000 | $600.00 | $720.00 | $120.00 |
How the numbers work. At 3,000 units under graduated pricing: 1,000 units at $0.10 is $100, and the next 2,000 at $0.08 is $160, so $260. Under volume pricing all 3,000 units fall in the second band, so 3,000 x $0.08 is $240. At 10,000 units under graduated pricing: $100 for the first 1,000, $320 for the next 4,000 and $300 for the final 5,000, so $720. Under volume pricing all 10,000 units are charged at $0.06, so $600.
What the example shows.
Volume pricing has cliffs. A customer using 1,000 units pays $100. A customer using 1,001 units pays $80.08, which is $19.92 less for using more. At 5,000 units the bill is $400. At 5,001 it is $300.06, which is $99.94 less.
Graduated pricing never falls. The bill rises with every unit, so a customer is never better off using less.
The gap grows by band. The difference between the two structures steps from $20 to $120 as the customer reaches the third band. A finance team comparing quotes will see it.
Effective rates differ. At 3,000 units, volume pricing works out to $0.08 a unit. Graduated pricing works out to about $0.087.
The cliff problem, and what customers do about it
A cliff is a point where using more costs less. It sounds like a gift to customers, but it creates behaviour you did not design for:
Holding back usage just before a threshold, or timing it to land in a later period, to stay under or push over a band boundary.
Splitting accounts to stay in a cheaper tier for each, or merging accounts to reach a higher one.
Seat capping where tiers are defined by user counts.
Quote disputes, because the same customer can get a lower bill for more usage and will ask why.
We did not find research that measures how common these behaviours are, so treat them as risks to check, not as established findings. The design choices that avoid them are simple:
Use graduated pricing when you want the bill to rise smoothly with usage.
Use volume pricing deliberately, when you want a visible reward for crossing a threshold, and set the cliff so that it is small and explainable.
Add a minimum charge at the boundary if you must keep volume pricing but want to remove the drop.
Show the calculation on the invoice, so a customer can reproduce the bill.
Flat fees per tier, and what happens at zero usage
Tiers can carry a flat fee as well as a per-unit price. Stripe’s documentation works through an example with a 12-unit quantity, per-unit prices of $5, $4, $3, $2 and $1 across five bands, and flat fees of $10, $20, $30, $40 and $50. In volume mode the total is $66: 12 units at $3 plus a $30 flat fee. In graduated mode it is $111, because each band’s units and flat fee are summed.
Two practical details from the same documentation. First, at a quantity of zero, the first tier’s flat fee is still billed, so a customer who has not used anything can still receive a charge. To bill nothing at zero usage, Stripe suggests a first tier that covers one unit at the flat rate with no flat amount. Second, a price can be edited only until a subscription uses it, which matters when you plan changes to tiers (see the migration section below). Other billing systems differ, so check the rules for the system you use.
Flat-rate pricing vs tiered pricing
Flat-rate pricing charges one price for the same thing. Common forms are one plan with everything included (for example $49 a month), one fee per account regardless of size, or one price per unit however many units are bought. It is simple to explain and to bill, and there is no decision for the customer.
Tiered pricing trades some of that simplicity for fit. Choose flat-rate when your customers are alike and your product is narrow. Choose tiers when they are not, and when you want price to rise with the value a larger customer gets.
How to design pricing tiers
The steps below apply to package tiers. They are our method, built from how pricing works, not findings from a study.
Start from segments, not features. Decide who each tier is for before deciding what goes in it. A tier with no clear buyer will not sell.
Choose a value metric that sets the limits. Seats, usage or records can define how a customer moves up, so the price rises as their value does.
Choose the tier count deliberately. Three is common because buyers can compare three options quickly. Add a custom plan for large buyers. See the evidence box below before you take a number as fact.
Fence the right features. Put the things bigger customers need, such as higher limits, advanced controls or support levels, in the higher tiers. Do not lock away basics that every customer needs.
Make each step up clearly better. The jump in price should be matched by a visible jump in value.
Name tiers after the customer. A name that says who it is for helps a buyer choose faster than a name that only signals size.
Check unit economics. For each tier, work out what a typical customer costs to serve, and check the margin is acceptable at the limit of the tier, not just at the average.
How to choose tier boundaries from usage data
If your tiers are defined by usage or limits, set the boundaries with data, not instinct. This is our method:
Pull a month of usage per customer and sort it.
Look at the distribution. Usage is usually very uneven: a long tail of small customers and a few very large ones.
Try boundaries at natural gaps or percentiles of the distribution, so each tier holds a meaningful group.
Calculate revenue and margin per tier at each candidate boundary, using each tier’s price and the cost to serve.
Check that few customers sit right at a boundary. A cluster just under a limit suggests people are managing their usage around it.
Test with a cohort of new customers before you change existing ones.
What the evidence does and does not say about tier counts
A lot of advice states that “three tiers” or “three to five tiers” converts best. We looked for the research behind that and could not find a verified study of SaaS pricing pages that supports a specific number.
What the underlying research says is more mixed:
Choice overload. A meta-analysis by Scheibehenne, Greifeneder and Todd of 50 experiments, covering over 5,000 participants, found a mean effect close to zero, with large differences between studies. The authors identified possible preconditions for overload but no conditions that reliably produce it. A popular-science summary of the same literature describes the evidence as mixed, with failed replications and effects that depend on context.
Decoy effects. Heath and Chatterjee’s meta-analysis and experiment found that adding a decoy option reduced the share of lower-quality alternatives more than higher-quality ones. A conventional sample showed that pattern and a less conventional sample reversed it, so the authors say it may not hold across all segments.
