work / 02 / balance-transferSantander · 2024

Designed for the majority without failing the minority

Most customers hold one credit card; a few hold several and need to compare offers. One journey adapts to both, tested over two rounds with 145 responses.

Decision
One journey that adapts, not one for everyone
Role
Lead designer (UX Manager)
Team
Product, Engineering, Research, Copy, Compliance
Platform
Web and mobile app
Timeline
2024, handed over part-way on promotion
50%
task success, up from 31%
73%
rated it easy or very easy, up from 36%
88%
expected a confirmation at the end, up from 56%

Figures are rounded. Exact figures are shared privately on request.

01 · The problem

One journey had to serve two kinds of customer who needed opposite things.

Most credit-card customers hold one card. For them a balance transfer should be quick: see the offer, choose an amount, confirm.

A small minority hold several, and may have a better offer on a different card from the one in front of them. They need to compare before they commit.

It also had to work on the web and in the app, which have different ways in and behave differently.

Figure: the two groups of customer and what each needs from the journey
Depth: How the groundwork was done
  • Wrote a research plan with the user researcher to test the ways in across web and mobile.
  • Mapped the whole journey and where it crossed existing card-detail and servicing flows.
  • Ran sessions with Engineering to surface edge cases and limits early.
  • Used customer data to weigh the needs of single-card and multi-card customers.

02 · The constraint

The way in made a promise that the best offer might break.

In the app, most people arrive from one card's detail page. That sets an expectation: this transfer is about this card.

A customer with several cards might have a better offer on another one. Show only this card's offer and they miss it. Show every offer to everyone and the majority wade through a choice they do not have.

Designing only for comparison would slow most customers down. Ignoring multi-card customers would confuse them and risk a transfer on the wrong card.

03 · The call I made

Not one journey for everyone: one that adapts to how many cards you hold.

I proposed a conditional journey. With one eligible card it goes straight through. With several, it adds a comparison and makes any change of card explicit. Switch between the two customers to see what changes.

Same transfer, two customers

  1. Step 1Arrive from a cardSame for everyone
  2. Step 2See the offerSame for everyone
  3. Step 3Choose an amountSame for everyone
  4. Step 4ReviewSame for everyone
  5. Step 5SubmitSame for everyone

No steps added. With one eligible card there is nothing to compare, so the journey never asks.

How I made the case

In design review I quickly prototyped an alternative comparison pattern, so the trade-off could be seen instead of argued. It was refined with Product and Engineering, built, and tested with users.

  • Every eligible offer in one list that can be scanned.
  • The card the customer came in from is marked.
  • Choosing an offer on another card says so plainly, and asks before anything changes.
  • Promotional rate, standard rate, duration and fee shown for each offer.
  • The same pattern on the web and in the app, despite the different ways in.

04 · What shipped

A direct path for one card, and an honest comparison for several.

With one card there is no selection step at all: the way in, the offer, the amount, a review, submit. Attention stays on the offer, the transfer details and the confirmation.

With several, customers see every eligible offer, the card they arrived from highlighted, and a clear signal when an offer belongs to a different card.

Flow film: the single-card journey, then the multi-card comparison
Depth: Explaining what happens after a transfer

Plainer explanation helped customers understand what using the card for purchases means once a balance transfer is in place.

05 · What changed

More people finished, it felt easier, and the ending was clear.

Research combined two rounds of prototype testing with wider comprehension studies: 145 completed responses in all.

Task success rose from 31% to 50%. Ratings of easy or very easy rose from 36% to 73%. The share who expected a confirmation at the end rose from 56% to 88%.

It also showed what was left to fix: 65% understood how the fee was paid, and answers on choosing a card pointed to clearer “from” and “to” labels.

What I handed over

Part-way through I was promoted and moved off the team as a new designer joined. I set out who owned what in a RACI, built onboarding materials, ran paired sessions to pass on the context and the reasons behind decisions, and moved myself to an “informed” role: present in reviews, without taking the lead back.

06 · What I took from it

Three things I carry into the next project.

  • Explain things plainly. Customers understood the consequences of a transfer once the journey said them outright.
  • Design conditionally. The best answer was not one journey for everyone, but one that adapts to the customer's situation.
  • Lead through ambiguity. Clear ownership, a structured handover and staying on as an adviser protected quality through a change of designer.