Skip to main content

Trends, tech news, and tool comparisons.

How to price a SaaS product at the early stage: a decision rule

· 6 min read
Dealpuffin
Proof points and live coupons for early adopters

To price an early-stage SaaS product, choose one value metric that grows with the benefit a customer gets, anchor the price to the cost of the problem it replaces, offer two or three simple tiers, and test the number in conversations with real buyers before you build the billing. Expect to change the price in the first year, and keep the first customers' price fixed for a stated period so changing it stays fair.

This guide is a method, not a benchmark. Pricing data differs a lot between markets, and I cite only vendor documentation for the facts below, read on 9 November 2026.

Step 1: pick the value metric​

The value metric is what you count to charge: seats, projects, events, API calls, tracked pages. Stripe's documentation on recurring pricing models lists the common shapes: flat rate (a tier at a fixed price), per-seat, tiered (price per unit changes with quantity), and usage-based.

A useful test: if a customer gets twice the benefit, would they accept paying more? If yes, that unit is a candidate.

  • Per-seat suits tools that each person on a team uses separately.
  • Flat tiers suit products where usage is hard to meter or customers dislike surprise bills.
  • Usage-based suits products whose cost to you rises with use, such as AI calls or data processing. It brings more engineering work: you need metering, and invoices that vary. The post on usage-based billing after Metronome and Orb covers the tools.

If you are unsure, start with flat tiers and a usage limit on the lowest one. They are the easiest to explain and the easiest to build.

Step 2: anchor to the problem you replace​

List what the customer does today without you: hours of manual work, another tool's subscription, an agency, or nothing at all. Put a rough monthly cost on that. A price well below it is easy to justify.

Worked example: a tool that generates compliance reports saves a founder about three hours a month. If their time is worth $50 an hour, the problem costs $150 a month. A $29 plan is a clear yes, and a $149 plan asks for the whole saving. A rule of thumb I use is to start near a fifth of the cost removed, then adjust from the answers you get. That is my heuristic, not a published figure.

If the honest answer is "nothing", the product may be a nice-to-have. Many first users will pay little for those, so expect a slow start rather than a pricing problem.

Step 3: set two or three tiers​

Three tiers is the common pattern, and it helps because each plan answers a question for a different buyer.

TierWho it is forWhat changes
StarterA solo user trying it for real workLow limit on the value metric; core features
TeamA small group using it every weekHigher limit; collaboration; export
ScaleA company with a process around itHighest limits; support terms; invoicing

Keep the differences in limits and support, not in whether the product basically works. Hiding the main benefit behind a higher tier tells early adopters you are playing games with them.

Avoid a free forever plan until you can measure what it costs you and what share upgrades. A time-limited trial or a generous first tier has a built-in end.

Step 4: check the price against fees​

Low prices carry a fixed cost per transaction. Using the published rates read today, a $9 monthly plan loses about $0.56 to Stripe's standard US card fee of 2.9% + 30¢ (6.2%), or $0.95 to the 5% + 50¢ that Paddle and Polar's free plan charge (10.6%), before taxes and payout fees. At $29 the same two rates take $1.14 and $1.95 (3.9% and 6.7%).

Higher fees are not automatically worse: the merchant-of-record providers also handle sales tax and VAT, which Stripe's basic payments leave to you. The post on Stripe vs Paddle vs Polar fees works through the trade, and merchant of record explained covers what the term means. Very low prices put a fixed fee on every payment, so a $5 plan needs a real reason to exist.

Step 5: test before you build​

You can test a price without a billing system.

  1. Show the pricing page to 10 to 20 people who fit your target customer.
  2. Ask "which plan would you pick, and what would make you not buy?" rather than "would you pay $X?". The second question gets polite answers.
  3. Count the answers. If almost everyone picks your cheapest plan, your middle plan is probably overpriced or poorly defined. If nobody objects to the price, it is probably too low.
  4. Ask for a card or a signed intent from a few. A pre-order or a paid pilot says more than approval.

Step 6: plan for the first change​

You will adjust the price. Decide the rules ahead of time and write them on the pricing page: early customers keep their price for 12 months, or for as long as they stay subscribed, and later customers pay the new price. Changing a price quietly is the fastest way to lose the users who helped you most.

For more on finding those first buyers, see how to get your first 100 users for a SaaS.

Frequently asked questions​

How much should I charge for my first SaaS product?​

Start from the cost of the problem you remove, then price near a fifth of that and adjust from buyer feedback. There is no standard number; the rule is a heuristic you should test.

Should an early-stage SaaS use usage-based pricing?​

Only if your own cost rises with use or customers clearly think in units, such as AI calls. Usage-based pricing needs metering and produces varying invoices, so flat tiers are simpler to start with.

Should I offer a free plan?​

Only if you can measure what it costs and what share of users upgrade. A time-limited trial gives you the same learning with a clear end.

Can I raise prices after launch?​

Yes, and most founders do. State the policy for early customers up front, such as keeping their price for 12 months, so a change does not break trust.