#PRODUCT STRATEGY #PORTFOLIO STRATEGY #AI PRODUCT

Building a UX team from zero, in a space that didn't trust UX yet

Workday's Customer Deployment Tools org was a highly technical, high-stakes space — and it had no UX team and a bad history with the function. I built one from scratch, including making the harder call to move on an existing designer who wasn't the right fit, hiring for logic over pedigree, and coaching a struggling researcher through a confidence crisis rather than a skills one.

COMPANY
Workday

ROLE
UX Manager

TIMELINE
2018–2021

THE SITUATION

An open headcount of four, and a function with a bad reputation before I walked in

Customer Deployment Tools (CDT) was the org behind getting new Workday customers live — deployment, integrations, data migration. It was deeply technical, and deployment was slow and expensive: over a year for some segments, multiples of ACV in cost, with customer satisfaction taking a hit during onboarding. There was an open UX headcount of 4, and the org had a bad history with the function — UX had been tried before and hadn't landed. One designer was already in the org when I joined, and the relationship was strained.

WHAT WE DID

Making the harder call first

Before I hired anyone, I had to deal with who was already there. The existing designer was talented, but a more conceptual, creative designer in a domain that needed rigorous, logical, systems-level thinking. That mismatch — not effort or attitude — was the root of the strained relationship. I moved him into a space better suited to his strengths, where he went on to do well. Then I hired for the actual job: people with strong logic and a cross-skill across design and research, because we needed to move fast and couldn't afford handoffs between disciplines for everything.

Over the following year I hired three product designers and a researcher, building the team out quarter by quarter through 2019.

Coaching through a confidence problem, not a skills one.

The researcher struggled in her first few months. CDT was more technically complex than she'd expected, and under pressure she started doing what PMs asked rather than thinking for herself — at one point writing leading questions into a survey that made the results unusable. My first read was a skills gap. It wasn't — it was confidence. I started joining her research conversations, modeling the strategic questions for the brief myself rather than handing her the answers, and helping her connect her findings back to the specific problems each PM was working on. I gave her room to rebuild that muscle rather than taking the work back from her.

Building the operating system for a team that didn't exist yet

Beyond the people, the team needed infrastructure: project charters, a research audit, a defined CDT UX process, and a mindset shift across the wider org toward "design experiences, not features." We ran a week-long cross-functional "Improvathon" — 80+ engineers, 11 PMs, 3 UX, 4 end users, across two locations — to build empathy with real users and fix what surfaced, together. For teams with no UX support at all, we ran weekly CDT UX office hours so they could still get research and design help.

IMPACT

One strained relationship to a 5-person team in a year, cutting deployment time 45–50%

  • UX function built from a single strained relationship and an open headcount of 4, to a functioning 5-person team within about a year

  • The solutions the team designed contributed to a 45–50% reduction in total deployment time across the client examples we tracked (a design contribution alongside PM and engineering work, not a UX-only number)

  • 519 customers went live on Workday through CDT in 2019 alone; 167K individual customer migrations supported overall

  • The team continued to perform well after I left CDT in April 2021 — it didn't depend on me to keep functioning

WHAT THIS TAUGHT ME

Knowing which lever to pull is most of what building a mature team means

Not every people problem is the same problem. Amanda's was confidence, and the fix was presence and modeling, not instruction. The other designer's was fit, and the fix was moving him somewhere he could do his best work, not managing him harder.

The other lesson: in a domain with no track record for UX, the team has to match how the domain actually thinks. Hiring for logic and cross-skill is what let the team move at the pace the business needed.