How much should I charge for my product?

Written by Dante, who built Prep66, PrepAMT and Ancestorii

Charge well under what the problem costs your buyer, and well over what the product costs you to run. A buyer judges a price against what the problem costs them today and what else they could do about it. Y Combinator's talk on startup pricing gives a rule of thumb: the buyer should easily see the value as ten times the price. Say, as an example, your product saves a freelancer 5 hours a month and they bill $20 an hour: the problem costs them $100 a month. Under that rule a first price is $10 a month. Then test it: change the price and count who still pays.

What a price is compared with

Nobody reads a price on its own. A buyer holds it against what the problem costs them now, in money, hours or risk, and against the other ways out: another product, a spreadsheet, or doing nothing.

Two talks in Y Combinator's library use the same three figures: Startup pricing 101 by Kevin Hale, and the talk on YC's guide to business models by Aaron Epstein. Cost is what it costs you to serve a customer. Value is what the customer sees the product as worth. Price is what you charge, and it has to sit between the two. Kevin Hale calls the gap between price and cost your incentive to sell, and the gap between price and value the buyer's incentive to buy. Both say to set the price from the value, not from the cost.

The common pricing models

A model is how the buyer pays, not how much.

Model	How the buyer pays	What to know
One off	Once, and keeps the product.	Nothing recurs, so you need new buyers every month.
Subscription	Every month or year, until they cancel.	YC's guide to business models calls recurring revenue the most valuable revenue. It lasts only while customers stay, which is retention.
Usage	For what they use: per request, per record or by data used.	The same guide says it grows as your customers grow, and warns not to count it as recurring revenue.
Free plan	Nothing, for a limited part of the product.	Every free user costs you something to serve, so it earns its place only if some move up.
Free trial	Nothing for a set time, then the full price.	The buyer has to decide by a date. On the App Store, Apple's subscriptions page says to show clearly how long the trial lasts and the price billed once it is over.

Stripe's documentation splits a subscription further: a flat rate, a price per seat, tiers and usage. Two of them in its own words:

Customers choose a service tier (for example, Basic, Starter, or Enterprise) and pay a flat rate for it.
Charge customers based on their usage of your product or service.

Stripe, Recurring pricing models

How to work out a first price

What the problem costs the buyer. In money or hours. Ask your users what they hoped the product would do for them.
What one customer costs you. Hosting, payment fees, support. This is the floor, not the price.
What else they could do. The other products and what each charges, and what doing nothing costs.
A number between the floor and the value. Kevin Hale's rule of thumb in Startup pricing 101 is a price at a tenth of the value the buyer can easily see. It is one speaker's rule of thumb, not a law.
Stop at the right size of number. Aaron Epstein's test in the talk on YC's business models page: if you charge $10 and buyers would pay $100, change it. If they would pay $15 or $20, leave it.

How to test a price

Change it and count. Startup pricing 101 says this needs no chart, only a table: each price you tried, with its conversion rate, its sales and its revenue. The best row is the one that brings in the most money.

Price (an example)	Visitors	Buyers	Revenue a month
$10 a month	200	10	$100
$20 a month	200	7	$140

In this made up case the higher price loses 3 buyers and still brings in more. But 10 against 7 is too few to be sure of, and the conversion rate page says why.

Both talks say to test upwards. Kevin Hale's rule of thumb, which he says matters most when selling to businesses, is to raise the price 5% at a time until you are losing 20% of deals. Aaron Epstein says to keep raising it until customers complain and still pay.

The mistakes a new founder makes with it

Not charging. The talk on YC's business models page calls this the most common mistake Y Combinator sees founders make. Charging is how you find out whether anyone will pay at all.
Charging too little. Both talks say most startups undercharge. Kevin Hale calls it the number one piece of advice Y Combinator gives startups to fix their pricing.
A complicated pricing page. Aaron Epstein's last point is to keep it simple, because complexity adds friction between a customer and paying you.
A free plan nobody leaves. People sign up but nobody pays tells one case: a free plan whose users never once reached the checkout.

Where pricing matters

What you can afford to spend. The price, and how long a customer stays, decide what a customer is worth to you, your lifetime value. That limits what you can spend to win one, your customer acquisition cost.
Which channels you can use. In Startup pricing 101 Kevin Hale says a low priced product that people buy on their own cannot carry a sales team or much ad spend, so almost all its marketing has to be inbound. Check that before testing Google Ads or Meta Ads.

How Sami uses it

The honest read AskSami writes lists the competitors it found by live web searches, each with what it charges and a link to its own page. Your price is then set beside real ones.

The price you tell Sami and the payments your own payment webhook reports become dated records. Where there is no data on what a price did, Sami says so and never invents a figure. It does not change your price. You do.

Questions people ask

Should I be cheaper than my competitors?

Not as your only difference. In the talk on YC's business models page Aaron Epstein says a lower price is not a lasting advantage, because a larger competitor can go lower than you can afford to. He adds that buyers read a price as a sign of what a product is worth.

Can I raise my prices after I launch?

Yes. In the talk on YC's business models page Aaron Epstein says pricing is not permanent. He gives two ways to raise it: keep existing customers on their old price and charge new ones more, or give notice before the rise.

Is a free trial better than a free plan?

Neither is right by default. A trial ends, so the buyer has to decide. A free plan does not, so some people never do. Decide from what your free users do, and count payments, not signups.

Should I charge monthly or one time?

Match how the product is used. Used month after month, a subscription fits. One job done once, a one off price fits, and you will need new buyers every month.

Sources

Y Combinator, Startup pricing 101, a talk by Kevin Hale, page created 6 September 2019. Read on 8 October 2026.
Y Combinator, YC guide to business models, with a talk by Aaron Epstein, page created 15 December 2022. Read on 8 October 2026.
Apple, Auto renewable subscriptions. Read on 8 October 2026.
Stripe, Recurring pricing models. Read on 8 October 2026.

Checked on 17 October 2026. Every rule, price and figure on this page was read on the page it links to, on the day shown beside it above. They change, so open the source before you rely on one.