indie-hacker·9 min read·
Stripe vs PayPal vs GoCardless for the UK indie hacker (2026 fee comparison + decision tree)
Stripe charges 1.5% + 20p on UK cards, GoCardless caps Direct Debit fees at GBP 4, and PayPal sits at 2.9% + 30p with a brand-trust premium. None of the global comparison posts work the actual GBP arithmetic for a UK indie hacker shipping in 2026. This is the comparison that does: real fee math on a GBP 14.99 transaction, churn implications for recurring billing, chargeback exposure in plain English, and a four-question decision tree that gets you to the right processor in under a minute.

Search "Stripe vs PayPal vs GoCardless UK" from a London postcode and the first page is a wall of US comparison sites quoting dollar fees, a Stripe blog post that conveniently forgets GoCardless exists, and three affiliate-padded listicles last updated in 2023. None of them work the actual GBP arithmetic on a GBP 14.99 transaction. None of them tell you that GoCardless caps fees at GBP 4 per Direct Debit. None of them mention that Stripe Billing adds a 0.7% surcharge on top of the standard 1.5% + 20p UK card rate the moment you turn on subscriptions. And none of them have a clear decision tree for a UK indie hacker shipping in 2026 with three product shapes - one-off digital sale, recurring SaaS subscription, marketplace-style consumer trust play.
This is the comparison none of them wrote: real fee math in GBP, churn implications for recurring billing, chargeback exposure in plain English, and a four-question decision tree that gets you to the right processor in under a minute.
The three processors in plain English
Three options, three different shapes. The trick is that they are not direct substitutes - they solve different problems and the fees only matter once you have picked the right shape.
Stripe is a card processor. It charges the customer's Visa, Mastercard or Amex (plus Apple Pay, Google Pay, Klarna and a long list of local methods), settles to your UK bank account in two working days, and dominates the developer experience because the SDK is the cleanest in the industry. UK indie hackers reach for Stripe by reflex. For most product shapes, the reflex is right.
GoCardless is a Direct Debit processor. The customer authorises a one-off or recurring pull from their UK bank account via the BACS or Faster Payments rails. There is no card, no expiry date, no PCI surface, and the fee is dramatically lower than card fees because Direct Debit is bank-to-bank infrastructure run by the UK clearing system. The trade-off is friction at signup (a UK bank account check is slower than a card form) and a 3-5 working day collection window.
PayPal is a wallet plus a card processor plus a brand. Customers pay with their PayPal balance, a linked card, or a one-tap "Pay with PayPal" button that bypasses your checkout entirely. The fee is materially higher than Stripe for the same card transaction, but a meaningful slice of UK consumers - especially older buyers, marketplace shoppers and anyone who has ever been burned by an online retailer - actively prefers paying with PayPal because of the dispute-resolution layer.
Three shapes, three jobs. The fees only become the right question once you know which shape your product is.
Real GBP fee math: a worked example on GBP 14.99
The cleanest way to compare processors is to take a single common UK indie price - GBP 14.99 - and walk it through each one. Numbers as of May 2026.
| Processor | Headline fee | On GBP 14.99 | Net to you |
|---|---|---|---|
| Stripe (UK card, one-off) | 1.5% + 20p | 22p + 20p = 42p | GBP 14.57 |
| Stripe (UK card, recurring with Stripe Billing) | 1.5% + 0.7% + 20p | 33p + 20p = 53p | GBP 14.46 |
| Stripe (EU card) | 2.5% + 20p | 37p + 20p = 57p | GBP 14.42 |
| Stripe (rest-of-world card) | 3.25% + 20p | 49p + 20p = 69p | GBP 14.30 |
| GoCardless (UK Direct Debit) | 1% + 20p, capped at GBP 4 | 15p + 20p = 35p | GBP 14.64 |
| PayPal (UK card, standard) | 2.9% + 30p | 43p + 30p = 73p | GBP 14.26 |
On a single GBP 14.99 transaction the spread between cheapest and most expensive is 38 pence. Looks small. Multiply by a thousand transactions a month and the spread becomes GBP 380/month, or GBP 4,560/year - which is a Vercel Pro plan, a Claude Code Max subscription, and a long weekend in Lisbon.
A few things worth noting. Stripe's standard UK rate of 1.5% + 20p only applies to UK and EEA cards - international cards bump up to 2.5% (EU) or 3.25% (rest-of-world). For a UK-focused indie product the international mix is usually under 15%, so the blended rate sits around 1.7-1.8%. Stripe Billing's 0.7% surcharge is non-negotiable the moment you switch on subscriptions, free trials, or proration logic - it is the price of having Stripe handle the recurring lifecycle so you do not. GoCardless's 1% + 20p Direct Debit fee drops to 0.5% over GBP 100k/month (when you graduate to their Plus tier at GBP 99/month) - well past the indie hacker bracket but worth knowing. PayPal's 2.9% + 30p UK rate is the headline - the brand-trust premium is real, but it is also a near-doubling of the per-transaction fee.
Subscription churn: the 0.7% that compounds
For a recurring SaaS at GBP 14.99/month, the fee story changes shape. Two effects compound.
