~$10M in card revenue retained through a merger

Turned a high-stakes card switch into a three-field flow with clear next steps, so clients kept using their cards.
Activate your new credit card screen in a browser window

~$10M in card revenue retained through a merger

Led all design for the flow HSBC Bank Canada clients used to activate their new RBC credit cards before the merger, reducing support friction and helping clients activate early.
Activate your new credit card screen in a browser window
The activation form: full bank names in every label, plain helper text and a date of birth split into day, month and year.
Role
Senior Product & Content Designer
company
RBC Royal Bank of Canada
duration
6 weeks
setting
Engineering, Product, various credit cards teams
context & stakes

Activating cards before two banks became one

RBC acquired HSBC Bank Canada, and every HSBC Bank Canada credit card was replaced with an RBC card. Clients could activate their new card before Legal Day 1, so it would work the day the merger took effect.

I was brought into this project on my own and led product and content design from discovery to implementation in 6 weeks. I worked with product and engineering teams from different areas and other designers, and met weekly with the credit card C-suite. RBC’s design system sets the atoms, like spacing, colour and type, so the screens and components in this flow are my own designs.
Results at a glance
68,000+
card activations in the first three weeks
~3,500
activations above target, ahead of the merger
~$10M
annual card revenue retained (estimated)
business & user impact

Every new card had to work on day one

Clients who activated early had a card that worked the day the merger took effect. For RBC, each new cardholder brings annual card revenue, so every activation mattered.
Keep cardholders
Each activation kept a client’s card with RBC through the merger.
Ready on Legal Day 1
Activated cards worked on April 1, 2024, with a clear next step to set up online banking.
No one to call
With no support line before the merger, the screens were the only help clients had.
problem & bet

A letter that didn’t say which card replaced which

Cards arrived with one letter that didn’t pair each RBC card with the HSBC Bank Canada card it replaced. Another team sent the cards and wrote the letter, so we responded to their strategy rather than shaping it.
No customer service
The banks weren’t integrated yet, so there was no support line. Every screen had to answer its own questions.
No card mapping
The letter didn’t pair cards, so people with several cards could enter the wrong pair.
Strict naming
HSBC required its name to appear as HSBC Bank Canada, in full, everywhere.
Fixing the letter
Executives wanted the flow to cover the letter’s gaps, even on screens people rarely read.
1 · Activation link in mailed letter
Clients arrive from the link they were sent, in English or French.
→
2 · Enter details
RBC card numberlast 4 of HSBC Bank Canada carddate of birth
→
3 · Match the cards
Both card numbers and the date of birth are checked against client data.
→
4 · Outcome
successtry againout of trieserror code
Activated cards work on April 1, 2024. Every other outcome has its own screen.
the bet
If every screen answered its own questions, named each bank in full and gave a clear next step for every outcome, clients could activate on their own, without a support line.
constraints

Designing for every scenario

Product, engineering and HSBC Bank Canada partners shared their requirements and limits. The data available to us shaped a simple task flow that worked for every party.
Six weeks
No time for a new research round, so I built on earlier RBC card research and the patterns it had proven.
A fixed letter
The card and letter team worked separately, so the flow had to absorb decisions made upstream.
Business clients
The flow was built for personal clients, but business clients could land in it anyway.
Minimum age of 14
Cardholders had to be at least 14 to activate, explained on the date of birth field.
Annotated desktop flow showing every path from activation to success, failure and system states
The final flow, annotated for engineering and QA: every path from cancelling to running out of tries, with notes on validation, language and where each link goes.
three key decisions

Clear names, a path for every mismatch and errors that fix themselves faster

With no one to call, each decision had to answer a question before a client needed to ask it.
decision 01

Full bank names, everywhere

HSBC required its name to appear as HSBC Bank Canada everywhere. I extended that to RBC and used full bank names in labels, helper text, tips and final screens, so it was always clear which card was which.
The activation form
The activation form: full bank names in every label, plain helper text and a date of birth split into day, month and year.
Error summary
The error summary links to each field, and every label names its bank in full.
decision 02

A clear path when cards don’t match

Clients get three tries, since data showed no one held more than three cards. Each retry shows the tries left and the most likely fix: use the number from another card. After three tries, activation pauses and the screen explains what to do on April 1, 2024.
Cards don't match
Cards don’t match: tries left, plus the most likely fix.
Out of tries
Out of tries: what to do on April 1, 2024.
decision 03

An error code on every technical error

When clients called about an issue, the team always asked for the error code, because it pointed to the exact problem. So every technical error screen shows its code with a Copy button, turning a vague “it didn’t work” into an issue Product and engineering could fix quickly.
Technical error
Technical error, with a code that can be copied.
Maintenance
Maintenance: an expected return time, agreed with the senior executive team.
outcomes

Ready on day one

Clients activated their new cards ahead of Legal Day 1, without a support team behind the flow.
68,000+
card activations in the first three weeks
~3,500
activations above target, about 5%, ahead of the merger
~$10M
annual card revenue retained, estimated from the revenue each new cardholder brings
3
tries at most, enough to cover every card a client held while limiting fraud
solution

Click through the flow

Walk through every scenario in Figma: the happy path, missing details, cards that don't match, a technical error and maintenance. Fill in the form, trigger the errors and try leaving midway, on desktop or mobile.
Card activation form shown on desktop and mobile
what I’d do next

Fix the gap upstream, then measure where clients get stuck

Next time, I’d work with the letter team from the start, so the card mapping lives in the letter instead of being patched in the flow. After launch, I’d track failed matches and FAQ visits by screen to see where clients still got stuck.
Share early, every week
Weekly reviews with the credit card executives kept decisions moving and answered questions before they became blockers.
Let the flow shape the FAQ
Mapping every path showed what the screens couldn’t cover, so the FAQ was written to fill those gaps.
Design around upstream
We couldn’t change the letter, so the flow covered its gaps.