SaaS Pricing Models: Tiered vs Per-Seat vs Usage-Based — How to Choose the Right One

What You'll Learn

• ✓Compare the 5 major SaaS pricing models (tiered, per-seat, usage-based, freemium, hybrid) and their economics
• ✓Match pricing model to product type, customer segment, and stage of growth
• ✓Understand the metric implications (ARR vs revenue, churn rate, expansion) of each model
• ✓Avoid the common pricing mistakes that cap SaaS growth

The Direct Answer: 5 Models, Each With Specific Strengths for Specific Products

The five major SaaS pricing models each work well for specific product types and customer segments: **Tiered pricing** (e.g., Starter/Professional/Enterprise): multiple fixed price points with progressively more features or limits. Works well when features or limits naturally segment customers. The simplest model for customers to understand. Examples: Slack, Notion, Figma (for teams). **Per-seat (per-user) pricing**: charge a monthly fee for each user added to the account. Scales linearly with team size. Works well for collaboration tools where each user gets distinct value. Examples: Slack, Notion, Linear, Asana. **Usage-based pricing** (metered pricing, pay-as-you-go): charge based on consumption — API calls, data volume, compute hours, messages sent. Aligns cost with value but creates unpredictable bills. Works well for infrastructure and developer tools. Examples: AWS, Stripe, Twilio, OpenAI API, Snowflake. **Freemium**: free tier + paid tier. Free tier functions as marketing and discovery; conversion to paid drives revenue. Works when the free tier is valuable enough to attract wide adoption but limited enough to drive upgrades. Examples: Dropbox, Zoom, Canva, Spotify. **Hybrid**: combination of two or more models. Common pattern: tiered structure + per-seat within each tier + usage-based overages. Captures the benefits of multiple approaches but adds complexity. Examples: HubSpot (tiered + per-seat), Datadog (usage-based + per-seat for some features). The biggest pricing mistake: picking a model because competitors use it, without considering whether it fits YOUR product's value proposition. The right question is not 'what do similar companies charge?' but 'what unit of value does my product deliver, and how do I meter that?' Snap your product description and customer data to BusinessIQ and it recommends the pricing model that best matches your product's value delivery, customer segment, and competitive landscape. This content is for educational purposes only.

Tiered Pricing: The Default for Most SaaS

Tiered pricing is the most common SaaS model because it works for almost any product and is the easiest for customers to understand. A typical tiered structure has 3-4 plans with progressively more features, users, or limits. The entry tier captures price-sensitive customers; the middle tiers capture the majority of revenue; the enterprise tier captures large-budget customers with custom needs. Common tier structures: - **3-tier**: Starter / Professional / Enterprise. Simple, clear, guides customers to the middle option. - **4-tier**: Free / Starter / Pro / Enterprise. Adds a free tier for acquisition. - **5-tier**: Gets complex. Usually a sign the pricing page is trying to capture too many segments at once. **Design principles for tiered pricing**: 1. **Price anchoring**: the highest tier makes the middle tier look reasonable. Set the enterprise tier at a premium (2-5x the middle tier) so customers perceive the middle tier as a good deal. This is standard retail psychology applied to SaaS. 2. **Goldilocks effect**: most customers pick the middle tier. Design the middle tier to be your target revenue point. Most pricing pages display three tiers with the middle one highlighted as 'most popular' — this is deliberate. 3. **Feature differentiation**: the difference between tiers should be meaningful features or limits, not trivial ones. 'Up to 3 projects' vs 'Unlimited projects' is meaningful. 'Up to 10 users vs 12 users' is arbitrary and feels like price gouging. 4. **Avoid sandbagging**: do not deliberately cripple a lower tier to force upgrades. Customers can tell when a tier is designed to be painful. It works in the short term but destroys trust. 5. **Grandfathering**: when you raise prices, existing customers should keep their current rate for at least 12 months. New customers pay the new price. This prevents churn from price shocks. **Common tiered pricing mistakes**: - Too many tiers (decision paralysis). - Unclear differentiation between tiers. - Pricing the top tier too low (leaves money on the table with large customers). - No free tier or trial (creates friction for product-led growth). - Flat pricing within tiers (no per-seat scaling) for collaboration tools where larger teams get more value. The tiered model works particularly well for products where usage patterns are hard to predict or where the value delivered is feature-based rather than consumption-based. It does NOT work well for infrastructure services (where usage varies 10x or 100x between customers) or for consumption-oriented products (where the value is proportional to use). BusinessIQ helps you design tier structures with specific recommendations on feature differentiation, price anchoring, and the target customer for each tier.

