How to Set Up Pricing Tiers

Pricing-tier design is one of the few places in a business where small design decisions have large revenue consequences. The same underlying product, priced as three tiers vs. five, or priced per-seat vs. flat-fee, can earn very different ARR per customer, as the worked example below shows. This is not cosmetic, the tier structure determines how your customers grow with you.

1. Pick a value metric that scales

The value metric (Nagle calls it the price metric^[1]) is the unit that ties price to customer success. Good value metrics share three properties:

• They scale with customer success. As the customer gets more value, the metric grows. Seats for collaboration tools, API calls for developer tools, transactions processed for payment tools, revenue managed for accounting tools.
• They are easy to measure. Both sides can see and agree on the number without ambiguity.
• They don't penalise usage. A metric that causes customers to restrict their use of the product (e.g., "every notification sent") is worse than a metric aligned with product value.

Flat-fee pricing works for narrow products where the customer's value does not grow much with use (e.g., single-use utilities). For SaaS with expansion potential, a value metric is usually the better choice, because it lets revenue grow without a renegotiation.

Common mistakes: pricing per-seat when usage is primarily asynchronous (penalises organizations that invite broader participation); pricing per-feature in a way that forces artificial unbundling of naturally-grouped capabilities; pricing on metrics customers can't predict, which drives unpredictable invoices and churn.

2. Three tiers, not five

The standard structure: Starter, Growth (or Pro), Enterprise. Three tiers, mapped to three customer segments.

Why three:

• Clarity: each extra tier is another comparison the buyer has to make. The case is not that more options always lose sales; a meta-analysis of choice-overload studies found the average effect close to zero, with large differences between settings^[2].
• Three is enough to cover the natural segments (solo/small, growth, enterprise) without creating competition between adjacent tiers.
• With three tiers, the middle tier becomes the anchor for decision-making, and upgrades from Starter step to the next tier rather than jumping to Enterprise.

Five-tier structures are usually a sign of indecision, a company that couldn't commit to three clear customer segments. The result is that customers within a tier are heterogeneous, which complicates everything downstream (packaging, support, sales motion).

Legitimate exceptions: genuinely complex products (data platforms with very different use cases at scale) and regulated industries (where compliance tiers add real cost). Even then, four is usually enough.

3. Anchor high, price the middle

Price anchoring is one of the best-known findings in behavioral pricing: in Ariely, Loewenstein and Prelec's experiments, even arbitrary numbers shifted what people would pay^[3]. A later replication found weaker effects and urged caution about building on single studies^[5], so treat anchoring as a nudge, not a lever that sets the price for you.

Practical implications for tier design:

• Show the premium tier first visually. Order tiers highest-price to lowest-price when the page is scanned, even if the buyer's likely choice is the middle tier.
• Make the middle tier obviously better value than the bottom tier. Bottom tier should feel too restrictive for anyone with real needs. Middle tier should feel like what most legitimate users will need. Top tier should contain enterprise features that occasional buyers genuinely need.
• The 3:1 ratio. One starting point: bottom $X, middle ~$3X, top ~$10X. Not a rule, but it creates clear separation.

Willingness-to-pay research (Van Westendorp PSM or similar) should set the price points^[4]. A price set by gut feel has no way to tell you whether it is too low.

4. Build clear upgrade triggers

Expansion revenue comes from customers crossing tier boundaries. Every tier boundary should have an obvious trigger, a limit or feature the customer hits that forces the upgrade conversation.

Good upgrade triggers:

• Usage-based limits. "Starter includes up to X seats / API calls / projects." When customer exceeds, they see a clear upgrade path with pricing attached.
• Feature gates tied to customer maturity. Advanced analytics, integrations, SSO, audit logs. Features that genuinely matter to larger/more sophisticated users, gated at the tier those users occupy.
• Support-level tiers. Response-time SLAs, dedicated account management. These are real value for enterprise, useful for mid-market, and largely irrelevant for individual users.

Bad upgrade triggers: arbitrary feature gates that just feel punitive, essential functionality gated behind paywalls (hurts adoption more than it earns), and gates that force customers to contact sales for pricing visibility (mid-market customers often walk away rather than engage).

Communicate upgrades proactively. In-product notifications when a customer approaches a tier limit, with the specific next tier's price visible, turn the limit into a decision the customer can plan for instead of a wall they hit mid-task.

Review tier structure annually. As the customer base matures and the product evolves, what was the right tier structure two years ago rarely still is. Packaging is a lever worth pulling deliberately.

5. Grandfathering existing customers

When you change pricing or tier structure, existing customers need a separate policy. Three common approaches:

• Full grandfather: Existing customers stay on old pricing indefinitely. Protects renewal rates but creates operational complexity and a permanent revenue gap between cohorts.
• Grandfather with sunset: Existing pricing honored for 12–24 months, then migrated. Communicated well in advance. Balances protection with eventual alignment.
• Migrate at renewal: New pricing applies at next annual renewal. Most common for SaaS. Can be softened with a discount against the new rate for the first year on new terms.

The choice depends on the size of the change and the customer relationship depth. A modest increase with clear notice is a different event from a restructure that raises some customers' bills by 40%: the bigger the change for an individual customer, the longer the transition it needs to avoid damaging renewals.

