Best processing for an online store, and the thing worth more than the rate.
Every online sale is card-not-present, so online rates start higher and stay higher. Within that, the published options separate cleanly by what you are building — and the decline rate most stores never measure is worth more than any of them.
What this page settles.
Three questions, in the order most people ask them.
The cost you cannot negotiate away
Every online sale is card-not-present. The card was never read, the chip never authenticated, and the issuing bank carries more fraud risk as a result. Interchange reflects that, and the gap is roughly 0.3% to 0.5% above the equivalent card-present tier on the same card.
That is the networks’ own schedule, not a markup anyone invented, and no amount of shopping around removes it. What interchange is.
Three things follow, and they shape everything below.
Comparing your rate to a shop’s is meaningless. A coffee counter at 2.3% and an online store at 3.4% can be paying identical margins.
Card mix hits you twice. Rewards cards cost more, and they cost more again card-not-present. Online, your customers are disproportionately rewards-card holders. What rewards cards cost.
Downgrades still exist. Card-not-present is a tier, not a floor. Missing AVS data, missing descriptor data or late settlement push you below it. Card-not-present downgrades.
So the honest framing is: accept the structural cost, pick sensibly within it, and then go after the things that are actually yours — which are worth considerably more.
The published options, by what you are building
You are building your own checkout, or payments live inside a product. Stripe, at 2.9% + 30¢ for domestic cards. It is the benchmark online rate, you are underwritten as a merchant in your own right rather than as a sub-merchant, and the billing engine has no equivalent at this price if you charge repeatedly. Watch the published surcharges: +1.5% international, +1% currency conversion, +0.5% manually entered. For a cross-border store those are the dominant cost, not the headline. Cross-border fees.
You are on Shopify. Shopify Payments, and the decision is largely made for you. 2.9% + 30¢ on Basic falling to 2.25% + 30¢ on Plus, and Shopify adds an extra transaction fee of 2% to 0.2% if you process with anyone else. On Basic that surcharge makes an alternative impossible; on Plus at 0.2% it is worth pricing one. Stripe against Shopify Payments.
You want the PayPal button. Add it, alongside whatever else you use. PayPal’s online rates are the highest here — 2.99% + 49¢ for standard cards, 3.49% + 49¢ for Checkout — and that is not the point of it. A share of buyers will only pay that way, and if it converts orders that would otherwise be lost, it pays for itself on those orders several times over. Measure whether it actually is before assuming. Stripe against PayPal.
You also sell in person. Keep the channels separate in your head and price them separately. Square is 3.3% + 30¢ online on its free plan and 2.9% + 30¢ on the paid tiers, which is competitive only if you are already using Square for the counter. Square against Stripe.
You are doing real volume. See below.
The number worth more than all of them
Most online sellers can quote their processing rate to two decimal places and have no idea what their decline rate is. The decline rate is worth more money, usually by an order of magnitude.
A store doing $80,000 a month with a 9% decline rate is having roughly $7,900 of attempted purchases refused. If a third of those are recoverable — and on stored-card and subscription volume, considerably more than a third usually is — that is $2,600 a month. A 0.2% improvement in your processing rate on the same volume is $160.
Four things move it, in order of return:
Account updater. The network service that updates expired and reissued card numbers automatically on stored cards. Without it, a renewed card is a decline. This is the single highest-yield fix for any business billing repeatedly, and it is frequently just switched off.
Retry logic that distinguishes hard from soft declines. Retrying a hard decline repeatedly gets you flagged. Retrying a soft decline after a sensible interval recovers a real share. The distinction is in the response code. Why transactions get declined.
AVS handling set deliberately. Rejecting every street-address mismatch rejects a meaningful number of legitimate customers who typed their billing address slightly differently. Decide your threshold rather than inheriting a default.
3-D Secure configured rather than just enabled. Done well it shifts fraud liability and barely touches abandonment. Done badly it adds a challenge screen to good customers.
Where the second-biggest leak is
Disputes, and the largest category is not stolen cards. It is customers who made the purchase and disputed it anyway.
Check your billing descriptor first. If what appears on the statement is not obviously your store, you are generating disputes from your own paying customers. It takes one phone call and it is the cheapest fix available. Billing descriptors.
Then delivery evidence. Tracking, confirmation, signature on higher-value items. An item-not-received dispute is won with the carrier record and lost without it. How to fight a chargeback.
Then the ratio, not the count. Both networks monitor your chargeback-to-transaction ratio and the consequences of breaching their thresholds are serious — Visa’s VAMP merchant ratio fell to 1.5% on 1 April 2026. Chargeback monitoring programs.
And check whether you are paying for a gateway twice. A processor’s monthly gateway fee alongside a platform that already includes one is common and nobody notices. Gateway fees.
When to leave a flat rate
Above roughly $40,000 to $60,000 a month online, your own underwritten merchant account on interchange-plus generally costs less than any of the published flat rates, and the gap widens from there. The crossover is higher online than in person because card-not-present interchange is higher, which leaves less margin for a flat rate to be carrying.
Two things complicate it. If you are on Shopify below Plus, the outside-processor surcharge probably makes it impossible regardless of the arithmetic. And if your payments are deeply built into Stripe’s billing or marketplace products, the migration cost is real and should be counted.
Work it out rather than assuming: take your effective rate from a statement, and compare it to interchange + 0.75% + 25¢ at your own card mix. The method.
What we charge
Interchange + 0.75% + 25¢ for card-not-present, interchange + 0.50% + 10¢ in person, with no contract, no monthly fee and no monthly minimum.
Online, the most useful thing we do costs nothing and is not a sales pitch: separate the three costs on your statement — processing, gateway, platform — which on most online statements are tangled together. Then measure the decline rate. If the decline rate is the problem, switching processors will not fix it and we will say so.
While we are on the subject
Why is my online rate higher than a shop's?
Because the card was not physically read, so the issuer carries more fraud risk and charges more interchange. Roughly 0.3% to 0.5% above the card-present tier, and not negotiable.
Is Stripe or PayPal better for an online store?
Stripe is cheaper on cards and far stronger for anything you are building. PayPal adds a payment method some buyers insist on. Most serious stores run Stripe with PayPal as a second button.
Should I use Shopify Payments?
On Shopify, below the Plus plan, yes — the outside-processor surcharge of 2% to 0.6% makes any alternative uncompetitive. On Plus at 0.2% it is worth pricing an alternative.
What is the single biggest saving available online?
Usually the decline rate, not the processing rate. Account updater, sensible retry logic and a correctly configured 3-D Secure recover more money than any provider switch.