Designing Trust at Scale: Open Banking Infrastructure Across GCC Banks

The funnel showed where people stopped. The interface helped explain why.

The baseline funnel recorded 530 sessions at launch, 282 at provider selection, and 136 at consent. It showed substantial loss before people reached the bank-authentication handoff, giving us a clear part of the journey to investigate.

Alongside the user sessions and support feedback described in the original study, I reviewed what the interface was asking people to understand. The central design problem was confidence: recognising the bank, understanding the request, and feeling in control before continuing.

As Lead Product Designer, I led the design work with partner banks and worked across research review, design exploration, user comparisons, and the detailed states prepared for development.

The old journey brought sensitive risks into focus before building confidence.

The permission screen explicitly named payments, bank usernames, and passwords in a section explaining what access did not allow. That reassurance was factually intended to reduce concern, but it put the most alarming possibilities directly in the user’s reading path.

The user also had to connect several separate ideas: a service requesting access, a 12-month period, a generic “Confirm” action, a redirect, and an account-selection screen. Each step could leave a different question unresolved.

I organised the redesign around those questions, making the scope, purpose, control, and next step visible at the moment each mattered.

Make the bank feel familiar before asking for permission.

The earlier list already included bank logos, but presented them as small entries in a dense sequence. I gave each bank more visual space, with a larger recognition target and a consistent relationship between its logo and name.

This was the first opportunity to make the flow feel deliberate and understandable. The same card system could accommodate the bank sets for Saudi Arabia and Bahrain while preserving a familiar interaction.

Explain the access being requested, in language people can act on.

I shifted the emphasis toward a concrete explanation of the request: the bank being connected, Tarabut acting on behalf of the requesting service, why data is needed, and the categories of data involved.

The final screen separates that explanation from the legal terms. Users can expand a category to inspect its contents while keeping a manageable overview. The interface leads with what the request means, allowing the person to judge it before moving on.

Bring the sharing period and the right to revoke into the decision.

The earlier design already mentioned a 12-month period and the ability to revoke consent. I made that control more visible: explicit sharing dates and a dedicated revocation notice appear below the data categories and close to the primary action.

The expanded view adds examples of the data within each category. Users can see the overview, inspect the details, and return to the same decision without losing the sharing period or the explanation of control.

Explain where the user is going, then help them find their way back.

The earlier redirect paired a bank handoff with “get approved instantly,” introducing another outcome while the user was still connecting an account. I kept the transition focused on the destination and the connection task.

A short “How it works” explanation introduces the sequence before consent. On the return journey, the interface names the original app and explains that the user has been logged out of the bank. Together, these states make the boundary between organisations easier to follow.

Support meaningful choices without hiding the requirements.

The permission system had to accommodate differences between markets and use cases. The Bahrain flow documents account details as required, while transaction details depend on the service being provided. The developer notes identify income checks and personal finance categorisation as cases that need transaction data.

I designed the responses to those choices: a required-data explanation, a route that allows limited data where supported, and an inline message when the selection is insufficient to continue.

This made the handoff more specific than a set of ideal screens. It connected the interface to the conditions under which a user could proceed, and gave market-specific disclosures a defined place within the layout.

Keep the same reassurance in both reading directions.

Arabic and English were designed across the journey, including selection, permissions, help, and the handoff states. Permission descriptions, dates, navigation, and legal copy all had to remain readable within the same system.

Placing the versions side by side makes this work visible: text aligns with the reading direction, navigation follows it, and the relationship between the bank, the request, and the primary action stays clear.

Preserve the decision as the layout changes.

On desktop, the same content takes advantage of a wider canvas. Bank options sit within a contained grid; the permission layout separates the request from the sharing summary and action. On mobile, those elements follow a single reading sequence.

The handoff includes desktop and mobile designs, alongside intermediate viewport layouts for permissions, so the experience is defined beyond one ideal screen size.

Keep the user oriented when authentication moves to another device.

The handoff also covers a connection that starts on one device and uses another for bank authentication. I designed a choice between continuing on the current device and logging in elsewhere, followed by a QR screen that explains how to make that transition.

The original device then shows a waiting state that explains where the user is in the process. On the second device, the completion screen confirms the account connection, states that the user is logged out, and tells them to return to the original device.

Designing those states made the start, wait, and finish of the connection visible on both sides of the handoff.

People preferred the clearer structure.

In the documented Arabic bank-selection comparison, 77% of 26 participants preferred the grid layout. The accompanying feedback repeatedly mentions clarity, organisation, and the visibility of bank icons.

A separate Arabic permission-layout comparison recorded a 63% preference among 30 participants for the design with more visibly grouped information. These comparisons helped inform the direction. The final expandable permission system continued to develop beyond the layouts shown in that test.

These are preference results. They give evidence about the designs people favoured, rather than directly measuring journey completion or comprehension.

A more understandable request. A more confident journey.

The project’s reported outcome was a 27% improvement in consent-journey completion. The redesign also established reusable patterns for bank selection, permission review, market-specific requirements, and device handoffs.

The central contribution was making confidence part of the interaction: a recognisable bank, an understandable request, visible control, and a clear path through authentication and back to the app. Each decision addressed a place where uncertainty could make someone stop.