Price Testing on a Landing Page

What is price testing on a landing page?

Price testing on a landing page means showing real visitors a real price and measuring what they do: click “Buy”, start checkout, or pay. You run two or three prices, either side by side for different visitors or one after another over time, and compare the share of visitors who take the paying step at each price.

The difference from a survey is the whole point. In a Van Westendorp survey, people tell you what feels fair. On a landing page with a checkout button, they show you, with a credit card in hand and no one watching. Behaviour at checkout is the most honest pricing data an early founder can get.

When should you use it?

• Use it once your landing page test shows people want the thing, and you need to choose between prices.
• Fits best for B2C products, templates, courses, apps and small-business SaaS where people buy online without a sales call.
• Skip it for enterprise deals and anything sold through conversations. There, test prices across sales calls with a one-page offer sheet.
• Skip it if you can’t get at least 100 visitors per price in a month. Tiny samples produce noise you will mistake for insight.

How do you do it, step by step?

1. Pick 2 or 3 prices that are far apart. Testing 19 against 21 euros tells you nothing at early-stage traffic. Test 19, 29 and 49. Take the range from your survey or from a competitor pricing map.

2. Choose split or sequential. Split: make 2 or 3 copies of your landing page, change only the price, and send equal traffic to each copy, say through two ad sets with identical targeting or two links shared in comparable places. Sequential: show price A for one week, price B the next. Split is cleaner; sequential is simpler and fine for organic traffic, as long as nothing else changes between weeks (no launch, no holiday, no newsletter mention).

3. Decide what counts as the signal before you start. Ranked from weakest to strongest: click on “Buy”, entering the checkout page, completed payment. Write it down: “Winner = highest revenue per 100 visitors, measured by completed payments.” Revenue per visitor matters more than conversion rate alone; a higher price with slightly fewer buyers can still win.

4. Wire up real checkout. Create one Stripe payment link per price and put each on its own page version. If you are not ready to deliver, use a fake-door checkout instead: the button leads to a page saying “We’re not taking orders yet. Leave your email and we’ll hold this price for you.” Read the fake door test guide for the honest version of this.

5. Track visitors and clicks per version. Install a free analytics tool such as PostHog or Google Analytics on each page (see analytics setup). Log daily in a sheet: visitors, buy clicks, checkouts started, payments, revenue.

6. Follow fair rules while you test. Our rules: anyone who paid the higher price gets the difference refunded once you settle on a lower one; anyone who asks why a friend saw a different price gets a straight answer; every buyer gets the same product. Also check the consumer rules in the countries you sell to, especially around displaying prices and discounts.

7. Run to a fixed stop, then decide. Our rule of thumb: stop at 2 weeks or 150 visitors per price, whichever comes later. Don’t peek and stop early when one price is ahead after 30 visits.

What does a worked example look like?

A real case first. In a post on the Buffer blog dated 16 February 2011, co-founder Joel Gascoigne described adding a pricing page between his landing page and the email sign-up, before the product existed. Clicking a plan led to the sign-up, so the extra step showed which plan people picked and whether they still wanted in. He reports that people kept clicking through and a small number chose paid plans, which convinced him to start building. You can read his original write-up on Buffer’s site.

Now a hypothetical walk-through with numbers. Say Zeynep sells a client-management template pack for freelance photographers. She builds 3 identical Carrd pages at 19, 29 and 49 euros, each with its own Stripe payment link, and splits Instagram ad traffic evenly. After 2 weeks:

The 19 euro page converts best, but 29 euros earns the most per visitor. She picks 29, refunds 20 euros to each of the 3 people who paid 49, and plans a 39 euro bundle test for next month.

Which free tools help?

What mistakes should you avoid?

• Changing two things at once. New headline plus new price means you can’t tell which one moved the number.
• Mixing traffic sources. Newsletter readers and cold ad clicks behave like two different audiences. Give each price the same mix, or run them sequentially from the same source.
• Judging by conversion rate alone. The cheapest price tends to convert best. Compare revenue per visitor.
• Stopping as soon as a winner appears. Early leads flip all the time at small numbers. Set your stop rule in advance.
• Keeping the extra money quietly. If you lower your price after the test, refund or reward the people who paid more. Your first buyers talk.

How do you know it worked?

You have an answer when each price got at least 100 to 150 visitors from the same kind of traffic (our rule of thumb) and one price leads on revenue per 100 visitors by 20% or more.

Go: a clear winner at a price that covers your costs with room to spare. Make it your launch price. Rerun: no price leads by 20%. Test prices further apart or get more traffic. No-go: fewer than 1% of visitors reach checkout at any price. The problem sits higher up the page: message, audience or offer.

What should you do next?

If you have enough traffic later, move to proper A/B testing of the pricing page itself. Start taking real orders at your winning price with a payment link test. If you want someone to sanity-check your test setup, the F/MS startup game walks you through validation with an AI co-founder.

FAQ

Is it ethical to show different prices to different people?

It is fair when the test is short, every buyer gets the same product, and you refund or compensate anyone who paid more than the price you settle on. Be honest when someone asks. Avoid targeting higher prices at people based on personal data.

How much traffic do I need for a price test?

Our rule of thumb is 100 to 150 visitors per price from the same source, over at least 2 weeks. With less, run prices one after another and treat the result as a direction, never a final answer.

Should I test prices with a fake checkout?

A fake-door checkout measures intent and works before you can deliver. Real payments are the stronger signal. If you use a fake door, tell people right away that orders are not open yet and offer to hold the price for them.

Updated 26 October 2026 · Written by the Fe/male Switch team

Mark this method as done