eBay - Design Process

eBay - Design Process

Design Operations, Design Process

Design Operations, Design Process

2023

2023

Sr. Design Manager

Sr. Design Manager

Project info

Rewriting the Playbook: A Case Study on Redesigning the Design Process at eBay

When I joined eBay, the design org was full of talented, passionate people—but the way they worked was anything but unified. Each team operated in its own silo, with its own tools, rituals, and definitions of success. There was no shared blueprint for how design should happen, no consistent process guiding teams from insight to execution.

The result? Inconsistent qualityfragmented experiences, and an organization struggling to scale good design practices across products and platforms.

My charter was to change that: to create a unified, scalable design process that worked for everyone.

But I knew from experience—you can’t impose process from the top. It has to be built with the people who use it. This is the story of how we did that.


Step 01: Researching the Landscape

Before jumping into solutions, we committed to understanding the full picture.

We launched a deep discovery phase that included:

  • Dozens of interviews with engineersdesigners, and program managers

  • Surveys across the global design organization

  • Audits of design files, specs, documentation, and code repositories

We looked at everything. Our goal wasn’t just to identify broken processes—we wanted to understand the root causesand the unique needs of every team. What we uncovered was a complex ecosystem of overlapping toolsinconsistent rituals, and significant variation in design maturity from one team to the next.

We captured it all in a detailed Current State Report—a document that painted a clear picture of how design was actually happening across eBay, and where the most critical gaps were.


Step 02: Designing Together

Armed with a shared understanding of the problem, we moved into solution mode—but we didn’t do it alone.

We created a cross-functional Design Process Task Force, bringing together stakeholders from designengineering, and product. The goal wasn’t just to define a process that could work in theory—it had to work in practice, for everyone.

This task force played two roles:

  • First, as co-creators, helping us shape the design process based on real-world needs

  • Then, as governors, maintaining and evolving the process after it was launched

The result was a sense of shared ownership. This wasn’t a top-down directive. It was a process the teams had built themselves.

In parallel, we also established a Design System Council—a governance body focused specifically on the design system component library and its evolution, ensuring quality and consistency across platforms.


Step 03: Creating the Playbook

A process isn’t truly usable unless it’s clearly documented and easy to follow. So we created the Design Best Practices Playbook—a comprehensive guide that outlined the new process, from kickoff to handoff.

The playbook included:

  • Clear design stages and decision points

  • Expectations for cross-functional collaboration

  • Guidelines for documentationspecs, and handoff quality

  • Links to supporting tools and resources

It didn’t just tell teams what to do—it showed them how, and why. Paired with the custom tools we built (detailed in a separate case study), the playbook helped ensure that the new process was practicaladoptable, and scalable.


Outcome: From Fragmentation to Flow

This wasn’t just a process redesign. It was an organizational shift—from isolated teams working in parallel to a collaborative network guided by shared principles.

We:

  • United siloed teams under a common framework

  • Brought designersengineers, and PMs into the process design itself

  • Created a playbook that made the process easy to adopt and scale

  • Formed a Design Process Task Force to define and govern the new approach

  • Established a Design System Council to guide design system evolution

  • Laid the groundwork for automation and AI-powered workflows in the future

This work wasn’t about perfection—it was about progressmomentum, and building the kind of operational maturitythat design at scale requires.

And it worked—because we built it together.

Rewriting the Playbook: A Case Study on Redesigning the Design Process at eBay

When I joined eBay, the design org was full of talented, passionate people—but the way they worked was anything but unified. Each team operated in its own silo, with its own tools, rituals, and definitions of success. There was no shared blueprint for how design should happen, no consistent process guiding teams from insight to execution.

The result? Inconsistent qualityfragmented experiences, and an organization struggling to scale good design practices across products and platforms.

My charter was to change that: to create a unified, scalable design process that worked for everyone.

But I knew from experience—you can’t impose process from the top. It has to be built with the people who use it. This is the story of how we did that.


Step 01: Researching the Landscape

Before jumping into solutions, we committed to understanding the full picture.

We launched a deep discovery phase that included:

  • Dozens of interviews with engineersdesigners, and program managers

  • Surveys across the global design organization

  • Audits of design files, specs, documentation, and code repositories

We looked at everything. Our goal wasn’t just to identify broken processes—we wanted to understand the root causesand the unique needs of every team. What we uncovered was a complex ecosystem of overlapping toolsinconsistent rituals, and significant variation in design maturity from one team to the next.

We captured it all in a detailed Current State Report—a document that painted a clear picture of how design was actually happening across eBay, and where the most critical gaps were.


Step 02: Designing Together

Armed with a shared understanding of the problem, we moved into solution mode—but we didn’t do it alone.

We created a cross-functional Design Process Task Force, bringing together stakeholders from designengineering, and product. The goal wasn’t just to define a process that could work in theory—it had to work in practice, for everyone.

This task force played two roles:

  • First, as co-creators, helping us shape the design process based on real-world needs

  • Then, as governors, maintaining and evolving the process after it was launched

The result was a sense of shared ownership. This wasn’t a top-down directive. It was a process the teams had built themselves.