Effect one: the Stripe Billing 0.7% surcharge. On GBP 14.99 it adds 11p to every transaction. On a 1,000-customer SaaS with GBP 15k/month MRR, that is GBP 105/month - GBP 1,260/year - going to Stripe purely for the privilege of using Billing's lifecycle endpoints. For most indie hackers that is a fair trade. For a SaaS at scale, it is a number worth interrogating once you cross GBP 50k MRR.
Effect two: card churn. Roughly 3-5% of UK card subscriptions fail to renew each month because of expired cards, lost cards, or Strong Customer Authentication challenges that the customer never completes. Stripe's network of card-update services and SCA exemption logic recovers most of those, but a residual 1-2% monthly involuntary churn is normal. For a SaaS with GBP 15k MRR, that is GBP 150-300/month leaking out for reasons unrelated to product quality.
GoCardless changes both effects. The fee is 1% + 20p capped at GBP 4 - on GBP 14.99 that is 35p, less than Stripe Billing's 53p. There is no recurring surcharge because GoCardless is built around Direct Debit's natural recurring shape. And card churn drops to roughly 0.5-1% monthly because UK bank accounts do not expire, get re-issued, or fail SCA. The Direct Debit guarantee means the customer can claw back a payment, which is its own risk - but on a properly described, fairly priced UK SaaS the actual claw-back rate sits below 0.3% of transactions in the data we see.
The blended picture for a UK-only SaaS at GBP 14.99/month: GoCardless usually saves 15-25p per transaction on fees and another 1-2% per month on involuntary churn. For a UK B2B SaaS with predictable monthly billing, that is the strongest argument for picking Direct Debit over cards full stop.
Chargeback exposure: the asymmetric risk
Chargebacks are the silent killer of consumer-facing UK indie products, and the three processors handle them differently.
Stripe chargebacks land at roughly 0.05-0.15% of transactions for a healthy UK indie product. Each one costs GBP 15 in dispute fees on top of the refunded transaction, win or lose. Stripe's dispute UI is the best in the industry - clean evidence upload, automated checks against the customer's transaction history, win-rate of around 40-50% on legitimate disputes if you have your evidence in order. Total expected cost for a UK indie product doing 1,000 transactions/month: roughly GBP 15-45/month in chargeback exposure plus the refunded value.
PayPal chargebacks (or "disputes" in PayPal language) sit at roughly 0.3-0.5% for the same product - notably higher than Stripe because PayPal customers have a one-click "raise a dispute" button inside their account that is much easier to find than the equivalent on a Visa card. PayPal's seller protection covers physical shipped goods better than digital goods, and their dispute-resolution UX can hold money for 21 days while a case resolves. Expected cost for the same 1,000 transactions/month: GBP 60-150/month plus the refunded value, which on a digital product with a 30% margin starts to bite.
GoCardless chargeback exposure is structurally different and much lower. UK Direct Debits can be reversed under the Direct Debit Guarantee (for genuine errors or unauthorised collections), but the path is bank-mediated, takes 5-10 working days, and the rate sits at roughly zero to 0.1% for compliant UK SaaS. There is no per-dispute fee from GoCardless. The risk is not chargebacks - it is the Direct Debit Guarantee being invoked maliciously, which does happen but is rare on B2B SaaS.
For a one-off consumer purchase under GBP 50, chargeback exposure is a footnote. For a SaaS with 1,000+ recurring customers, the asymmetry between PayPal and GoCardless is hundreds of pounds a month.
The decision tree
Four questions. Answer them in order. The processor falls out at the end.
Question 1: Is this a one-off digital purchase, a recurring subscription, or a marketplace consumer trust play?
- One-off digital purchase under GBP 100 → Stripe. The 1.5% + 20p UK rate is the lowest practical fee on cards, the SDK is the cleanest, and Apple Pay and Google Pay are one toggle away. Skip the rest.
- Recurring subscription → Question 2.
- Marketplace, secondhand goods, anything where buyer trust is the conversion bottleneck → PayPal as a checkout option (alongside Stripe), not as the only one. The brand premium pays for the higher fee on the cohort that would not otherwise convert.
Question 2: Is the recurring product UK-only, primarily B2B, with monthly billing of GBP 20+?
- Yes → GoCardless. Lower fee, lower churn, near-zero chargeback exposure. Stripe Billing as a fallback for international customers.
- No (international, mostly consumer, or sub-GBP 20 monthly) → Question 3.
Question 3: Is your customer base primarily UK consumers paying by card, with international (especially US) customers under 30%?
- Yes → Stripe + Stripe Billing. The 0.7% surcharge is the price of the lifecycle endpoints; the SDK is the price of your sanity.
- No, international/US is the majority → Stripe + Stripe Billing still wins for the SDK, but consider adding PayPal as a second checkout option for the US consumer cohort that strongly prefers it.
Question 4: Does your product surface require buyer-trust signalling (secondhand goods, marketplace, regulated category)?