The honest takeaway: fewer, clearer options are a reasonable design principle, and decoy-style comparisons can influence choice in some settings. Neither justifies a specific number of tiers without testing it on your own customers.
What tiers ask of your billing system
Tiers mean that customers move. They cross thresholds, upgrade in the middle of a month, downgrade, pause and sometimes ask for a refund. Each move has to be billed correctly.
Prorations apply to charges billed in advance. Stripe’s documentation gives a simple example: upgrading from a $10 plan to a $20 plan halfway through the period bills an extra $5, made up of a $5 credit for the unused part of the old plan and a $10 charge for the new one.
Usage-based charges are not prorated in that documentation. So with tiered usage pricing, a mid-period change does not produce a prorated usage charge; the usage is rated by the structure in force.
Credits are not always refunded. In Stripe’s documentation, negative prorations are not refunded automatically, and credit calculations can differ between billing modes.
Preview before you change. Billing systems with invoice previews let you see what a change will charge before you make it.
We could not verify from the documentation how a quantity change interacts with graduated or volume tiers mid-period, so test that case in your own billing system before you launch. Do not assume.
For the general case, a billing system needs to measure quantity accurately, apply the right structure at the right moment, work out part-month charges, credits and refunds, and explain each line on the invoice. OpenByte works out part-month charges, pauses, credits and refunds automatically, so every change is billed correctly without anyone having to touch it. See the Billing page or read more about recurring billing software.
How to migrate customers to new tiers
Run every existing customer through the new structure and list the winners and losers. Anyone whose bill rises sharply needs a conversation before the change.
Decide on grandfathering. Keep existing customers on their old tiers for a fixed period or indefinitely. Indefinite grandfathering means supporting two structures.
Move customers at renewal where you can, so a change lands at a natural decision point.
Preview invoices for a sample of customers before you switch.
Check the mechanics. In some systems, editing a price is possible only until a subscription uses it, so a tier change means creating new prices and moving subscriptions to them.
Watch the first cycle. Look at churn, downgrades, failed payments and support tickets.
For more on changing prices, see our guides to SaaS pricing strategy and SaaS pricing models.
Common tiering mistakes
Using the word “tiered” without saying which kind. Buyers, sales and finance will each assume a different structure.
Too many tiers. Choice becomes confusion, and buyers delay.
Tiers that differ only trivially. If customers cannot see why they would pay more, they will not.
Limits customers hit by surprise. An unexpected cap is a support ticket and sometimes a cancellation.
Boundaries that create cliffs and invite gaming.
No upgrade path. Customers who outgrow a plan need an obvious next step.
Changing tiers without a plan for existing customers.
Vendor labels: why you cannot trust the word
Before you compare vendors or read a pricing page, apply the arithmetic test.
Stripe’s documentation uses “tiered” as the umbrella and separates “volume” and “graduated” modes.
Some vendors and blogs use “tiered” to mean graduated and “volume” to mean all units at one rate.
At least one billing vendor’s documentation uses “volume tiers” to mean a single flat fee for whichever tier total usage lands in.
Some glossaries say “tiered” and “volume” are interchangeable and then define them as different things.
If you are told a structure is “tiered”, ask for the bill at three usage levels either side of a boundary. The numbers will tell you what you are buying.
Frequently asked questions
What is tiered pricing?
A structure where price depends on which band or package a customer falls into. It can mean package tiers, graduated pricing or volume pricing.
What is the difference between tiered pricing and volume pricing?
Under graduated (often called tiered) pricing, each unit is charged at the rate of the band it falls in. Under volume pricing, the whole quantity is charged at the rate of the band it reaches. In the worked example above the two differ by $20 at 3,000 units and by $120 at 10,000.
Why does my bill go down when I use more?
Because of volume pricing. When usage crosses a threshold, all units are re-priced at the lower rate, so the total can fall. Graduated pricing does not do this.
How many pricing tiers should I have?
There is no verified study that gives a number. Fewer, clearer tiers are a reasonable principle. Test options with your own customers.
How do I choose tier boundaries?
Use the distribution of customer usage, try boundaries at natural gaps or percentiles, check margin per tier and test with a cohort before changing existing customers.
How does proration work with tiers?
For charges billed in advance, a mid-period change produces a credit for the unused part and a charge for the new plan. In Stripe’s documentation, usage-based charges are not prorated. Test how your billing system treats quantity changes across tier boundaries.
Can I combine package tiers with usage tiers?
Yes. A common hybrid is a package plan with an included usage allowance and graduated pricing beyond it. See SaaS pricing models for how hybrids are built.
What is a flat fee per tier?
A fixed amount attached to a tier in addition to, or instead of, a per-unit price. Whether the flat fee of every tier is summed or only the reached tier’s fee applies depends on the mode.
Key takeaways
“Tiered pricing” means three different things: package tiers, graduated pricing and volume pricing. Always ask which one someone is describing.
The same usage gives different bills under each. In our worked example, volume pricing creates cliffs where a customer pays less for using more.
Package tiers work best with a clear step between plans, a middle option most people choose, and limits customers can understand.
Tiers change what happens at upgrade, downgrade and proration, so check how your billing system handles each.
Be wary of vendor labels. Ask for an example calculation before you trust the word “tiered”.
Before you publish a structure, run a customer’s usage through it. The numbers here are invented for illustration, so swap in your own.
Keep reading
SaaS pricing models: types, examples and how to choose
Seven ways to charge for software, one worked example across five structures, a scorecard to choose, and what each model asks of your billing system
SaaS pricing strategy: how to price your SaaS product
Three ways to set a price, six steps with the arithmetic shown, and how to raise prices without losing customers
SaaS revenue models and billing models explained
Three layers that get confused, worked examples of timing, trials, proration and failed payments, and what each model asks of your operations
