Global biotechnology company
Scaling UX impact across 12 engineering teams
Coordinated a distributed UX initiative that helped 12 engineering teams introduce user research into their work, with different teams progressing through design-system adoption, independent research, and specialist collaboration.
12 participating engineering teams
Research artifacts
JTBD and personas created by all participating teams.
Integration or a path
Every team integrated the design system or planned an upgrade.
Specialist partnership
Three teams consulted UX for complex work; follow-through groups may overlap.
Two UX specialists and reusable research resources supported embedded engineering participants—one per team—who brought research and improvements into 12 teams across three verticals. Participation did not mean every team reached the same maturity.
01 / The challenge
High ambition. Limited specialist capacity.
Many internal applications were maintained by individual engineers, with fragmented interfaces and limited access to dedicated UX support. Modernization was a priority, but another large redesign program would have competed with existing delivery commitments.
Distributed ownership
12 engineering teams across three verticals, each with its own products and priorities.
Limited UX capacity
Two central UX specialists could not guide every application individually.
Initial skepticism
Engineering leads needed to see user value without taking on a competing program of work.
02 / The approach
Specialist guidance. Local ownership.
With my manager's support and approximately 10% of my time, I volunteered to coordinate participation. I worked with leads on capacity, facilitated learning, and connected engineers with the two UX specialists.
Engineers brought research and manageable improvements into their own backlogs. The specialists retained ownership of their discipline and remained available for complex work; I did not author the existing design system.
- Enable
Make interviews, JTBD and personas approachable.
UX specialistsInterviews · JTBD · personas · shared learning
- Embed
One engineering participant in each team.
Embedded engineers - Support
Keep specialist guidance available for complex needs.
- Apply
Bring research and improvements into local backlogs.
12 teams' local backlogs
Interviews · JTBD · personas · shared learning
Learning fed into local backlog work, with specialist support available when questions became more complex.
A custom GPT helped translate interviews into jobs-to-be-done (JTBD) and persona artifacts; it did not replace specialist judgment.
03 / Key decisions
Adapt the model. Make it easy to say yes.
The original participation plan did not survive reorganization. I adapted the model around engineering ownership, addressed skepticism with concrete user needs, and kept the commitment modest enough to fit existing priorities.
- Adapt after reorganization. A reorganization disrupted my original product-team participation plan. I shifted to one engineering representative per team, preserving a route into each team without relying on the former structure. The tradeoff: research needed to be approachable for engineers with limited UX experience.
- Address skepticism with user value. I connected observed user pain points to organizational priorities and a commitment agreed with leads: roughly three hours per engineer per month, including learning and backlog work. Useful, prioritized improvements could fit existing roadmaps rather than compete with them.
- Build capability, not dependency. Resources and specialist partnership helped engineers conduct routine research locally, with an expert path for harder problems. Independence was a direction to support—not an assertion that all 12 teams became self-sufficient.
04 / Results and reflection
Shared participation. Different paths forward.
All 12 teams participated and produced JTBD and persona artifacts. Additionally, every team either integrated the existing design system or established an upgrade path. Follow-through varied, rather than advancing uniformly across the network.
Organizational sponsorship and embedded ownership were both necessary. Leadership made room for the work, specialists provided guidance, and engineers carried the practices into their teams. My role was to connect those groups and make collaboration workable.
Scale capability, not dependency
The lesson I carry forward is to scale capability without making every decision depend on a central team.
- Teams participated
- 12All created JTBD and persona artifacts
- Teams completed upgrades
- 3Existing shared design system
- Teams continued research independently
- 3Beyond the structured initiative
- Teams consulted UX specialists
- 3For more complex work
Leadership principle
Lower the cost of participation, maintain trust in expertise, and give people a useful next step in their own work.