- Yes → PayPal alongside your main processor. The "Pay with PayPal" button on a marketplace listing converts 1.3-1.8x better than a card form alone for older UK buyers. Eat the higher fee on that cohort.
- No → you are done. Stick with the answer from Question 1, 2 or 3.
For the typical UK indie hacker shipping a recurring B2B SaaS at GBP 14.99-49/month, the answer almost always lands on GoCardless for the UK customer base, Stripe Billing for the international tail. For a one-off digital product, it is Stripe full stop. For a marketplace or trust-sensitive consumer play, it is Stripe primary plus PayPal as a second button.
Stack-level wiring in 2026
The good news for UK indie hackers shipping in 2026: all three processors integrate cleanly into the AI-native stack. A typical Vercel + Stripe wiring is a 30-minute job with Claude Code or Cursor - the SDK is so well-trodden that the AI tools generate the webhook handler, the customer portal redirect, and the SCA challenge logic on the first prompt. GoCardless takes slightly longer (90-120 minutes) because the mandate flow requires a redirect to the bank's authentication page and a webhook for the mandate-confirmed event, but the GoCardless SDK is documented well enough that Claude Code handles it in one pass. PayPal's REST API is the oldest of the three and the integration is fiddlier - budget half a day if you are wiring it as a primary processor, an hour if you are dropping in the smart button as a secondary option.
Webhooks matter. All three processors send asynchronous events for failed payments, refunds, disputes, mandate cancellations and (for subscriptions) renewal events. Wire the webhooks into a Supabase or Vercel Postgres table and you get a clean, queryable payment ledger that any future dashboard can read from. The cost is one afternoon and the upside is that you never lose track of a customer's payment state when something goes wrong.
Want a data-backed UK business idea every week? Free reports drop every Thursday - keyword volumes, SERP analysis, builder prompts. Browse the latest free report on IdeaStack.
Frequently asked
Should I bother with PayPal at all in 2026 if my fees are higher?
For a software-only UK indie product targeting digital-native buyers under 35, PayPal usually costs more than it earns - the brand-trust premium does not register because the audience trusts a Stripe checkout just fine. For consumer-facing products (especially marketplaces, secondhand goods, regulated categories like supplements or refurbished electronics) the calculus flips: the cohort that only converts because PayPal is the option pays for itself even at 2.9% + 30p. The pragmatic move is to launch with Stripe-only, watch the abandonment funnel for two months, and add PayPal as a second button if the bounce-at-checkout rate suggests buyer-trust friction.
Does GoCardless work for international customers or only UK?
GoCardless supports Direct Debit equivalents in the UK, the EU (SEPA), the US (ACH), Canada, Australia and New Zealand - a meaningful international footprint. The fees and timelines vary by country (SEPA is 1% capped at EUR 4, ACH is 1% capped at $5), and the UX of authorising a non-UK bank pull is sometimes slower than a card form because of local banking requirements. For a UK-first SaaS, default to GoCardless for the UK customer base and offer Stripe as the international fallback - the operational simplicity of two processors on two clearly delineated cohorts beats trying to force one processor across all geographies.
Is Stripe Billing's 0.7% surcharge worth it or should I roll my own subscription logic?
Stripe Billing is worth the 0.7% for any UK indie hacker doing under GBP 50k MRR. Rolling your own subscription engine - proration, dunning, free trials, plan upgrades, SCA challenges, payment retries, customer portal - is a multi-week project that you will get wrong in five different places before you get it right in the sixth. For 11p on a GBP 14.99 transaction you get the entire lifecycle handled, a hosted customer portal, a webhook event for every state change, and Stripe's payment-retry logic that recovers roughly half of involuntary churn for free. Past GBP 50k MRR the 0.7% becomes meaningful enough to consider building parts in-house - but only the parts where you have a specific, measurable advantage.
What about Apple Pay and Google Pay - do they cost extra?
Apple Pay and Google Pay are tokenised card networks that route through your existing card processor, so they pay the same fee as the underlying card. On Stripe a UK Apple Pay transaction is the same 1.5% + 20p as a UK Visa transaction. The conversion lift, however, is real - UK mobile checkout conversion roughly doubles when Apple Pay or Google Pay is enabled because the customer skips the form-filling step entirely. Turning them on in Stripe is a single toggle in the dashboard plus a domain-verification file. For any UK indie product with mobile traffic above 30%, switching them on is a free conversion lift.
Will this advice change if HMRC or the FCA rolls out new payment regulations?
The fee structures quoted are May 2026 numbers and the underlying regulations - PSD2 / SCA, the Direct Debit Guarantee, FCA conduct rules on consumer credit - are stable enough that material change inside the next 12-18 months is unlikely. The two areas worth watching are the variable recurring payments (VRP) rollout, which could give GoCardless-style economics on a card-style UX and would be the biggest payments shift since SCA, and the FCA's ongoing review of buy-now-pay-later providers, which could change the Klarna / Clearpay piece of Stripe's offering. Neither changes the decision tree above today, but a UK indie hacker shipping a recurring product in 2027 should re-check the VRP landscape before committing to a long-term processor strategy.
Filed under





