Stripe vs. Square vs. PayPal for Small Businesses: Fees, Payouts, Hardware, and the Effective-Rate Math

On a $10 online sale, the gap between the cheapest and the priciest of the big three processors is 25 cents — pocket change. On $20,000 a month of those same sales, the gap is over $190 a month, every month, taken straight out of your margin before you see a dollar of it. Processing fees are the rare business cost you lock in with a single signup decision and then pay on autopilot for years, which makes the choice worth an hour of math up front.

Stripe, Square, and PayPal all charge zero monthly fees on their entry plans and all quote rates that look nearly identical at a glance — something in the high two percents plus a fixed handful of cents. The real differences hide in four places this guide compares head to head: online checkout depth, in-person hardware, payout speed, and the effective-rate math at your volume and average ticket size.

The Short Answer​

• Pick Stripe if you sell online first — especially subscriptions, marketplaces, or anything custom-built. Its 2.9% + 30¢ online rate beats both rivals' list prices, and Stripe Billing, Payment Links, and the API cover billing models the others cannot touch.
• Pick Square if you sell in person first — retail counter, food truck, salon, market stall. Its free POS software, $49-and-up reader lineup, and next-business-day deposits are built for card-present life, and it charges no fee when a customer disputes a charge.
• Pick PayPal if your buyers already live there — eBay-style marketplaces, international customers, invoice-based freelancing. Its checkout rate is the highest of the three, but the branded button, Venmo, and Pay Later options can lift conversion enough to earn the premium.

At $20,000 a month in $50 online sales, Stripe costs about $700, Square about $780, and PayPal Checkout about $894. By the end of this guide you will know how to run that same calculation on your own numbers.

Online Rates: Where Stripe Wins on Price​

For card-not-present sales — your website, payment links, invoices paid online — the published US rates stack up like this:

Three things to notice. First, Stripe's standard online rate is the lowest headline number, and you get it with no monthly fee. Second, Square only matches Stripe online if you pay for the Plus plan, a monthly per-location subscription that also trims the in-person rate — worth it once your volume makes the 0.4-point online spread exceed the subscription. Third, PayPal has two online prices and most small sellers land on the expensive one: the 3.49% + 49¢ branded Checkout rate applies whenever buyers pay through the familiar PayPal button, while the 2.99% + 49¢ rate covers unbranded card fields. The 49-cent fixed fee is also the sting in small tickets — on a $10 sale it alone adds nearly five points to your effective rate.

Keyed-in cards cost more everywhere: expect around 3.4% + 30¢ on Stripe and 3.5% + 15¢ on Square. If you take many phone orders, that surcharge deserves its own line in your comparison.

Online Depth: Checkout, Subscriptions, and Global Reach​

Rates decide what you pay per sale; platform depth decides what you can sell at all.

Stripe is the developer's platform that non-developers can still use. No-code sellers get hosted Checkout pages and Payment Links you can paste into an email or invoice. But the reason venture-backed startups default to Stripe is everything behind that: Stripe Billing handles subscriptions, trials, proration, and usage-based pricing; Connect powers marketplaces that split payments between sellers; and the API supports 135-plus currencies with localized payment methods. If your billing model might ever get complicated — tiers, seats, metered usage, multi-party splits — starting on Stripe means never migrating processors mid-growth.

Square covers standard small-business online selling. Square Online gives you a basic storefront, Square Invoices handles one-off and recurring invoices for free, and appointments, gift cards, and loyalty plug into the same dashboard. What you do not get is Stripe's programmability: custom checkout flows, marketplace splits, and exotic billing models are outside Square's design center. Square also operates in roughly eight countries, so internationally ambitious sellers hit a wall early.

PayPal sells trust and reach. Hundreds of millions of buyers already have accounts, the branded button measurably lifts conversion on unfamiliar stores, and PayPal operates in 200-plus markets with Venmo and Pay Later (buy-now-pay-later) bundled in. The trade is control: PayPal's risk system is famous for sudden holds and reserves on new or spiky-volume accounts, and its subscription and API tooling lag Stripe's by a generation. Choose PayPal for buyer familiarity, not for billing sophistication.

In-Person Rates and Hardware: Where Square Wins​

Walk the comparison into a physical store and the ranking flips. Card-present rates:

PayPal's Zettle reader actually posts the lowest in-person percentage, with Square and Stripe close behind — and note how the fixed fee shuffles the order. Square's 2.6% beats Stripe's 2.7% on percentage, yet Stripe's nickel fixed fee versus Square's 15 cents makes Stripe cheaper on that $40 ticket ($1.13 vs. $1.19). This crossover is the whole game for counter businesses: below roughly a $10 ticket, the fixed fee dominates and Stripe's 5 cents wins; above it, percentages decide. Run your own average ticket through both formulas before concluding anything from the headlines.

Hardware deepens Square's in-person lead:

• Square ships a complete shelf: a free magstripe reader to start, a $49 contactless-and-chip reader, a $149 counter stand that turns an iPad into a register, and a $299 standalone Terminal with built-in receipt printing. The POS app itself is free and genuinely good — inventory, staff permissions, offline mode, and end-of-day reports included.
• Stripe Terminal offers well-built readers that drop into a custom point of sale you (or your developer) assemble with the Terminal API. The hardware is solid and the omnichannel reporting — one dashboard for online and in-person — is excellent, but there is no free Square-style POS waiting for you. Stripe in person suits businesses with developers, franchises, or platforms embedding payments.
• PayPal Zettle sells an inexpensive reader and a basic free POS aimed at market stalls and pop-ups. It is the cheapest way to start taking cards in person, but inventory, staff management, and reporting are thinner than Square's, and the ecosystem beyond the counter is PayPal's online world with all its holds-and-reserves quirks.

