SaaS pricing tiers work when each one maps to a real customer segment, not when you split your feature list into three even piles. Most early-stage founders build tiers backward: they start with "good, better, best" as a template, then stuff features into each bucket until the price feels justified. That is exactly the order that breaks.
A McKinsey analysis of S&P 1500 companies found that a 1% price improvement lifts operating profit by roughly 8%, nearly three times the impact of an equivalent 1% gain in sales volume. Tiering is the single biggest lever you have to capture that gain, because it decides who pays more and why. Get it wrong and you leave your best customers underpriced while confusing everyone else, a mistake covered in more depth in why founders undercharge for their SaaS.
What actually determines how many tiers you need
The number of tiers you need is determined by how many genuinely different customer segments you serve, not by convention. If your buyers are largely similar in need and willingness to pay, three tiers is the right default. If they diverge sharply by use case or company size, a rigid three-tier ladder will misfit half of them.
Stripe's guidance on this, based on patterns across thousands of SaaS businesses, lands on two to four tiers as the sweet spot. Beyond four, buyers struggle to compare plans and conversion drops. Fewer than two and you have nothing to segment against.
Before you pick a number, answer one question honestly: can you describe who each tier is for in a single sentence? "The solo founder testing the product," "the small team that needs collaboration," "the company that needs SSO and audit logs." If you cannot write that sentence for a tier, that tier does not represent a real segment yet. It is just a price point with features attached to justify it.
This is the step most founders skip. They go straight to feature allocation because it feels productive. Segmentation feels slow and abstract by comparison. The pattern shows up constantly in early pricing pages: a three-tier ladder copied from a competitor's site, with no sentence behind any of the three tiers. But every tiering failure below traces back to skipping this step.
The mistake that kills most tier structures
Customer success platform Gainsight learned this the expensive way. For years it sold mid-market customers a classic good-better-best lineup, three tiers priced roughly $1,000, $2,000, and $3,000 a month. On paper it looked like sound segmentation. In practice, almost every deal landed on the middle tier, the top tier had to be discounted to close at all, and higher tiers were stuffed with features that went unused, what pricing consultants call shelfware.
The root cause was not the pricing, it was the segmentation underneath it. Gainsight had treated every mid-market customer as one homogeneous group when in reality a fintech buyer and a healthcare buyer wanted almost entirely different capabilities. No three-tier ladder could serve both well, so the tiers blurred together and customers gravitated to whichever one felt "safe."
The fix was not a fourth tier. Gainsight rebuilt its packaging around modules mapped to actual use cases, letting a customer combine only what they needed instead of graduating through a fixed ladder. Mixpanel ran into a similar problem with its usage-based metric (event volume), which customers could not predict or control, so it introduced a "known users" plan alongside optional add-ons for teams with heavier data needs. Both cases are documented in detail in Monetizely's analysis of good-better-best pricing failures. If your own metric feels arbitrary to customers, the same trade-offs show up in usage-based pricing vs subscription pricing for SaaS.
The lesson generalizes past both companies: if your top tier is rarely bought at list price, or your customers cluster overwhelmingly into one tier regardless of company size, the structure is not reflecting real segments. No amount of feature reshuffling fixes that. Only re-segmenting does.
Structuring SaaS pricing tiers: a three-tier framework for early-stage founders
Most early-stage B2B SaaS companies do not have Gainsight's segmentation complexity yet, which is exactly why a disciplined three-tier structure is usually the right starting point, not a compromise.
- Entry tier: covers one clear job to be done, with a usage or seat cap low enough that a real, engaged user grows past it within a few months. This tier's purpose is proving value fast, not maximizing revenue. Do not gate the feature that delivers your product's core value behind a higher tier here. If a trial user cannot experience the reason they signed up, no upgrade path will save the account.
- Growth tier: this is your workhorse. It should be the default recommendation on your pricing page and the plan most customers land on. Add the capabilities that unlock value as a team scales: more seats, integrations, higher usage ceilings, basic collaboration features. Price it so upgrading from entry feels obviously worth it within weeks, not a marginal improvement.
- Scale or enterprise tier: reserved for governance and control features a growing company needs, not for features you wish smaller customers would pay for. Single sign-on, audit logs, custom contracts, dedicated support. Gate these here specifically because only a subset of customers value them enough to pay a premium, and bundling them lower just creates shelfware.
Two packaging levers do the real work inside this structure. Feature gates should protect access to entirely new capabilities a segment specifically needs, not throttle the core workflow that made someone sign up. Usage limits should sit just below where a genuinely engaged customer naturally lands, so hitting the ceiling feels like a growth milestone, not a wall.
Pick your value metric before you finalize tier pricing, not after. If the unit customers pay for does not track how they get value (seats for a collaboration tool, API calls for infrastructure, records processed for a data tool) every tier built on top of it will feel arbitrary no matter how carefully you name the plans. The full framework for setting the actual numbers inside each tier is in how to price your SaaS product, and if you are revisiting prices on an existing tier structure, see how to raise SaaS prices without losing customers.
What good tiering actually looks like once it works
You do not need a dashboard full of pricing metrics to know if your tiers are working. Three signals tell you almost everything.
Expansion revenue as a share of total growth. If close to none of your growth comes from existing customers moving up a tier, your upgrade path is not doing its job, either because the growth tier is not meaningfully better or because nobody hits its edges.
Plan distribution. If 80% or more of customers sit on your cheapest tier regardless of company size or usage, your growth tier is not priced or positioned to capture the value bigger customers get. If everyone lands on your top tier, you priced it too low for what it delivers.
Self-serve upgrade rate. The percentage of upgrades that happen without a sales call is a direct read on whether a customer can look at your pricing page and immediately see why paying more makes sense for them. Low self-serve upgrade rates usually mean the difference between tiers is unclear, not that customers do not want more.
None of these numbers matter in isolation. Track them together for 60 to 90 days after any tiering change before you decide whether it worked.
What to do this week
Do not redesign your entire pricing page. Pick one thing: write a single sentence describing who each of your current tiers is actually for. If you cannot, that is your answer. Fix the segmentation before you touch a single feature gate or price point. Everything downstream depends on getting that sentence right first.
Frequently asked questions
How many pricing tiers should a SaaS startup have?
Most early-stage SaaS companies should start with three tiers: an entry plan to prove value fast, a growth plan most customers land on, and a scale or enterprise plan for governance features. Go beyond four tiers only once you have clearly distinct customer segments that justify it.
What is the good-better-best pricing model?
Good-better-best is a three-tier structure where each plan expands on the last with more usage, features, or support at a higher price. It works well when your customers are relatively uniform in needs and willingness to pay, and breaks down when segments diverge sharply, as Gainsight's mid-market pricing did.
What features should go in the lowest pricing tier?
The lowest tier should include whatever is necessary for a real user to experience your product's core value, not a crippled demo. Gate advanced governance, integrations, and support tiers higher up, but never gate the reason someone signed up in the first place.
How do I know if my SaaS pricing tiers are wrong?
Three warning signs: almost all customers cluster in one tier regardless of size, your top tier is rarely bought without a discount, or expansion revenue is close to zero. Any of these usually points to a segmentation problem, not a features-and-price problem.
Should I use feature-based or usage-based pricing tiers?
Use feature-based tiers when customer segments need qualitatively different capabilities. Use usage-based tiers when value scales predictably with a metric like seats, API calls, or records processed. Many SaaS companies blend both: a base tier structure with usage limits inside each tier.
Getting your tiers right is a segmentation exercise before it is ever a pricing exercise. Start with the sentence, not the spreadsheet.