How to price your first product (without guessing)

A practical method for pricing a first product: anchor on what customers do today, pick a simple structure, ask for money early and test price with three metrics.

Most first-time founders set the price last, in a hurry, by looking at a competitor or choosing a number that feels safe. That is guessing. A price is a statement about how much value you think you deliver, and it is also the cleanest signal you will ever get about whether someone wants what you made.

This guide gives you a method you can follow without a spreadsheet: reason from the customer’s side, choose a simple structure, ask for money early, and test price with three plain numbers. I do not claim any particular price works. What I can give you is a way to find out for your product.

The short answer. Find what your customer does today and what it costs them. Price clearly below the value you save or create. Keep one simple price at first, ask real people to pay before you build everything, and judge the result with visits, sign-ups that pay and returns. Change one thing at a time.

Start with value, not cost

Your costs set a floor: you cannot sell below what it takes to run the product for a customer. They do not tell you the price. The price lives between two numbers:

  • The floor, which is what it costs you to serve one customer.
  • The ceiling, which is what the result is worth to them.

Your job is to find the ceiling. Ask: what changes for the customer if this works? Do they save hours? Win more clients? Avoid a mistake that costs money? Write that down in money or time, then convert time to money using what their time is worth to them. A price that is a small fraction of a clear benefit is easy to say yes to. A price that is close to the whole benefit is a hard sell, even if the product is good.

Anchor on what they use today

People never evaluate your price in a vacuum. They compare it to what they already do. That comparison is your anchor, and you should find it deliberately.

Common anchors:

  • A tool they pay for now (a spreadsheet add-on, an app subscription).
  • A person they pay (an assistant, a freelancer, a bookkeeper).
  • Their own time, spent on a manual routine.
  • Doing nothing, with a cost of lost orders or mistakes.

The last two are the ones founders miss. If the real alternative is “I do it myself on Sundays”, your competitor is a Sunday, not another app. Ask customers what they use today, what it costs them, and what annoys them most. Then your price can be framed against that: cheaper than the person, faster than the Sunday, or clearly better than the tool.

Pick a simple structure

Complex pricing is a way of avoiding the decision. Start with one of these.

Structure How it works Good when Watch out for
One price, one-off Customer pays once for the product or result A single deliverable, template or report No repeat income; support costs continue after the sale
One price, monthly Customer pays every month while using it Ongoing use, ongoing costs for you People cancel if they stop seeing value, so watch returns
Free trial, then paid Customer uses it free for a short time, then pays Product that shows its value quickly Many trial users never pay; trial needs a clear end
Pre-order Customer pays before it is finished You want proof of demand before building You owe delivery; be exact about the date and refund
Usage-based Customer pays by how much they use Your costs rise with use Hard to predict bills, which makes buyers nervous

For a first product, one price with one plan is usually easiest to explain, easiest to test and hardest to get wrong. Add tiers only when different groups clearly want different things, and you have seen it in real conversations.

Ask for money early

Praise is cheap. “Looks great, I would use that” costs the person nothing. A request for money is the honest test.

You can ask long before the product is finished:

  1. Describe the result and the price in a short page or message.
  2. Offer a pre-order or a paid pilot, with a stated price, a clear delivery date and a refund promise you can keep.
  3. Count real commitments, not compliments: money paid, a signed note of intent, or a booked time to start.

Be careful and honest. Say exactly what exists today and what does not. Never take money for something you cannot deliver, and do not use pressure or false deadlines. If no one will pay even a small amount, that is a result. It is much cheaper to learn it now than after months of building. If you are still at the stage of finding the problem, the interview questions guide shows how to talk to customers first.

Interview questions about price

You can learn a lot about price without ever naming yours. Ask about the past and the present, not what people think they might do.

  • “What do you use today to deal with this?”
  • “What does that cost you per month, in money or hours?”
  • “When did you last pay for something to solve it? What was it, and what made you decide?”
  • “What is the cost if this is not solved?”

Then, once you have listened, test a number: “I am planning to charge about X a month. Does that feel low, fair or too much? What would you compare it to?” Do not follow up with a discount. Listen to the comparison they name, because that is your anchor.

Test price with three metrics

A price test is a small experiment. Show a price to real visitors and watch what they do. The same three numbers from the launch guide apply, with money added:

  1. Visits. How many people saw the price.
  2. Activation. How many took the step that matters: started a trial, paid, or pre-ordered.
  3. Return. How many came back, or kept paying, a week or a month later.