If your week is mostly face-to-face selling, Square's free POS plus its reader lineup is the package to beat. If you need in-person payments inside software you are building, Stripe Terminal is the only serious option of the three.

Payout Speed: When the Money Reaches Your Bank​

A processor that holds your money an extra day is an interest-free loan you never agreed to. Standard schedules:

• Square: next business day. Sales taken today land tomorrow on the standard free schedule — the fastest default of the three, and a genuine cash-flow advantage for businesses that turn inventory weekly. Instant transfers to a debit card cost 1.75%.
• Stripe: two business days. Standard payouts arrive about two business days after the charge (T+2), and your very first payout typically waits 7 to 14 days while Stripe vets the new account. Instant payouts cost 1% with a 50-cent minimum.
• PayPal: immediate to your balance, 1–3 days to your bank. Money hits your PayPal balance the instant the buyer pays, which feels fast — but moving it to your bank takes one to three business days free, or minutes for a 1.75% instant-transfer fee (capped at $25). New sellers should also budget for holds: PayPal routinely reserves a slice of incoming funds for days or weeks until the account builds history.

For a business running on thin cash — buying Friday's supplies with Thursday's sales — Square's next-day default can matter more than a tenth of a point on the rate. Price the float, not just the fee.

Disputes and Chargebacks: The Fee Nobody Quotes You​

Every processor passes card-network dispute costs to you, but the surcharges differ:

• Square charges no dispute fee. Win or lose, Square absorbs its own cost — a quiet perk worth real money if you sell in a category with friendly-fraud risk.
• Stripe charges $15 per dispute, refunded if you win.
• PayPal charges $15 or more per dispute depending on your volume tier, on top of the original transaction fee it keeps.

None of the three will save you from the dispute itself — that fight is always about your receipts, delivery proof, and refund policy — but Square's zero-fee stance is one less insult added to the injury.

The Effective-Rate Math: Run Your Own Numbers​

Headline rates mislead because the fixed per-transaction fee hits small tickets far harder than large ones. Your effective rate is simply total fees divided by total volume, and you can estimate it in two minutes:

1. Take last month's card volume and transaction count.
2. For each processor, multiply volume by its percentage rate, add the fixed fee times the count, and divide by volume.
3. Add monthly plan fees, hardware amortized over its life, and one dispute fee per dispute you typically eat.

Worked example — $20,000 a month online at a $50 average ticket (400 transactions):

• Stripe: 400 × ($1.45 + $0.30) = $700.00 (3.50% effective)
• Square free: 400 × ($1.65 + $0.30) = $780.00 (3.90% effective)
• PayPal Checkout: 400 × ($1.745 + $0.49) = $894.00 (4.47% effective)

Same sales, $194 a month between first and last — about $2,300 a year. Now rerun it with your numbers, because the ranking moves with ticket size: at a $200 average ticket the fixed fees fade and the spread compresses to pure percentage points, while at a $6 coffee-shop ticket the fixed fee is nearly the whole story. If your volume ever crosses roughly $10,000 a month with steady tickets, also get an interchange-plus quote from a traditional merchant account — flat-rate simplicity starts costing real money at scale, and the processors above know it.

Booking It Right: Gross Revenue, Separate Fees​

Whichever processor you pick, book every sale the same disciplined way: record the full gross amount as revenue and the processor's cut as its own expense — never book the net payout as revenue. Net booking understates both your top line and your costs, which corrupts every margin, tax, and loan-application number downstream. Each payout then reconciles as gross sales minus fees, refunds, and any reserve hold, matching the bank deposit to the cent.

In a plain-text ledger the pattern looks like this:

2026-10-18 * "Stripe payout #1234"
  Assets:Checking                      1,934.20 USD
  Expenses:Payment-Processing-Fees        65.80 USD
  Income:Sales                         -2,000.00 USD

One transaction, three legs: what arrived, what the processor kept, and what the customer paid. Keep a separate fees subaccount per processor while you compare them — Expenses:Payment-Processing-Fees:Stripe versus :Square — and your ledger answers the effective-rate question with real data instead of brochure math. The Beancount documentation walks through account setup and balance assertions if you are starting from scratch.

Reconcile each payout the day it lands: match the processor's payout report line by line against the bank deposit, and chase any gap — a held reserve, a refund you forgot, a fee change you missed — while the memory is fresh. Processors change fee schedules with an email you will not read; your reconciliation is the backstop that catches it.

Making the Call​

List your last three months of volume, transaction count, average ticket, online-versus-in-person split, and dispute count. Run the effective-rate formula for all three, add the value of next-day versus two-day payouts to your cash position, and weigh the platform question honestly: Square's free POS today, Stripe's billing depth tomorrow, or PayPal's buyer trust where your customers already are. The cheapest processor is the one that costs least on your tickets, pays out on your schedule, and never forces a migration the month you can least afford one.