How Should You Structure Your SaaS Pricing Tiers?

How Many Pricing Tiers Should You Have?

Three or four, and the reason is not aesthetics. Each tier has to do a distinct job for a distinct buyer, and most companies cannot name four genuinely different buyers. Add a fifth and you are usually splitting one audience in half rather than reaching a new one.

We build the pricing pages these decisions land on, which means we see the packaging argument at the exact moment it stops being theoretical. A tier that made sense in a strategy document often cannot be explained in a table.

Here is a framework for deciding the structure, what the available data shows about how other companies do it, and the questions we ask before adding anything.

What Does the Tier Data Actually Show?

Clustering around three to four. SaaS Research Lab published a study in March 2026 of 110 SaaS products with public self serve pricing pages, sourced from Product Hunt, G2, Capterra, TrustRadius and Crunchbase. It found an average of 3.5 tiers per product.

The methodology is worth stating because the sample is small. Every page was read by hand and scored against the same 24 field rubric, with two readers and disagreements resolved by a third, and all prices normalised to monthly list rates in US dollars for a first time visitor.

Two other findings from that study are useful anchors. It reports that 54.1 percent of products offer a permanently free tier, and that 38.2 percent offer a time boxed trial. It also found that 67 percent of four tier products include freemium, against 41 percent of three tier products, which suggests the fourth tier is often the free one rather than a new paid segment.

What Job Is Each Tier Doing?

This is the question that settles the count. Write down, in one sentence each, who the tier is for and what problem it solves that the tier below does not. If two sentences describe the same person, you have one tier wearing two names.

The usual honest set is three. An entry tier for an individual or a small team who need to get started without approval. A middle tier for a team that has adopted the product and needs collaboration, permissions and integrations. A top tier for an organisation that needs security, control and support commitments.

Notice those are three different buyers, not three sizes of the same buyer. That is the test. Tiers that differ only by volume are a pricing metric problem, not a packaging problem, and adding tiers is the wrong fix.

Should You Have a Free Tier?

Only if it produces something you need. A free tier is a marketing cost with a support burden attached, and it is worth it when it generates either qualified usage data or a network effect. It is not worth it as a substitute for a clear value proposition.

The data suggests it is close to a coin flip in practice. With 54.1 percent of the studied products offering a permanent free tier and 38.2 percent offering a time boxed trial, neither approach is the industry default. That is a genuine choice rather than a convention you have to follow.

The design question underneath it is what the free tier is limited by. Limit by capacity and people upgrade when they grow, which is healthy. Limit by removing a feature that makes the product work and people churn instead, having concluded the product does not work. We covered the adjacent decision in our piece on free trial or demo.

What Should Each Tier Be Priced On?

On a metric that grows when the customer's value grows, and that they can predict. Seats work when the product is used by people. Usage works when value is transactional. The failure case is a metric that rises with something the customer cannot forecast, which turns every renewal into a negotiation.

Keep the metric the same across tiers wherever you can. Charging per seat on the middle tier and per transaction on the top tier makes the upgrade impossible to evaluate, and a buyer who cannot compare two of your own plans will delay rather than choose.

If you are considering a usage component, treat it as a separate decision from tiering rather than folding it in. It has its own failure modes, which we set out in our guide to usage based pricing.

Which Features Go Where?

Put the features that scale with organisational risk at the top, and the features that scale with team size in the middle. Single sign on, audit logs, advanced permissions and compliance documentation belong to the tier bought by companies who have a security review. Collaboration and integrations belong to the tier bought by teams.

Resist gating the thing that makes the product good. Every packaging exercise produces a suggestion to move a core capability up a tier to drive upgrades. It drives churn instead, because the entry tier stops demonstrating why anyone should pay more.

Keep the feature list short enough to read. A pricing table with forty rows is not communicating, it is defending itself against sales objections that should be handled in a conversation. Our guide to pricing page design covers how to present what remains.

Should You Publish the Enterprise Price?

Publish something, even if it cannot be a single number. Nielsen Norman Group's B2B research found that people view companies that hide costs as being evasive and untrustworthy, and that participants went to competitors' sites to find pricing, making those competitors more likely to reach the shortlist.

Its research also reports that business customers identified pricing as the top most needed piece of information online. Contact us is not an answer to that. It is a request for the buyer to spend their time before you spend any of yours, which is the wrong way round early in an evaluation.

Where an exact figure is impossible, Nielsen Norman Group's guidance from its earlier work is to show prices for typical scenarios. A starting from figure, or a worked example at a representative size, gives an evaluator enough to keep you on the list. That is all it needs to do at this stage.

How Does the Buying Committee Change Your Packaging?

It means your tiers are read by people with different jobs. The 6sense B2B Buyer Experience Report for 2025 found buying groups averaging more than ten members on deals averaging 250,000 dollars, with buyers evaluating 5.1 vendors on average.

Ten people reading the same pricing table want different things from it. The champion wants to know it does the job. The finance ratifier wants to know what it costs at renewal. The security reviewer wants to know which tier includes the controls. If your table only answers the first, the other two go looking elsewhere.

6sense also found the point of first contact sits at 61 percent of the way through the journey, so most of that reading happens with nobody from your company present. The page has to survive being read alone, forwarded, and argued over in a meeting you are not in.

When Should You Change Pricing?

When the packaging stops matching how customers actually buy, not on a calendar. The signals are consistent: deals stalling at the same tier boundary, sales routinely discounting the same line, or customers buying a tier and using only one feature from it.

Change the packaging before you change the numbers. A price increase on a structure that already confuses people amplifies the confusion. A clearer structure often improves realised revenue on its own, because buyers stop defaulting downwards out of uncertainty.

Then grandfather existing customers explicitly and say so publicly. The cost of that is smaller than the cost of the trust damage, and it removes the single biggest internal objection to ever touching pricing.

Where Should You Start?

With the one sentence test. Write who each tier is for and what it solves, one sentence per tier, and show it to someone outside the company. If they cannot place themselves, the structure is wrong regardless of what the table looks like.

Then check that a first time visitor can find a number. Not a range for every configuration, just enough to know whether they are in your bracket. That single change fixes more pricing pages than any restructure.

Finally, read your own pricing page as each member of a ten person buying committee in turn. It is a tedious exercise and it will show you exactly which question your page does not answer.

If you want help pressure testing a packaging change before it ships, we are happy to work through it with you. Find us at phoenix.studio.