If visits are fine and activation is low, the price or the message is the problem. If activation is fine and return is low, people paid and then did not find value. The metrics guide explains how to read these for a small product.

Rules for testing:

  • Change one thing at a time, and give each version enough visitors to mean something.
  • Show different prices to different visitors, not to the same person at different times.
  • Write the result down before you decide, including what you expected. Small samples are noisy, so treat early results as hints, not proof.

A worked example (invented for illustration)

This is a made-up business with round numbers, to show the reasoning. It does not describe real customers or real prices.

Imagine a simple booking tool for independent dog walkers.

  • Anchor: today they take bookings by text and a paper notebook. An assistant would cost far more than they would pay, so their real alternative is their own Sunday evenings, plus the odd double booking.
  • Value: suppose one double booking or forgotten request costs a walker about 30 in lost income, and it happens a couple of times a month. That is roughly 60 a month of avoidable loss, plus a few hours of messaging.
  • Floor: suppose serving one walker costs 3 a month.
  • Candidate price: 15 a month, clearly below the 60 of loss, well above the 3 floor.
  • Structure: one plan, monthly, with a 14-day free trial to see the first week of bookings.
  • Ask early: before building everything, show 20 walkers a one-page description with the price and a pre-order button for the first month.
  • Read it: if 2 pay, that is a signal to build. If 0 pay and several say “too much”, ask what they compare it to and try 10 with a smaller first version.

The numbers are invented, so use the method, not the figures.

Common mistakes

  • Too cheap. A very low price can signal low value, attract people who will not care, and leave no room to serve them. If everyone says yes instantly, you may be too low.
  • Many tiers. Three or four plans before you have ten customers is guesswork with extra columns. Start with one.
  • Discounts as a reflex. Cutting the price when someone hesitates teaches people to hesitate. If you discount, make it a clear, time-limited offer with a reason.

Payments safety

Once money is involved, a mistake costs real people real money.

  • Keep payment keys in a secret store, never in the code, never pasted into a chat or a prompt. A key in a prompt should be treated as leaked.
  • Require approval before anything touches money: charging, refunding, changing a price or sending a payment link. A small pause is cheap.
  • Use a payment provider’s hosted checkout rather than building your own card form, so card details never pass through your product. Building a small shop without card handling shows the idea.
  • Say clearly what happens on refunds and cancellations. The no-code security checklist covers the rest.

Where equipe fits

equipe is an AI startup team for people who cannot code: a PM, engineers, a designer, QA and a data analyst. It can build the page that shows your price and a pre-order step, and it can count visits and sign-ups without cookies. Anything that goes live, costs money or reaches customers waits for your approval.

As an honest example of a structure, equipe’s own pricing is a two-part model: a flat Team plan at $19 a month or $190 a year, plus AI work paid from credits at the model list price plus a margin, shown to the cent per task, with domains at cost. We use it as an illustration, not as proof it suits your product. The pricing page has the full terms, and the cost article walks through what running a small app costs. We do not publish a “typical monthly spend”, because we do not have real usage data yet. For a 30-day path from idea to first payments, see the 30-day plan.

A pricing checklist

  • I know what my customer does today and what it costs them
  • I wrote the value in money or time
  • My price is clearly below that value and above my cost to serve
  • One plan, one price, easy to explain in a sentence
  • I asked at least ten real prospects, not friends
  • I counted payments or pre-orders, not compliments
  • Keys are in a secret store and money actions need approval
  • I changed one thing, measured, and wrote down the result

Frequently asked questions

How do I decide the price of my first product?

Start from the customer, not your costs. Find what they do today to solve the problem, what that costs them in money or time, and what a fixed result is worth. Then set a price clearly below the value, test it with real people, and adjust from what they do.

Should I charge before the product is finished?

Asking early is the strongest test of demand, because a promise to pay is more honest than praise. You can offer a pre-order at a stated price with a clear refund promise. Be exact about what is built and what is not, and never take money for something you cannot deliver.

Is it better to charge monthly or one time?

Monthly suits a product people use again and again, because your costs repeat too. A one-time price suits a single result, like a template or a report. Choose the one that matches how the customer gets value, and keep it to one price at first.

How does equipe price its own product?

equipe uses a two-part model: a flat Team plan at $19 a month or $190 a year, plus AI work paid from credits at the model list price plus a margin, shown to the cent per task. That is one example of a structure, not proof that this model fits yours.

Got an idea? Tell the PM.

Your AI startup team builds, launches and keeps improving it. You approve what goes live. You do not need to know how to code.

Free to start · $5 credit · No card needed

Keep reading