In parallel, we also established a Design System Council—a governance body focused specifically on the design system component library and its evolution, ensuring quality and consistency across platforms.


Step 03: Creating the Playbook

A process isn’t truly usable unless it’s clearly documented and easy to follow. So we created the Design Best Practices Playbook—a comprehensive guide that outlined the new process, from kickoff to handoff.

The playbook included:

  • Clear design stages and decision points

  • Expectations for cross-functional collaboration

  • Guidelines for documentationspecs, and handoff quality

  • Links to supporting tools and resources

It didn’t just tell teams what to do—it showed them how, and why. Paired with the custom tools we built (detailed in a separate case study), the playbook helped ensure that the new process was practicaladoptable, and scalable.


Outcome: From Fragmentation to Flow

This wasn’t just a process redesign. It was an organizational shift—from isolated teams working in parallel to a collaborative network guided by shared principles.

We:

  • United siloed teams under a common framework

  • Brought designersengineers, and PMs into the process design itself

  • Created a playbook that made the process easy to adopt and scale

  • Formed a Design Process Task Force to define and govern the new approach

  • Established a Design System Council to guide design system evolution

  • Laid the groundwork for automation and AI-powered workflows in the future

This work wasn’t about perfection—it was about progressmomentum, and building the kind of operational maturitythat design at scale requires.

And it worked—because we built it together.

Examples

Design Process Audit: Mapping the Landscape. We began by conducting a comprehensive audit of all existing design workflows across eBay. Using a standardized template, we collected and analyzed team-specific processes—spanning research, design, content, and engineering—to identify overlaps, inconsistencies, and gaps. This qualitative and quantitative effort revealed the need for a unified operating model. The audit served as a critical baseline, ensuring the redesigned process reflected both the reality of day-to-day work and the aspirations of cross-functional teams.

Design Process Audit: Mapping the Landscape. We began by conducting a comprehensive audit of all existing design workflows across eBay. Using a standardized template, we collected and analyzed team-specific processes—spanning research, design, content, and engineering—to identify overlaps, inconsistencies, and gaps. This qualitative and quantitative effort revealed the need for a unified operating model. The audit served as a critical baseline, ensuring the redesigned process reflected both the reality of day-to-day work and the aspirations of cross-functional teams.

Survey Findings: Organizational Misalignment and Process Gaps. To inform the redesign of the eBay design process, we conducted a comprehensive survey that exposed critical breakdowns in alignment and execution. Only 47.5% of designers reported consistently following the official design process, while 31% hadn’t even heard of it. Just 25% said they consistently received sufficient information—such as business requirements, success metrics, or technical constraints—at project kickoff. Confidence in cross-functional collaboration was also low, with 47% unsure how to engage with design technology, 40% with content, and 32% with the design system team. These findings highlighted the need for a unified process model, clear engagement protocols, and shared expectations across teams.

Survey Findings: Organizational Misalignment and Process Gaps. To inform the redesign of the eBay design process, we conducted a comprehensive survey that exposed critical breakdowns in alignment and execution. Only 47.5% of designers reported consistently following the official design process, while 31% hadn’t even heard of it. Just 25% said they consistently received sufficient information—such as business requirements, success metrics, or technical constraints—at project kickoff. Confidence in cross-functional collaboration was also low, with 47% unsure how to engage with design technology, 40% with content, and 32% with the design system team. These findings highlighted the need for a unified process model, clear engagement protocols, and shared expectations across teams.

Phase 1: Discovery – Aligning on the Problem Space. The Discovery phase centers on surfacing and articulating the customer problem through cross-functional collaboration. Design, product, and external partners work as a unified pod to define goals, success metrics, and guiding principles. Key activities include stakeholder kickoff meetings, design sprints, competitive analysis, and the use of standardized templates (e.g., JTBD, E2E journey maps). The outcome is a well-formed design brief grounded in customer insights and business objectives, setting the stage for aligned and focused design execution.

Phase 1: Discovery – Aligning on the Problem Space. The Discovery phase centers on surfacing and articulating the customer problem through cross-functional collaboration. Design, product, and external partners work as a unified pod to define goals, success metrics, and guiding principles. Key activities include stakeholder kickoff meetings, design sprints, competitive analysis, and the use of standardized templates (e.g., JTBD, E2E journey maps). The outcome is a well-formed design brief grounded in customer insights and business objectives, setting the stage for aligned and focused design execution.

Phase 2: Define – Aligning on Goals and Direction. In the Define phase, teams synthesize insights gathered in Discovery to explore design directions and align on product objectives. Design, product, and engineering pods work together to evaluate feasibility, timeline, and scoping requirements. Tools like prioritization mapping and design kickoff templates help teams clarify the BRD and PRD. The output of this phase is a well-scoped direction, enabling early alignment with downstream stakeholders and setting a foundation for efficient implementation.

