How to Price a SaaS Product: Tiers, Anchors and Trials (2026)

Payment fees

Stripe’s pricing page lists 2.9% plus 30¢ per successful domestic card charge in the US, and Stripe Billing on pay-as-you-go adds 0.7% of billing volume. International cards add 1.5% and currency conversion adds 1%.

Worked example: on a $5 monthly plan, fees are $0.145 + $0.30 + $0.035, about $0.48, or 9.6% of the price. On $20 they are about $1.02 (5.1%). On $50 they are about $2.10 (4.2%). The fixed 30¢ is why very cheap monthly plans are hard to run. Annual billing helps: a $200 annual charge costs about $7.50 in fees, against about $12.24 for twelve $20 monthly charges.

Running costs per customer

Add what each active customer costs you each month: hosting, any AI or API calls, email sending and support time. If your app calls an AI model on every action, that cost can exceed the card fees, so estimate it from your own usage rather than guessing.

Write the total as “cost per customer per month”. Your lowest paid tier should cover it several times over, because some customers will cost far more than the average.

The four common models

Stripe’s documentation describes four recurring pricing models: flat rate (customers pick a tier and pay a fixed amount), per-seat (each user needs a license), tiered (the unit price changes with quantity or usage) and usage-based (customers pay for what they consume, for example a fixed fee plus overage, pay as you go, or credit burndown). Most small SaaS products use flat-rate tiers with a limit on one metric, because they are easy to explain on a pricing page.

A good metric passes three tests

It tracks value: a customer with more of it is getting more out of the product. It is easy to understand: a customer can predict next month’s bill. And it is hard to game: people can’t avoid it by sharing one login.

• Tool for personal trainers: number of active clients.
• Booking app for clinics: number of practitioners or locations.
• Report generator: reports per month, or connected data sources.
• Internal team tool: seats, if every person gets similar value.

When per-seat pricing backfires

Per-seat pricing punishes customers for inviting colleagues, which can slow adoption inside a company. If your product gets more useful when more people use it, consider charging on a different metric (projects, workspaces, records) and letting seats be generous.

Name each tier after a customer, not a metal

Start by writing one sentence per tier describing who it is for: “a solo trainer with up to 15 clients”, “a studio with several trainers”. Then decide the limit on your value metric and the one or two features that only that customer needs. Features that everyone needs belong in every tier.

Worked example

Suppose you sell workout-planning software to personal trainers. Interviews tell you a solo trainer bills clients around $60 a session and saves about two hours of admin a week with your tool. Pricing at $19 a month is a small fraction of the value of two hours, and it clears the fee floor comfortably. A studio plan at $49 a month covers up to five trainers and adds shared client lists and branded exports. The numbers are illustrative; the logic is what matters.

• Solo, $19/month: one trainer, up to 25 active clients.
• Studio, $49/month: up to five trainers, shared clients, branded PDFs.
• Annual option on both at ten times the monthly price (two months free).

Leave room to grow

Set limits so a successful customer eventually outgrows the lower tier. If nobody ever hits the limit, the higher tier only sells on features, which is a weaker reason to upgrade. Track how close customers get to each limit once you launch.

The research behind it

A 1982 paper by Joel Huber, John Payne and Christopher Puto in the Journal of Consumer Research showed that adding an option that is clearly worse than one existing option can shift people’s choices between the original two. This is often called the decoy or attraction effect. Later researchers have debated how strong it is in practice, so treat it as a nudge, not a lever.

Use a real top tier as the anchor

The honest way to use anchoring is to show a higher tier that some customers genuinely need, placed next to the plan you expect most people to choose, and to mark that middle plan as the common choice. Inventing a tier nobody should buy erodes trust when customers notice, and they often do.

Annual pricing is also an anchor

Showing “$190 a year (two months free)” beside “$19 a month” anchors the monthly price as the expensive way to pay. Annual plans also reduce fee drag and give you cash up front. The tradeoff is refunds: decide your annual refund policy before you offer it.

Length

If a new user gets value in their first session, 7 days is enough. If they need to enter data, invite clients or wait for a weekly cycle, 14 days is a safer default. Longer trials mostly delay the decision. Stripe allows trial periods of up to 730 days, so the limit is your judgment, not the tool.

Card up front or not

Asking for a card at sign-up filters for serious buyers and converts automatically, but fewer people start. Not asking means more trials and a deliberate conversion step at the end. Stripe supports both: without a card, you choose whether the subscription cancels, pauses or invoices when the trial ends.

Reminders are required, not optional

Stripe’s trial compliance guide says you must comply with card network rules for trials that convert to paid. Its hosted reminder email goes out 7 days before a trial ends, and the receipt shows trial information and a cancellation link. If you send your own reminder instead, you are responsible for meeting the same requirements.

Signals that the price is too low

Almost nobody mentions price as an objection, trial conversion is high, and customers describe the product as essential. Raising prices for new customers first, while keeping existing customers on their current price for a while, is the gentlest way to test this.

Signals that the price or packaging is wrong

Many trials end without converting and people cite price, or customers cluster on the lowest tier and never hit its limits. The second usually means the limits are too generous, not that the price is too high.

Measure the right numbers

Stripe’s Billing analytics defines trial conversion rate as trials that converted in the last 30 days divided by trials that ended in the last 30 days, and ARPU as total MRR divided by active subscribers. Watch both after any price change, alongside churn.

How much should I charge for my SaaS?

Start from the value the product creates for one customer (time saved, revenue gained) and price at a modest fraction of it, then check that the price clears your costs and payment fees several times over. Treat the result as a hypothesis and revisit it after your first few dozen paying customers.

How many pricing tiers should a new SaaS have?

Two paid tiers is a good start, each written for a distinct type of customer. Add a third when you clearly serve another group, such as agencies or larger teams, not just to fill a grid.

Should I offer annual pricing?

Usually yes. A common pattern is annual at ten times the monthly price. It lowers fee drag per dollar, gives you cash up front and reduces monthly churn. Decide your refund policy for annual plans before launch.

Is $9 a month too cheap?

Not automatically, but check the arithmetic. At Stripe’s US rates a $9 charge loses about $0.62 (6.9%) to card and billing fees, before hosting, AI and support costs. Cheap plans also tend to attract customers who need as much support as expensive ones.

How long should a SaaS free trial be?

Long enough for a typical new user to reach the point where the product proves its value. Seven days suits products that show value in one session; 14 days suits products that need setup or a weekly cycle.

How do I raise prices on existing customers?

Give clear notice, explain what has improved, and consider keeping current customers on their price for a set period. Many founders raise prices for new customers first to test demand before touching existing subscriptions.