Rebuilding ClearSaver around trust and control
ClearSaver, our Open Banking product, asked users to link their bank account through in exchange for better credit offers. Half of users dropped off on the first landing page and half again on the second. I led the content redesign that collapsed two pages into one, restructured it around what users told us they needed to feel safe, and doubled bank account linking conversion.
29% → 51%
Users reaching the bank connection screen
11% → 22%
Overall bank account linking conversion, which I've doubled
+11%
Beat our squad target for conversation by 11 percent
2 → 1
Landing pages in the Open Banking activation journey
Role
Content designer in the Offers Innovation squad. I led the content strategy and copy for the relaunched journey, set up the card sort with the product designer and synthesised the usability testing into a content model.
Timeline
July to October 2022.
This was during the first four months at ClearScore in my probation period.
stakeholders
Product designer, product manager, engineers, compliance, copywriters, customer operations and the Home squad.
The problem
ClearSaver launched in May 2022 with two landing pages before the bank connection screen. The funnel told a blunt story: around 50% of users left on the first page, around 50% of the rest left on the second, and only around 10% completed the journey.
The squad’s first instinct was a light-touch fix. A tweaked version of the two-page flow was tested with six users in July. The research debrief was clear that it would not move the needle, because it still did not answer the questions that stopped people sharing their data. We had been optimising the words without fixing the structure.
Ahead of a marketing relaunch in October, the squad set a target of getting users to the bank connection page from 29% to 40%.
What we knew
Research
Three user needs had already been identified and kept resurfacing: I need to know I’m getting something in return for sharing my data, I need to feel in control of my data, and I need activating ClearSaver to be clear and easy. The July usability test showed the second need was the weak point. Confidence in sharing data was mixed, only half of participants said they were likely to connect an account, and users had a long list of unanswered questions about who could see their data, what lenders would do with it and whether it was safe.
Brand data
The brand tracker showed that sharing personal information was the single biggest factor holding back ClearSaver activation, with trust in Open Banking close behind.
Constraints
The announcement component on Home, our largest entry point into the journey, was unavailable during the relaunch. Any uplift had to come from the journey itself.
Place holder showing image of current journey
My approach
The July test told us copy tweaks on the existing structure would not work, so I argued for rebuilding the content model from the ground up. I suggested an open card sort to find out how users themselves would group and prioritise the information, and set it up with Alex. Thirty ClearScore users in the market for credit sorted 37 questions.
They converged on four groups: benefits and rewards, data privacy and security, linking a bank account, and general FAQs. When asked which single question mattered most for their confidence, “how secure is my information?” and “who has access to my personal information?” came top.
That gave me the structure and the vocabulary. Users said “data” where we said “information”, and “security” where we said nothing at all. I restructured the journey around their four groups, led with the proposition from the user’s point of view, then put data, privacy and control on the page rather than in a help link.
The rewrite introduced several things ClearScore had not said in product before:
-
Named Open Banking directly. Some users brought it up unprompted in testing and associated it with trust, so hiding the term was working against us.
-
Framed the data exchange as like for like: lenders get more confidence, you get better rates.
-
Added social proof: 6 million people in the UK use Open Banking and over 1 million have linked an account through ClearScore.
-
Put autonomy in the copy, making clear users can unlink at any time and control stays with them.
-
Answered the top card sort questions in plain English, including whether it affects your credit score and whether it is free.
I took the approach through a content crit with other content designers, then presented it at a zeroing session to mission leads and the Director of Design to get sign-off before build.
[IMAGE PLACEHOLDER: card sort cluster analysis or the resulting four-category content model, with the top six priority questions]
What I chose and what I ruled out
Rejected
Iterating the two-page flow
This was the cheap option and the one we tested first. The research said it would not have a material impact on conversion because it did not address data confidence. I used that evidence to make the case for a structural rebuild instead of another round of edits.
Rejected
Keeping privacy and security in the help section
Our instinct as a fintech was to lead with the benefit and tuck the data detail away. The card sort and brand tracker said the opposite: data security was the top priority and the thing users wanted answered before anything else. It became a dedicated section on the page.
rejected
Avoiding the term “Open Banking”
There was a view that the jargon would scare people. Testing showed the reverse for users who already knew it, and for those who didn’t, explaining it plainly was better than leaving a gap they filled with worry.
Chose
One page, structured in the users’ own categories
Fewer steps removed physical friction. Structuring the content around the four card sort groups, in users’ language, removed the hesitation that the steps had been hiding.
accepted
Relaunching without the Home announcement entry point
We could not use our biggest traffic driver during the relaunch. We shipped anyway and measured the journey on its own merits.
What we shipped
A single ClearSaver landing page replacing the two-page flow: a front-loaded heading explaining what ClearSaver does for you, the wider benefits (affordability report, improving offers over time, exclusive to ClearScore), a new privacy and data section with social proof and plain English answers to the top user questions, and clear messaging that the user stays in control. I also worked with customer ops to add matching content to the help centre so the story was consistent beyond the journey.
[IMAGE PLACEHOLDER: before and after, the original two-page flow next to the relaunched single page]
Results
Measured in Amplitude before and after the October relaunch:
29% → 51%
11% → 22%
Users reaching the bank connection screen
Overall bank account linking conversion, which I have doubled
Within 48 hours of launch I posted early signs to the content design team at 24%, roughly double the previous rate. It settled at 22% and held. The squad’s target had been 40% on the first metric. We passed 50%.
This was a before-and-after measurement, not an A/B test, and it coincided with a marketing relaunch and CRM push. The direction and size of the change are hard to explain by traffic alone, since more users arriving does not usually double the proportion who complete a journey, but I would want a controlled test to put a precise number on the content’s contribution.
"Jordan suggested doing a card sorting exercise with users which really helped us to redefine that content structure. The outcome of Jordan’s work saw a 10% uplift in conversion for users connecting a bank account through the ClearSaver journey."
Alex nicholson, Product designer
"Let’s get ClearSaver out of the way ;) WHAT A BEAST THAT WAS. Yet you trusted your instincts, collaborated with Alex to back up your ideas and pressed hard for a better experience for our users. One that resulted in big, big metrics to fall on: increasing click throughs to 51% – Jordan that’s over half all users clicking on that screen now! – and doubling (DOUBLING!) conversion rates."
anne marie greig, head of content design
What I learned or would change
Structure beats polish. The first round of work made the copy better and the research said it would change nothing, because users were not confused, they were worried. Finding the right content model mattered more than finding the right words, and it was a cheap unmoderated card sort that found it.
Users also handed me the vocabulary. Nobody in the business was saying “data security” in product copy until 30 participants told us that was the category they cared about most. I now treat open card sorts as a language research tool, not just an IA one.
If I did it again I would push for an A/B test rather than a before-and-after so the content’s contribution was isolated from the marketing relaunch. And I would have looked harder at the entry points sooner: within weeks the squad noticed CRM emails saying “link your bank account” were dropping people on the landing page instead of the connection screen, which blunted some of the gain. The journey was fixed, but the promise made before the journey was not.