A shared quality bar across four brands

Stakes

Paysafe’s consumer portfolio spans Skrill, Neteller, PaysafeCard and PaysafeCash, with shared capabilities across payments, wallets, engagement and regulated money movement.

Multiple roadmaps competed for the same design capacity. Senior coverage was thin. In-house designers and external partners needed to work as one practice. And quality had to remain consistent across experiences ranging from entertainment and engagement to onboarding, KYC and money movement.

In 2023 the design organisation was at its largest, with dedicated research, content and design-system roles. In August 2024 two consumer design teams merged into one under me. By October three senior designers were out. By January 2025 five designers covered four brands and the design system, and delivered 12 projects that quarter, before the team was rebuilt to a blended team of up to eight across three pods.

I grew with that problem: from the designer people came to for design-system decisions, to Lead Product Designer, Product Design Manager and eventually Senior Manager, Product Design.

• Capacity Attrition and a persistent senior gap meant junior, mid-level and partner designers needed to take meaningful ownership quickly.
• Complexity Four consumer brands, shared capabilities and parallel product roadmaps competed for the same design attention.
• Continuity Knowledge, standards and decision-making needed to survive team changes, handovers and restructuring.

My role

Decision log

The function evolved through a series of practical decisions made under constraint.

1. 01
   Build capability instead of waiting for perfect capacity

   A persistent senior hiring gap could easily have become a reason to slow down.

   Instead, I hired and onboarded junior and mid-level designers, gave them meaningful product ownership with structured support, and used external partners where specialised capacity was needed.

   The cost sat with me. Without a senior layer, most design reviews ran through me for much of 2024 and 2025, and when capacity couldn't cover a request, I declined it rather than let the bar slip.

   In September 2024 I told leadership in writing that the team could not commit to the same output with three senior designers out, and reset priorities with stakeholders around the capacity we had.

2. 02
   Make learning part of the operating rhythm

   Knowledge was too important to remain in individual files or disappear when someone changed teams.

   I established ProtoHub, Learning Camp and Self-Learning Hour as recurring ways to share research, craft and product knowledge. Bug-finding sessions, critique and clearer Figma organisation helped turn individual learning into team capability. We reorganised around 300 design files into one structure and naming standard, and each designer had written yearly goals by March.

   The purpose was practical: reduce dependency on tribal knowledge and make onboarding, collaboration and handover easier.

3. 03
   Make progression explicit

   When expectations between levels are unclear, feedback becomes subjective and career conversations become harder than they need to be.

   I created and used career-level expectations from intern through management as a shared reference for hiring, development conversations and promotion readiness.

   That gave designers clearer ownership of their growth and managers a more consistent basis for feedback and progression decisions.

4. 04
   Make quality inspectable before release

   In January 2025 I asked leadership for two rules: design involved at least two weeks before work starts, and a mandatory design review before production. The first became how we worked. The second was never mandated, so I wrote the standard down and ran it inside the team, because quality could not depend on experienced designers catching every problem late in delivery.

   The written bar came with clearer entry expectations and PASS / RISK / FAIL reviews — Design Gates & UX Quality Bar →. The bar also worked in reverse, on live product: in 2026 the team assessed PaysafeCard registration against it (full story →).

5. 05
   Use AI to strengthen the practice, not bypass judgement

   As AI tools entered the design workflow, I treated them as team capability rather than novelty. Human checkpoints stayed around problem framing, prioritisation, quality and final decisions — full AI case →.

What changed

• 4 → 1 design systemsConsolidated four fragmented design systems into a shared consumer foundation with 50+ core components and 100+ sub-components
• PASS · RISK · FAILWritten quality bar making unresolved UX risk visible before release pressure took over
• Designers owned their workEach designer presented their own work ahead of PI planning. Research and design-system ownership sat with designers, not with me.
• 80% on timeProjects delivered on time in the two months after the merge, with three senior designers out

The clearest signal was who held the work. A designer I mentored went on to run usability research for PaysafeCard, and the design system changed owners without stalling.

The product case studies show what this made possible: Sports Corner and trust-critical journeys were both designed and shipped inside this operating model.

In April 2026, I was promoted to Senior Manager, Product Design, recognised for contribution during a period of significant organisational change.

Later in 2026 Paysafe wound down its in-house consumer design team, and I led the handover. My final piece of work was making sure the practice outlasted the team: named owners and next steps for 50+ live initiatives, and written continuity plans for the design system and accessibility work.

Reflection

The most important shift in my own leadership was moving from improving individual design decisions to improving the environment in which those decisions were made. Strong craft still matters — but across multiple products and teams, quality cannot depend on one experienced person noticing every problem.

Hiring well, teaching deliberately, making expectations visible and introducing lightweight checkpoints gave designers more independence while keeping the bar clear.

What I would take elsewhere: explicit standards, meaningful development, evidence before opinion, and enough structure to expose risk early. The mechanics should always adapt to the organisation — good operating systems make strong work easier; they should never become the work themselves.