Building a UX function from the ground up as economic infrastructure. I joined as UX Manager reporting to the Director of Financial Products, inherited an execution-only design team, and restructured it into a governance-driven, P&L-connected organizational capability operating across 15 products and 7 channels.
I joined Midway as UX Manager reporting directly to the Director of Financial Products. What I found was a design team operating exclusively as an execution layer: designers assigned to squads, responding to briefs, shipping screens. No governance structure. No research function. No design system. No mechanism connecting design decisions to business outcomes. The function existed without the organizational capability to generate consistent, measurable output.
Seven designers distributed across squads without governance, without shared standards, without research capability, and without executive visibility. Design decisions were made locally, inconsistently, and without connection to P&L. The function was a cost center with no measurable return.
Design generates economic value when it operates as infrastructure: standardized, governed, measurable, and connected to revenue and risk signals. The transformation required building that infrastructure: team structure, research function, design system, and governance model. Each element was a prerequisite for reliable results at scale.
Scale the team from 7 to 11. Create UX Research & Strategy and Content Design as formal organizational functions. Implement a governed design system integrated with engineering. Establish formal governance via Azure, RACI, and design review rituals. Then connect all of it: journeys, blueprints, and research to P&L signals in real time.
Organizational transformation at this scale requires simultaneous construction across multiple dimensions. The interventions were not sequenced. Governance, team, research, and design system were built concurrently, each reinforcing the others. The decision to move in parallel rather than sequentially was itself a capital allocation choice: compressing the transformation timeline reduced the period during which the organization operated without full capability.
"Sequential infrastructure construction extends the period of partial capability. Every week without governance is a week of uncontrolled execution accumulating as technical and organizational debt."
Leadership principle · Midway · 2023The four interventions were: team scaling and capability design, formal governance architecture, UX Research & Strategy as a new organizational function, and Design System as economic infrastructure. Each is detailed on the right. Together, they produced a function capable of operating across 15 products and 7 channels simultaneously without proportional headcount growth.
Scaled from 7 to 11 designers. Structured seniority distribution (Jr → Specialist) to match domain complexity. Assigned designers to critical product domains rather than generic squads. Eliminated the single-point-of-failure structure where any departure collapsed coverage of a channel.
Implemented formal governance via Azure DevOps, structured roadmap with ownership and entry criteria, RACI across all critical delivery fronts, and Design Review as a mandatory quality gate integrated with Tech. Governance was institutionalized as a non-negotiable operating standard.
Created the UX Research & Strategy function from zero. Standardized the Double Diamond as the institutional design process, a required standard rather than a recommendation. Conducted 50+ research studies in the period and developed 15+ service blueprints detailing the entire operation across multiple touchpoints. Blueprints integrated physical and digital journeys, NPS signals, system logs, and research insights to accelerate strategic decision making and AI assisted insight generation. Established a filter for research alignment: studies not connected to strategic priorities were declined, preserving capacity for high-ROI work.
Built centralized component library with formal documentation, mandatory usage standards, DS Commit governance flow, Storybook integration, and direct code repository connection. Established a DS committee with structured adoption metrics per squad. Design System was framed and managed as a capital asset, not a tooling project.
Organizational transformation generates friction. The interventions described above arrived in an organization with established power structures, competing priorities, and existing operating patterns. The governance model held because specific decisions were made with clarity, documented, and enforced. Three conflicts were defining.
Organizational transformation is only credible when it produces measurable economic outcomes. The interventions built above generated returns across three categories: revenue performance, operational efficiency, and capital efficiency through system leverage.
The operational difference between a team and a system is predictability. A team produces output when individuals are performing well. A system produces output consistently, because infrastructure, governance, and standards operate independently of any single person. That was the design goal from the start.
"A design function that depends on the manager's presence to maintain quality is not a function. It is a dependency. The infrastructure built was designed to operate independently of any individual."
Retrospective · Midway · 2025The organizational model operated 11 designers across 15+ simultaneous critical fronts: Midway App, Riachuelo App, physical store (PDV), Cockpit, WhatsApp, and Web. No middle management layer. Direct executive interface with Tech, Product, Risk, and C-suite was maintained throughout. Complexity was managed through governance architecture, not through headcount.
The Midway transformation is replicable. The conditions that made it possible, and the decisions that defined it, are transferable to any regulated product environment operating at scale.
Ungoverned execution accumulates as organizational debt: rework, misalignment, quality variance, and escalation cost. Investing in governance architecture early is not overhead: it is risk reduction with a measurable return. The Design Review ritual and DS governance paid back their implementation cost within the first quarter of operation.
Refusing scope creep, declining misaligned research, and establishing role clarity under active resistance are capacity protection decisions with direct impact on delivery quality. Organizations that cannot maintain role clarity cannot build scalable functions. The capacity freed by saying no is always redeployed to higher-ROI work.
When framed as infrastructure with adoption metrics, governance, and quantified economic impact, the Design System secured executive sponsorship and engineering integration. The system generated measurable capacity gains without budget expansion. That is a capital efficiency outcome, not a design quality argument.