SaaS pricing models compared, with worked examples for each

What a pricing model actually decides

A pricing model answers one question: what does the customer pay more for as they get more value? Price levels, discounts and plan names come later. The model is the shape of the curve.

Choose a model where the curve follows value and revenue grows with your customers without anyone renegotiating. Choose a model where it does not, and you either overcharge small customers or undercharge your largest ones. Usually both.

This guide covers the six common SaaS pricing models, with a hypothetical example for each, then compares them side by side.

Flat rate pricing

One product, one price, everything included. A project tool at 49 USD per month for any team size is flat rate pricing.

Worked example. Take a hypothetical tool with 500 customers at 49 USD. Revenue is 24,500 USD a month. Suppose 60 of those customers are agencies with 30 or more users who would happily pay 200 USD. The business leaves 60 x 151 = 9,060 USD a month on the table, more than a third of current revenue.

When it fits. Narrow products with homogeneous customers, very early products that need the simplest possible message, and tools used by individuals.

When it fails. As soon as customers differ widely in size or usage. The flat price is either too high for the smallest or far too low for the largest.

Tiered pricing

Two to four plans that differ by features, limits or both. This is the default for most SaaS companies because it serves different customer sizes without complex billing.

Worked example. The same tool moves to three tiers: Basic at 29 USD, Team at 79 USD, Business at 199 USD. Of the 500 customers, suppose 200 choose Basic, 240 choose Team and 60 choose Business.

Revenue rises from 24,500 to 36,700 USD, about 50 percent, with the same customers. In reality some customers would leave or downgrade, so treat this as the upper bound of a range, not a forecast.

When it fits. Most B2B SaaS with customers of different sizes and needs.

When it fails. When tiers are built from internal feature lists instead of customer segments. If nobody can say who the middle plan is for, it is the wrong middle plan.

Per seat pricing

The customer pays per user. Collaboration and communication tools have used it for decades because value tends to grow with each person invited.

Worked example. A hypothetical design review tool charges 12 USD per seat. A customer with 8 seats pays 96 USD a month, one with 120 seats pays 1,440 USD. The model scales cleanly with team size.

The risk. If the product makes each user more productive, customers need fewer seats over time. A tool that automates half of a support agent's work gives its customers a reason to cut seats. AI features make this risk sharper because they substitute for people.

When it fits. Products where every user actively works in the tool and value per user is roughly constant.

When it fails. Products used by one person on behalf of many, products where the output matters more than the number of people, and any product that replaces human effort.

Usage based pricing

The customer pays for what they consume: API calls, messages sent, records processed, compute hours, documents generated. It aligns price with value very closely when usage tracks value.

Worked example. A hypothetical transcription service charges 0.02 USD per minute. A small customer transcribing 3,000 minutes pays 60 USD a month. A large customer transcribing 400,000 minutes pays 8,000 USD. Nobody overpays and the largest customers pay the most.

The trade-off. Revenue becomes less predictable for you and costs become less predictable for customers. Finance teams at larger customers often dislike bills they cannot forecast. Our detailed guide to usage based pricing covers metering, overages and how to keep bills predictable.

When it fits. Infrastructure, APIs, communication platforms and AI products where cost to serve also scales with usage.

Hybrid pricing

A base subscription plus a variable component. Most mature SaaS companies end up here. Typical combinations:

• Platform fee plus per seat
• Tiered plans with included usage plus overage charges
• Per seat plus paid add-ons
• Annual commitment with a usage allowance, billed for anything above

Worked example. A hypothetical email platform charges 99 USD a month including 50,000 sends, then 1.50 USD per extra 1,000 sends. A customer sending 180,000 emails pays 99 + (130 x 1.50) = 294 USD. A customer sending 30,000 pays 99 USD.

The base fee keeps revenue predictable. The variable part captures growth without forcing a plan change. That combination is why hybrid models are so common.

When it fits. Products with a clear core value plus a usage dimension that varies widely between customers.

When it fails. When the bill becomes too complex to explain on the pricing page in two lines.

Freemium versus reverse trial

These are acquisition models rather than pricing models, but they shape how pricing performs, so they belong in the comparison.

Freemium

A limited version of the product is free with no end date. Users upgrade when they hit a limit or need a paid feature.

Worked example. A hypothetical note tool has 50,000 free users. If 2 percent upgrade to a 10 USD plan, it earns 10,000 USD a month. If the free tier is too generous and only 0.5 percent upgrade, revenue falls to 2,500 USD while hosting costs for 50,000 users stay the same.

Freemium works when the free product spreads the paid one, for example through shared documents or invitations, and when serving free users costs very little. It works badly when the free tier already covers what most customers need.

Free trial

Full access for a fixed period, then payment or loss of access. It is simple and puts a clear decision point in front of every user. The main design choice is length, which should match time to value. A tool that shows its worth in a day does not need a 30 day trial.

Reverse trial

New users start on the full paid plan for a limited time, then drop to a limited tier unless they pay. They experience the paid value first, and the downgrade makes that value concrete.

Worked example. A hypothetical analytics tool compares two approaches with 1,000 signups each. With a classic limited free tier, 3 percent upgrade within 60 days, giving 30 paying customers. With a 14 day reverse trial, suppose 5 percent upgrade. That is 50 customers, a two thirds increase from the same signups. Your own numbers will differ. The point is that the reverse trial is cheap to test against your current approach.

The models side by side

No model wins on every column. The decision is which trade-off suits your customers and your stage.

How to choose between them

Three questions narrow the choice quickly.

1. What grows when a customer gets more value? Users, usage, records, revenue they process, locations? That is your candidate value metric.
2. Does your cost to serve grow with that same thing? If yes, usage or hybrid pricing protects your margin.
3. How much do your customers differ in size? A tenfold spread calls for tiers at least. A hundredfold spread calls for a usage or seat component.

Our SaaS pricing strategy guide goes further into choosing the value metric and packaging tiers around it.

Changing models without losing customers

Switching from flat rate to tiered, or from per seat to hybrid, is a bigger change than raising a price. A few practices reduce the risk:

• Apply the new model to new customers first and measure conversion for several weeks.
• Give existing customers a clear period at their current price before moving them.
• Show customers what they would pay under the new model before the switch.
• Make sure the smallest existing customers do not face a large increase overnight.

Measuring each step against a baseline matters more here than anywhere else. A change to the model affects conversion, expansion and churn at once, and only a measured comparison separates them. This is the kind of question pricing optimization software is built to answer.

A worked comparison on one customer base

Seeing all the models applied to the same customers makes the trade-offs concrete. Take a hypothetical customer base of three groups for a reporting tool:

• 300 small customers with 2 users and 1,000 reports a month
• 150 mid-size customers with 10 users and 8,000 reports a month
• 30 large customers with 40 users and 60,000 reports a month

The flat price undercharges large customers badly. Pure usage pricing makes small customers very cheap and reduces total revenue. Per seat and hybrid land close together, but the hybrid model keeps working if the product automates more and customers need fewer seats. None of these figures are a forecast. The point is that the same customers produce very different revenue depending on the shape of the price.

Where to go from here

If you are not sure your current model fits, start with your billing data. Sort customers by revenue and by usage or seats. If the top tenth of customers by usage pay barely more than the bottom tenth, the model is leaving money behind.

Monetisation is a monetization platform that reads your billing and product data read-only and shows where your pricing model leaves revenue uncaptured, each opportunity with an estimated range, confidence and the method used. Estimates are estimates, not a guarantee of results. If you want to see it on your own business, run a sample audit on your site, and the SaaS use case page shows what it looks at for subscription companies.