6. Signals that packaging needs to change

When should you revisit the tier structure? Pay attention to these signals:

• One tier dominates (80%+ of revenue). If almost everyone buys the middle tier and nobody upgrades, the upgrade triggers aren't working. The top tier may not solve a real problem the middle tier doesn't.
• Tier-mix hasn't changed in 18+ months. Healthy pricing drives tier progression as customers mature. Static mix suggests expansion paths are broken.
• Sales discounting consistently on specific tiers. If the enterprise tier is routinely discounted 40%+ to close, the list price is probably aspirational rather than reflecting value. Realign.
• Competitor with similar product charges 2x+ your price. Either they have meaningful differentiation you missed or there's unrealised pricing room in your tiers.
• Customer requests for bundled features that don't map to your tiers. Frequent ad-hoc requests for specific feature combinations signal that your tier boundaries don't match buyer segments.

None of these is conclusive on its own; two or more together is usually enough to trigger a packaging review. The worst time to review pricing is when cash is critical, the decisions tend to be reactive. Build annual packaging reviews into the planning cadence instead.

7. Numeric worked example — three-tier economics

A project-management SaaS has 1,200 customers, flat $29/month pricing, and $418k ARR. Switch to tiered pricing with seats as the value metric, reprice, and model the 12-month impact.

Starter   $19/mo   up to 5 seats, core features
Growth    $49/mo   up to 25 seats, integrations + reporting
Business  $149/mo  unlimited seats, SSO + audit log + priority support

Assumed mix after 12 months
  Starter   35%  →   420 customers × $228/yr  = $96k ARR
  Growth    55%  →   660 customers × $588/yr  = $388k ARR
  Business   8%  →    96 customers × $1,788   = $172k ARR
  Churn     2%  →    24 customers lost on repricing friction
  Total                                         = $656k ARR

The move from flat-fee to three tiers drives roughly 57% ARR lift ($418k to $656k), assuming the mix lands where the packaging design intends. The Growth tier does most of the work: 660 customers paying $240 a year more add $158k. Business adds $138k. Moving 420 customers down to the $19 Starter tier costs $50k, and the repricing churn $8k. The mix is the assumption to test: if half of the projected Growth customers stayed on Starter, the lift would fall to about $119k (28%). Check it against how many current customers actually use more than five seats before you launch.

8. Failure modes worth naming

• Value metric the customer cannot predict. API-call-based pricing sounds clean until a customer gets a $12k invoice after a debug loop. Tie usage pricing to committed-use plans with alert thresholds, or pick a metric customers can forecast — seats, documents, revenue processed.
• Middle tier as a clone of the bottom tier plus one feature. If the only difference between Starter and Growth is one feature, most customers default to the cheaper tier. The middle tier needs to feel like a category-change for the typical growing user, not a minor unlock.
• Hidden pricing on the enterprise tier when the buyer is still self-serve. "Contact us for pricing" on a tier targeted at mid-market SaaS buyers who research without sales contact pushes them to shortlist a competitor whose price they can see. Publish the price unless the deal genuinely requires negotiation (regulated industries, custom contracts, enterprise procurement).

Hybrid pricing, a platform fee plus a usage metric, is one way to get both: the predictability customers want and the expansion mechanics that reward customer success.

Frequently asked questions

How many pricing tiers should a SaaS have?

Three, not five — typically Starter, Growth (or Pro), and Enterprise, mapped to three customer segments. Three covers the natural segments without adjacent tiers competing, each extra tier is another comparison the buyer has to make, and the middle tier becomes the decision anchor, with upgrades from Starter stepping to the next tier. The case is clarity rather than a law of conversion: a meta-analysis of choice-overload studies found the average effect of more options close to zero. Five-tier structures usually signal indecision about segments and leave customers within a tier heterogeneous. Legitimate exceptions are genuinely complex products and regulated industries, where four is usually still enough.

How do I pick a value metric for tiered pricing?

Choose a metric that scales with customer success, is easy for both sides to measure, and does not penalize usage. Good metrics include seats for collaboration tools, API calls for developer tools, transactions for payment tools, and revenue managed for accounting tools. Avoid pricing per-seat when usage is asynchronous, per-feature in a way that forces artificial unbundling, or on metrics customers cannot predict (which produces surprise invoices and churn). A value metric lets revenue grow as each customer's use grows, which a flat fee cannot.

How should I price the tiers relative to each other?

Anchor high and price the middle. Anchoring is one of the best-known findings in behavioral pricing, though replications have found it weaker than the famous early experiments: the first price a customer sees can shift their perception of the rest, so show the premium tier first visually even though most buyers choose the middle. One starting ratio is bottom $X, middle about $3X, top about $10X, with the bottom tier feeling too restrictive for serious needs and the middle feeling like what most legitimate users require. Willingness-to-pay research such as Van Westendorp should set the actual price points, because a price set by gut feel has nothing to say whether it is too low.

When should I revisit my pricing tiers?

Review packaging at least annually, and act when you see warning signals: one tier dominating 80-plus percent of revenue with no upgrades (broken upgrade triggers), a tier mix unchanged for 18-plus months, consistent 40-plus percent discounting on a specific tier (aspirational list price), a competitor charging 2-plus times your price, or frequent ad-hoc requests for feature bundles that do not map to your tiers. Two or more signals together usually justify a review. The worst time to review is under cash pressure, when decisions turn reactive.