Phase 2: Define – Aligning on Goals and Direction. In the Define phase, teams synthesize insights gathered in Discovery to explore design directions and align on product objectives. Design, product, and engineering pods work together to evaluate feasibility, timeline, and scoping requirements. Tools like prioritization mapping and design kickoff templates help teams clarify the BRD and PRD. The output of this phase is a well-scoped direction, enabling early alignment with downstream stakeholders and setting a foundation for efficient implementation.

Phase 3: Explore – Iterating Toward a Design POV. In the Explore phase, cross-functional teams refine design approaches through iterative internal reviews and stakeholder feedback. Designers prototype and test multiple UI directions—often requiring support from horizontal teams like Design Tech or Systems for scalable component solutions. This phase includes formalized checkpoints such as Lead Reviews and Director POV Reviews, ensuring alignment at every level. BRDs and PRDs are finalized here, and the outcome is a clear, vetted design solution with executive and partner team buy-in, ready to enter the refinement stage.

Phase 3: Explore – Iterating Toward a Design POV. In the Explore phase, cross-functional teams refine design approaches through iterative internal reviews and stakeholder feedback. Designers prototype and test multiple UI directions—often requiring support from horizontal teams like Design Tech or Systems for scalable component solutions. This phase includes formalized checkpoints such as Lead Reviews and Director POV Reviews, ensuring alignment at every level. BRDs and PRDs are finalized here, and the outcome is a clear, vetted design solution with executive and partner team buy-in, ready to enter the refinement stage.

Phase 4: Refine – Finalizing Design Solutions and Specs. The Refine phase focuses on maturing the design POV into a polished, implementation-ready solution. Designers partner with the product pod and cross-functional teams to evaluate feasibility, scope releases, and ensure accessibility across use cases and locales. This stage formalizes the handoff with detailed spec documentation using standardized templates and tools (e.g., Figma templates, accessibility and handoff checklists). Final designs are locked, timelines are confirmed, and work is phased across release cycles to optimize delivery without compromising quality.

Phase 4: Refine – Finalizing Design Solutions and Specs. The Refine phase focuses on maturing the design POV into a polished, implementation-ready solution. Designers partner with the product pod and cross-functional teams to evaluate feasibility, scope releases, and ensure accessibility across use cases and locales. This stage formalizes the handoff with detailed spec documentation using standardized templates and tools (e.g., Figma templates, accessibility and handoff checklists). Final designs are locked, timelines are confirmed, and work is phased across release cycles to optimize delivery without compromising quality.

Phase 5: Build – Partnering for Implementation and Quality Assurance. In the Build phase, design shifts into a support role as PD teams begin implementation. Designers provide handoff files, QA documentation, and remain available for consultation. Dedicated tools like QA/VQA trackers and “source of truth” templates ensure teams stay aligned on quality and scope. Cross-functional rituals (standups, bug bashes) help surface last-mile issues, while product pods manage timelines and final delivery. This phase ensures builds meet quality thresholds and that fast-follow work is scoped and prioritized effectively.

Phase 5: Build – Partnering for Implementation and Quality Assurance. In the Build phase, design shifts into a support role as PD teams begin implementation. Designers provide handoff files, QA documentation, and remain available for consultation. Dedicated tools like QA/VQA trackers and “source of truth” templates ensure teams stay aligned on quality and scope. Cross-functional rituals (standups, bug bashes) help surface last-mile issues, while product pods manage timelines and final delivery. This phase ensures builds meet quality thresholds and that fast-follow work is scoped and prioritized effectively.

Phase 6: Ship & Learn – Driving Continuous Improvement Post-Launch. The final phase centers on validating outcomes against success metrics and surfacing post-launch insights. Design and product teams conduct structured retrospectives, leveraging tools like GCX learning sessions and analytics dashboards (e.g., Tableau) to track ramp-up and customer experience. This phase institutionalizes learning through documented shareouts, templated rituals, and stakeholder alignment. Insights are folded back into future work, completing the feedback loop and ensuring continuous refinement of both process and product.

Phase 6: Ship & Learn – Driving Continuous Improvement Post-Launch. The final phase centers on validating outcomes against success metrics and surfacing post-launch insights. Design and product teams conduct structured retrospectives, leveraging tools like GCX learning sessions and analytics dashboards (e.g., Tableau) to track ramp-up and customer experience. This phase institutionalizes learning through documented shareouts, templated rituals, and stakeholder alignment. Insights are folded back into future work, completing the feedback loop and ensuring continuous refinement of both process and product.

Design System Audit Tracker – A Confluence dashboard used to manage the comprehensive audit of eBay’s design system. Every component review task was tracked in Jira and linked here, following a structured agile process with daily scrums, retrospectives, and clear ownership. This system ensured accountability, cross-team visibility, and steady progress toward improving design quality and consistency.

Design System Audit Tracker – A Confluence dashboard used to manage the comprehensive audit of eBay’s design system. Every component review task was tracked in Jira and linked here, following a structured agile process with daily scrums, retrospectives, and clear ownership. This system ensured accountability, cross-team visibility, and steady progress toward improving design quality and consistency.