Klarna Costs 5.99% Where a Card Costs 2.9%: The Conversion Lift BNPL Has to Earn

Buy now pay later costs a merchant more than a card, and both large providers publish exactly how much. On Stripe's US pricing (opened 29 July 2026) Klarna is 5.99% plus 30 cents per successful transaction against 2.9% plus 30 cents for a domestic card. On PayPal's US merchant fee schedule (opened 29 July 2026) Pay Later is 4.99% plus the fixed fee against 3.49% plus the same fixed fee for standard PayPal Checkout, where the USD fixed fee is 0.49. Because the fixed component is identical on both sides of each comparison, the premium is not a moving target: it is 3.09 percentage points on Stripe and 1.50 on PayPal, on every order, whatever the basket size. The only question left is whether the button brings you enough additional business to cover it, and that has a formula.

What the premium costs at three basket sizes

Method: our arithmetic applying the two published US schedules above, rounded to the cent, for a single successful domestic transaction with no refund, no currency conversion and no dispute. The extra column is always 3.09% of the order on Stripe and 1.50% on PayPal, because the 30 cent and 49 cent fixed fees cancel. That constancy is the useful part: you do not need a table at all once you know your provider's rate gap.

Two costs sit outside it. Stripe charges 15 dollars per dispute received, and PayPal charges 15 dollars per claim under its standard dispute fee, rising to 30 dollars under its High Volume Dispute fee. Neither schedule we opened offers a discount on disputes for instalment orders, so assume your existing dispute exposure applies unchanged.

The formula that decides it

The mistake is comparing the fee to nothing. Most of the orders that go through the instalment button would have gone through anyway, on a card, so you pay the premium on those for no gain. The button only earns its place if enough genuinely additional orders arrive.

Write it as: required lift = (share of orders that shift to BNPL x rate premium) / gross margin, where the lift is expressed as extra orders per hundred. Basket size drops out of the equation for the reason above, which makes it unusually easy to run.

Worked, on Stripe with Klarna at a 3.09 point premium. Say 20% of your orders move to the instalment button and you keep 40% gross margin:

• Cost on 100 orders of $80: 20 orders x $2.47 = $49.40.
• Contribution from one additional $80 order at 40% margin: $32.00.
• $49.40 / $32.00 = 1.54, so you need roughly 1.6 extra orders per 100, a lift of about 1.6 percentage points.

Change one input and it moves fast. At 30% margin the same shift needs about 2.1 extra orders per 100. If 40% of orders move to BNPL rather than 20%, the 40% margin case needs about 3.1. On PayPal's narrower 1.50 point gap, the first case falls to about 0.75 extra orders per 100, which is a much easier bar and a good reason to check which provider you are already on before assuming BNPL is expensive.

Nothing in this article tells you what lift to expect, on purpose. Every published uplift figure we could find comes from a company selling the product, without a methodology we could inspect, so we are not repeating any of them. Your own number is measurable in 60 days and is the only one that matters.

Measuring your own lift in 60 days

1. Record a baseline for four weeks before switching anything on. Orders per week, conversion rate on the checkout step, average order value and gross margin. Without this the test has no meaning, and a seasonal week will fool you.
2. Turn the button on and leave everything else alone. No new ad campaign, no pricing change, no free shipping threshold in the same window.
3. After four weeks, pull three numbers: share of orders paid by instalment, conversion rate on the checkout step, and average order value on instalment orders against card orders.
4. Put the observed share into the formula and compare the required lift with your actual change in conversion. If conversion did not move, you have bought a more expensive way to take the same orders.
5. Check the refund column too. Look at whether instalment orders are returned more often than card orders in the same period. On a 3.09 point premium, a materially higher return rate can turn a small conversion win into a loss.

If you sell in the UK, the rules changed this month

The Financial Conduct Authority states plainly: "We started regulating Deferred Payment Credit (DPC), often known as Buy Now Pay Later, on 15 July 2026" (opened 29 July 2026). Four points from that page matter to a merchant rather than a lender.

DPC covers the case where "the lender and the supplier of goods or services are not the same person," which is the ordinary shape: Klarna or a similar firm lends, you sell. "Broking of DPC agreements is exempt from regulation," so introducing your customers to a regulated lender at checkout does not by itself make you a regulated credit broker. Agreements entered into before 15 July 2026 remain exempt, so this is about new business, not your back catalogue. And registration for the temporary permissions regime has closed, which means the practical check is on your provider: confirm the firm behind your button is authorised or holds a temporary permission, because if it is not, the button is a problem you did not create and will still own.

One live gap we could not close from a first-party source: whether specific selling models, such as in-home sales, need authorisation anyway. That claim circulates in law-firm commentary and is not on the FCA page above, so treat it as a question for your own adviser rather than a settled point.

A note on the platform version

If you are on Shopify, Shop Pay Installments is the built-in option, and Shopify's own help page (opened 29 July 2026) publishes the eligible order range, "$35 to $30,000 USD in the United States, including discounts, shipping, and taxes," but no merchant rate. Secondary blogs quote a figure; some of them also say rates are agreed store by store, which would explain why no public number exists. Do not plan on a rate you read on a blog. Open your own agreement, find the line, and put that number into the formula above. And if you already pay a penalty for using an outside gateway, factor that in too, because platform fees stack on top of processing fees.

The decision rule, then. If your gross margin is above 40%, your average order sits high enough that customers ask about splitting payments, and you are on the narrower PayPal gap, the required lift is under one order per hundred and the button is close to free. If your margin is thin, your basket is small, and you would be paying a 3.09 point premium on a fifth of all orders, you need a real conversion improvement to break even and you should measure before you commit. Passing the cost to the customer is not the escape hatch it looks like either: surcharging is governed by rules that vary by network and jurisdiction, and a fee at checkout tends to undo the conversion gain you enabled BNPL to get.