Auto Save

Smarter savings through open banking

To enhance Zopa's Auto-Save feature by leveraging open banking to give customers greater flexibility and control when funding their savings. The goal was to simplify the process by enabling users to automatically split deposits across multiple savings pots, reducing manual effort while maintaining transparency and choice.

Defining what good looks like

We set clear, measurable goals focused on adoption, engagement, and customer value:

Understanding the landscape

I conducted a comprehensive competitive analysis across both incumbent and challenger banks, understanding how the industry approaches payment flows, identifying best practices, and uncovering opportunities to differentiate.

Plum uses AI to analyse spending habits and automatically transfers an affordable amount to savings "pockets." Monzo's round-ups feature saves spare change from transactions into designated Pots. Starling's Savings Spaces act as virtual piggy banks for different goals within the main account.

Insights workshop

To set the foundations, I ran a collaborative workshop with product, design, and research teams. The goal was to align on what we already knew about customer saving behaviours and identify the knowledge gaps that needed validation.

Closing the gaps

To validate our assumptions and uncover new insights, I proposed and conducted customer interviews. This had immediate buy-in, it was clear we had a number of unknowns we needed to understand before moving into design.

• 8 interviews with single-pot savers, multi-pot savers, and existing auto-save users
• Explored how they decide where to allocate deposits
• Tested comfort levels with automation vs. control
• Understood expectations around notifications, confirmations, and flexibility

Key takeaways

From research to solutions

With clearer understanding of customer behaviours, we transitioned to design. We didn't have immediate alignment on the focus, so we arrived at two hypotheses to shape the direction:

Wireframes

I started by exploring different approaches, experimenting with navigation patterns, allocation inputs (sliders vs. number fields), fund split visualisations, and placement within the journey. These explorations helped surface early insights and shape hypotheses for usability testing.

Testing the thinking

We ran remote usability testing with a diverse mix of customers. We explored three different journeys, a simple single-rule setup, a flexible multi-rule version, and a hybrid approach. Tests revealed that customers valued simplicity first, but also wanted optional flexibility through overrides.

Insights from testing

High-fidelity

After validating hypotheses, I moved into high-fidelity design to refine interaction patterns, visual hierarchy, and information density while maintaining clarity and trust around automation. The key challenge was balancing flexibility and simplicity.

Usability testing

I ran a museum tour with the core team to walk through the high-fidelity thinking. Through dot voting and discussion, we landed on a core experience we were happy to proceed with, but knew we needed additional testing for confidence.

Development handover

I worked closely with engineers and the product team to prepare a structured handover. I created detailed user stories capturing outcomes, design references, edge cases (overdraft conditions, limits), and dependencies between Auto-Save, Pots, and Smart Saver features.

Exceeding every target

The results far surpassed our initial goals, delivering meaningful value for both customers and the business.

Reflections

What went well

• Customer-centric decisions: Grounding every decision in research and usability testing let us confidently balance simplicity with flexibility, giving customers control while keeping things lightweight.
• Cross-functional collaboration: Close alignment with engineering, product, and compliance early on helped us move faster and avoid late-stage surprises. Joint workshops and clear communication reduced rework.
• Exceeding impact targets: We set ambitious goals around adoption, savings amounts, and inflows, the actual results far surpassed them across the board.

What I'd do differently

• Design system constraints: Zopa's existing design system wasn't built for complex multi-step financial flows. We had to extend components carefully, introducing allocation controls and summary patterns, while collaborating with the design system team for consistency and scalability.