Two of those three can be compared and one of them can't. Where your customer decides when to act, card payments arrive on or before the due date 65.1% of the time, open banking 62.0% and bank transfer 57.9%, measured on 59,777 paid customer-initiated invoices over 26 months (Adfin platform data). Direct debit reads 0.20% on time against the same test, and that number describes the Bacs calendar instead of your customers.
In this article
The short version
- Among methods your customer starts: card 65.1% paid on or before the due date, open banking 62.0%, bank transfer 57.9% (Adfin platform data, 59,777 paid customer-initiated invoices across 1,042 businesses over 26 months).
- Direct debit reads 0.20% on time on 299,504 paid collections, because a Bacs collection settles a median of three days after your due date when nothing has gone wrong (Adfin platform data).
- So any league table with all three in it reports the settlement cycle for its largest row. Most comparisons of this kind make exactly that error.
- Hold one business constant and the customer-initiated gaps fall under 1.5 percent, on a business-by-business split close to a coin flip (Adfin platform data).
- Apple Pay and Google Pay belong to their own comparison.
- Direct debit takes away your customer's decision about when to pay. The other methods differ mainly in how many steps stand between wanting to pay and having paid.
Why direct debit can't go in the table
Ask which method gets paid on time and you're asking about your customer's behaviour. A direct debit doesn't record any. You hold a mandate, you give notice of the amount and the date, and then a file goes to Bacs on your customer's behalf.
Bacs describes the collection as taking "less than three days from start to finish", over an input day, a processing day and an entry day. In Adfin's own book a collection is entered against the mandate a median of two days before the due date and the money moves five days later, so your typical direct debit lands three days after your due date with nothing wrong anywhere (Adfin platform data). Measured against the date on your invoice, 0.20% of 299,504 paid direct debit collections read as on time.
Nobody was late. The scheme did what the scheme does, and your due date was never the date the cash was going to arrive. Put that 0.20% in a table next to 65.1% for card and you've published a Bacs timetable as a customer behaviour finding, and direct debit is 62% of the paid invoices there. A blended average across all methods hides the same defect, so the confident days-late comparisons you'll find elsewhere are mostly measuring settlement. Why direct debit payments look late when they are not works through what that does to your ageing report.
What direct debit can be measured on is whether your money arrives: 97.0% of collections clear on the first attempt and 98.7% are eventually collected (Adfin platform data, 269,518 collections first attempted between July 2024 and mid-July 2026).
The ranking for the methods your customer starts
Card, open banking and bank transfer are comparable with each other, because in all three your customer acts and your due date is the deadline they're acting against.
| Method | Paid on or before the due date | Invoices measured |
|---|---|---|
| Card | 65.1% | 38,284 |
| Open banking | 62.0% | 8,010 |
| Bank transfer | 57.9% | 2,852 |
Adfin platform data, 59,777 paid customer-initiated invoices across 1,042 businesses over 26 months. Apple Pay and Google Pay are in the same population and have their own page.
Now the caveat, in the same breath as the ranking, because the two don't separate. Compare methods inside a single business's own book, where the terms and the client list are held still, and the differences shrink to under 1.5 percent on a business-by-business direction close to even (Adfin platform data). Open banking looks about 1.5 percent behind card pooled and comes out slightly ahead inside one book, on a 51 to 44 split across 96 businesses. Bank transfer looks about 7 percent behind pooled and half a percent behind inside one book, on a 13 to 13 split across 26.
So the seven percent between card and bank transfer above is mostly a statement about which businesses offer which method, and which of their customers reach for it. The full league table with confidence intervals is in the UK Getting Paid Report, and the within-business test behind the caveat is set out in the method page.
One limit on the population before you plan around it: these are invoices raised by businesses collecting through Adfin, weighted towards small practices billing small recurring fees, so half of them are under £180. Your own client mix will differ.
Most of that gap describes who uses what
A business offering open banking on a payment link is usually a different business from one emailing bank details at the foot of a PDF, and their clients differ too. The pooled numbers pick up both differences and hand the credit to the rails.
