work / 01 / open-banking-paymentsSantander · 2025
£100m+ paid through a journey I argued for
Credit-card customers could not pay their card in the app. The agreed fix was a third-party web view. I made the case for native, and it became the pattern for Open Banking across the bank.
- Decision
- Web view → native
- Role
- Lead designer (UX Manager)
- Team
- Product, Engineering, Legal, Compliance, Copy, Financial Crime
- Platform
- Native iOS and Android
- Timeline
- Discovery to phased rollout, 2025
- 250k+
- confirmed payments, April to October
- 10.8×
- monthly users, six months after launch
- 13%
- of everyone paying their card, by October
Pay credit card · Step 2 of 4
Choose your bank
Choose the bank you wish to make your payment from.
- Santander
- Barclays
- NatWest
- Halifax
- Monzo
Pay credit card · Step 3 of 4
Approve in your Barclays app
We'll take you to Barclays to approve £600.00, then bring you straight back.
Payment confirmed
£600.00 paid to your credit card from Barclays.
Figures are rounded. Exact figures are shared privately on request.
01 · The problem
If your only product was a credit card, you could not pay it in the app.
The app is how most customers deal with the bank. Card-only customers had to leave it to make the one payment they make every month.
For the bank, the card payments that did happen ran over debit rails, at a significant processing cost each year, while competitors already offered pay-by-bank.
Depth: How the opportunity was chosen
I ran a workshop with Product to rank where Open Banking could add value across the app. Paying a credit card came out top: it closed a real gap for customers and carried a clear cost case.
02 · The constraint
The decision had been made before design started: use the provider's web view.
It was cheaper to build and would meet a fixed deadline. Nobody was being careless; on the information available it was a reasonable call.
My concern was what it did to the customer at the exact point they were being asked to trust a new way to pay.
Depth: What competitors did
Every competitor journey I reviewed kept the customer in the app until the unavoidable hand-off to their own bank. There was no shared name for the feature, and most journeys spent time explaining why it was safe, which told me where customers' worry sat.
03 · The call I made
I asked for the decision to be reopened, and brought evidence.
A web view sends the customer out of the app at step two. Flip between the two versions to see where the journey leaves Santander's hands.
Same payment, two builds
- Step 1Choose amountIn the app
- Step 2Choose your bankOutside the app
- Step 3Approve at your bankOutside the app
- Step 4Payment statusOutside the app
3 of 4 steps outside the app. The customer leaves before they have chosen a bank, and Santander no longer controls what they see.
Where I pushed back
A web view was faster to ship. It also took customers out of the app at the moment they most needed to trust it.
- Benchmarked competitor journeys to show native was the norm, not a luxury.
- Ran targeted user testing on comprehension and trust, which confirmed the security concern.
- Set out the trade-off for leadership: conversion risk, session time-outs locking people out mid-payment, and a dead end for the Open Banking features already on the roadmap.
Depth: The four risks in full
- Leaving the app early, on an unfamiliar payment method, raises anxiety.
- Paying outside the app teaches customers that this is normal, which helps fraudsters.
- Short security time-outs in the app would lock out people mid-payment.
- A web view limits how closely later Open Banking features can be built in.
04 · What shipped
Native where trust is won, with one honest hand-off.
Customers choose an amount and their bank inside the app. The app says plainly what happens next, hands off to their own bank to approve, and brings them back to a confirmed payment.
Most of the reassurance is done by words: buttons that say what you are agreeing to, and each screen setting up the next.
Depth: States, edge cases and hand-off to engineering
Payment-status states, failed and abandoned approvals, saved banks for repeat payments, and how each was specified for both platforms.
05 · What changed
It worked, and other teams stopped building their own versions.
More than 250,000 payments worth over £100m were confirmed between April and October, from around 300,000 started.
Android launched first and completes at a higher rate than iOS, which pointed to where to improve next.
What I gave away
Other teams, Financial Crime among them, were starting separate Open Banking work. I shared the research and the consent, approval and status patterns so customers meet one version of it.
06 · What I would do differently
Three things, in order of how much they would have helped.
- Agree what counts as started, confirmed and failed before launch, so the numbers need no reconciling afterwards.
- Have Engineering in the first workshop, not the second.
- Pilot on iOS sooner. The gap between platforms was findable earlier.