How to choose a SaaS pricing model: subscription, freemium or usage-based

Choose a SaaS pricing model by how customers get value: if value grows with headcount, charge per seat; if it grows with usage, charge for consumption; if the product is simple, use flat plans. A free tier or trial is not a pricing model but an acquisition tactic, and it should be modelled separately.

Here are the main models, how to pick a value metric and how to test your price before it is hard-coded.

Which SaaS pricing models exist?

There are a handful of core models, and most products end up combining them: tiers with different limits, extra charges per seat or for overage. The choice shapes not only revenue but also what you will have to build in billing.

Model	How it works	Who it suits	Development complexity
Flat subscription	One price for access	Simple products with one use case	Low
Tiers	Several plans with different limits and features	Most B2B and B2C SaaS	Medium
Per seat	Price × number of the customer's users	Collaboration tools	Medium
Usage-based	Pay for volume: requests, gigabytes, documents	APIs, infrastructure, AI services	High: metering, limits, invoices
Hybrid	Base plan + overage charges	Products with uneven consumption	High

What is a value metric and how do you choose one?

A value metric is the unit customers pay for, which grows with the value they get: users, projects, emails sent, documents processed. A good metric is clear to customers, grows with their business and is simple to measure.

How to test a candidate metric:

1. Customers understand it without explanation and can predict their bill.
2. The more value customers get, the higher the metric.
3. It does not penalise healthy usage — for example, inviting colleagues, if that makes the product more valuable.
4. It is easy and reliable to measure in the system.

Example: an email marketing service is logically priced by contacts or emails, not by the number of users in the dashboard — the customer's value is in reach, not in how many staff click the button.

Photo: Unsplash

Freemium or a free trial?

Freemium fits when free users bring in new ones or the product grows more valuable with its audience; a free trial fits when value becomes clear within a few days of active use. Both cost money: free users consume infrastructure and support.

How to choose:

• freemium — a mass-market product, low cost per user, viral growth; the free tier should be useful but hit a clear limit;
• free trial — a B2B product that needs setup; 14–30 days is usually enough for a customer to reach the key action;
• sales-led demo — a complex, high-price product where customers need help with implementation.

Model the economics of a free tier or trial together with the rest of your unit economics — see SaaS unit economics.

How does the pricing model affect development?

The pricing model determines what you must build in billing and in the product: usage metering, limits, plan changes, proration and invoicing. A complex model at launch can cost more than it earns.

What you may need, depending on the model:

• metering the value metric accurately enough for customers to trust;
• enforcing limits in the product with clear messages when they are reached;
• plan changes with proration for the unused period;
• invoices, receipts and documents for business customers;
• handling failed charges and retries.

Plans belong in configuration, not code

Define plans in configuration rather than hard-coding them:

{
  "plans": [
    { "id": "start",  "price": 0,   "limits": { "projects": 1,  "seats": 2 } },
    { "id": "team",   "price": 39,  "limits": { "projects": 10, "seats": 10 } },
    { "id": "growth", "price": 129, "limits": { "projects": null, "seats": 50 } }
  ]
}

That way prices and limits can change without a release, and existing customers keep their terms.

How do you test a price before building billing?

Test the price before building complex billing: through customer conversations, a pricing page with a sign-up button and the first sales handled manually by invoice. Automate the model that has already been proven by real payments.

The order we recommend:

1. Choose a value metric and two or three price hypotheses.
2. Test them in interviews: which price customers find expensive, cheap or reasonable.
3. Launch a pricing page and accept first payments by invoice or a simple payment widget.
4. Automate billing once the model is proven, and measure conversion and churn by plan.

We covered how to track the effect of price changes in product metrics after launch.

In short

• Choose the model by how customers get value: flat, tiers, per seat, usage-based or hybrid.
• A value metric must be clear, grow with value and be easy to measure.
• Freemium suits viral mass-market products; trials suit B2B products that need setup.
• Complex billing is expensive: start simple and define plans in configuration.
• Test the price with real customers before automating billing.