You can still use the ranking, as long as you read it as description and not as a lever you can pull. A client of yours paying by bank transfer has often been given nothing else to work with.
What each method decides for you
Set the pooled percentages aside and your methods differ in something simpler: what each one takes off your customer's plate.
- Direct debit removes their decision about when to pay. Your customer agreed once, you give notice, and the collection happens on the date you set. Nothing has to reach the top of their to-do list on any particular morning, so what's left to go wrong is the money not being there, or a cancelled mandate.
- A card or open banking payment leaves the decision with your customer and removes almost all of the work. Your amount, your reference and your details are already filled in, and approving takes a phone.
- A bank transfer leaves the decision with your customer and hands them a small task: open the app, add a payee, copy a sort code and an account number, type your reference, check it. Anything with that many steps competes with the rest of their week.
Your ranking follows that order, and so does the small residue that survives holding one business constant. Your method also affects whether your due date gets missed far more than how long you then wait, since among late payments the median delay is 8 days for card and open banking and 10 days for bank transfer (Adfin platform data, 59,777 paid customer-initiated invoices).
Comparing methods inside your own book
For an answer about your own clients, your accounting system probably holds enough to run the cut yourself.
- Split last year's paid invoices by the method each customer used, keeping direct debit in a column of its own throughout.
- For the customer-initiated columns, compare the payment date with the due date you set and read off the share paid on or before it.
- For direct debit, measure the share collected first time, the share eventually collected, and how many mandates were cancelled. Days against your due date can't work here.
- Treat a difference of a percent or two between customer-initiated methods as ordinary variation, because on a few hundred invoices that's usually all it is.
- Check your terms and your due dates before you conclude anything about rails. A weekend due date, or one already in the past, costs you more than your method does.
Where one of your clients is late most months, changing their method is unlikely to be enough on its own. Moving them off deciding when to pay is the change that tends to hold, and that's a conversation about a mandate and not about your checkout.
Common questions
Which payment method gets paid on time most often? Among methods where your customer chooses when to act, card leads at 65.1% paid on or before the due date, then open banking at 62.0% and bank transfer at 57.9%, on 59,777 paid customer-initiated invoices over 26 months. Direct debit can't be ranked alongside them, because a Bacs collection settles after the due date by design.
Why can't direct debit be compared with card on an on-time basis? Because the on-time test measures a customer's decision and a direct debit doesn't involve one. Only 0.20% of 299,504 paid direct debit collections read as on time against the invoice due date, since a collection is entered before the due date and settles a median of three days after it. That figure describes the Bacs cycle.
Does direct debit actually get you paid more reliably? On the measure that fits it, yes: 97.0% of collections clear on the first attempt and 98.7% are eventually collected, on 269,518 collections first attempted between July 2024 and mid-July 2026. Direct debit trades a few days of settlement for that reliability.
Is bank transfer really the worst option for getting paid on time? Pooled, it's the lowest of the three at 57.9%. Inside a single business's own book that gap almost disappears, at about half a percent behind card on a 13 to 13 split across 26 businesses, so it mostly tells you which businesses rely on transfers.
Will switching a client from bank transfer to card get them paying on time? This data can't promise that. The within-business comparison puts the difference under 1.5 percent, so a switch of method is worth making because it removes work for your client, and not on the strength of the pooled ranking.
What should I measure instead of a blended days-late average? Two numbers, kept apart. For invoices your customers pay themselves, the share arriving on or before your due date. For direct debit, first-attempt success, eventual collection and cancelled mandates. A single average across both mostly reports the Bacs cycle.
Sources
- Bacs — Direct Debit, getting started (accurate as of August 2026)
This article describes payment behaviour measured on Adfin's own platform and is not legal or financial advice. The on-time figures cover customer-initiated payments only, since direct debit settles after the due date through the Bacs cycle by design and cannot be blended with them. Last updated August 2026.