Per-Seat Pricing: The Collaboration Tool Default

Per-seat pricing charges a fixed monthly amount for each user in the customer's account. A team of 5 pays 5x; a team of 50 pays 50x. This model dominates collaboration tools (Slack, Notion, Linear, Asana, Figma) because each user gets distinct value from the product and the growth of a customer's team directly correlates with increased usage. **Why per-seat works for collaboration**: - Each user has their own workspace, account, or content. - Value scales directly with team size (more users = more value). - Revenue scales naturally with customer growth without requiring negotiation. - Expansion revenue happens automatically as customers grow their teams. **The economics**: per-seat pricing produces a clear ARR-per-customer metric. A $15/seat/month product with a 20-user customer generates $3,600 ARR ($15 × 20 × 12). If the customer adds 10 more users next year, ARR grows to $5,400 — 50% expansion without any upselling effort. This is called 'net dollar retention' (NDR) and is one of the most important SaaS metrics. The best SaaS companies achieve NDR of 120-150% (existing customers grow their spend by 20-50% per year even after accounting for any churn). **Per-seat vs per-active-seat**: some companies charge for every user on the account whether they log in or not. Others charge only for 'active seats' (users who logged in during the billing period). Active seat pricing is more customer-friendly but creates revenue unpredictability. Most successful SaaS companies charge for all provisioned seats. **Per-seat challenges**: - Customers try to share accounts (user A logs out, user B logs in). Most SaaS products detect and prevent this via simultaneous session limits or device fingerprinting. - Price sensitivity at scale — a 500-person company paying $15/seat is $7,500/month or $90,000/year. That draws procurement scrutiny and negotiation pressure. Most companies introduce volume discounts or enterprise pricing at that scale. - Not all users get equal value — some power users get massive value, others barely use the product. Flat per-seat pricing over-charges light users and under-charges power users. Some companies address this with 'light' and 'full' seat pricing. **The per-seat + tiered hybrid**: many products combine per-seat pricing with feature tiers. Notion, for example, has Free / Plus / Business / Enterprise tiers, each with per-seat pricing at different rates. A 20-user team on the Plus tier pays different per-seat rates than the same team on the Business tier. This captures both feature differentiation and team-size scaling. The decision: per-seat pricing is the right choice when (1) each user gets distinct value, (2) customers accounts grow with team size, and (3) you want expansion revenue to happen automatically. It is the wrong choice for infrastructure, consumer products, and any case where value is not tied to individual user accounts.

Usage-Based Pricing and When to Go Beyond Subscriptions

Usage-based pricing (also called 'metered' or 'pay-as-you-go') charges customers based on consumption of a specific unit of value — API calls, data transferred, compute hours, messages sent, transactions processed. This model has grown dramatically in popularity over the past decade because it perfectly aligns customer cost with value received. **Classic usage-based examples**: - AWS: pay per compute hour, per GB stored, per API call. - Stripe: percentage + fixed fee per transaction. - Twilio: per SMS sent, per voice minute. - OpenAI API: per token processed (input and output). - Snowflake: per compute credit consumed (roughly per query). - Mailgun / SendGrid: per email sent. **Why usage-based pricing works**: - Perfect alignment with value: customers who use more pay more. No one feels ripped off for paying for capacity they do not use. - Low friction to start: customers can try the product for pennies. No big commitment required. - Revenue grows with customer success: when customers succeed and use more of the product, you earn more. This creates a partnership dynamic rather than a zero-sum negotiation. - Easier product-led growth: self-serve customers can sign up with a credit card, start using the product, and grow organically. **Why usage-based pricing is hard**: - **Revenue unpredictability**: customers do not commit to specific spend, so you cannot predict next quarter's revenue as confidently as with subscription models. This is a real problem for SaaS financial planning. - **Customer bill shock**: customers whose usage spikes unexpectedly get surprise bills. This creates churn when customers feel they have lost cost control. - **Complex billing**: you need metering infrastructure to track usage accurately down to the unit. This is expensive to build and requires ongoing investment. - **Valuation pressure**: public market investors often prefer predictable recurring revenue to variable usage revenue. Companies have explored 'committed spend' models (customers commit to a minimum usage level in exchange for discounts) to bridge this gap. **When usage-based pricing works best**: - Infrastructure and developer tools where usage is the value. - Products with huge variance in customer size (a small hobby project vs a Fortune 500 deployment). - Products where charging per seat or per tier would either massively over-charge small users or leave money on the table with large users. **When it does NOT work**: - Products where value is experiential rather than metered (marketing tools, productivity apps). - Products with stable, predictable usage per customer. - B2B products where customers need budget predictability for procurement approval. **The trend**: most successful modern SaaS companies are moving toward hybrid models. Pure subscription with tiered features is still common, but adding usage-based components for high-variance metrics (storage, API calls, AI inference) captures additional revenue that flat pricing would miss. Snowflake, Datadog, and Twilio are examples of companies with usage-based cores that added subscription elements for predictability. Slack and HubSpot are examples of subscription cores that added usage-based overages (message retention, contact counts). BusinessIQ helps you decide whether to add usage-based elements to your subscription pricing — identifying which product dimensions have high variance across customers and would benefit from metered pricing.

Check Your Understanding

More Guides