How rebuilding a design team's process from the ground up — starting with a principles workshop and a fully cross-functional kickoff — turned the most overwhelmed team in the org into its most prolific.
The Credit Karma Autos design team was technically functional but operationally broken. Two product designers — one full-time, one contractor — were fielding continuous project pressure from 8 product managers across insurance and financing. There was no research function, no design manager, and no shared process. To product leadership, design looked like the bottleneck. In reality, design was the symptom.
When design is excluded from problem definition, it can only execute — not solve. The Autos team was shipping products built on untested assumptions, learning they were wrong after launch, and absorbing the blame for outcomes shaped long before a designer touched them. A new design manager had just been brought in to change this — and she asked me to come with her.
The first thing I did wasn't design. It was facilitate.
I brought the designers, PMs, engineers, and marketing together for a principles workshop — the first time the full cross-functional team had been in a room together to define what they were trying to accomplish and how they wanted to work. We built a shared set of design principles that everyone had a hand in creating and could align behind going forward.
From that point, every new project started with a fully cross-functional kickoff that I facilitated — not product. The shift was deliberate. PM-led kickoffs in this culture tended to narrow the aperture early: fewer opinions, faster movement, less friction. But fewer opinions meant biased assumptions went unchallenged, and asynchronous information distribution meant engineers and designers were always catching up. Diverse perspectives at the start weren't inefficiency — they were the work.
The kickoff format changed what happened next:
In six months, the team shipped more successful products than any other vertical in the org. Every shipped product was grounded in research validated before significant engineering investment.
A foundational research artifact mapping the holistic car-buying journey and the emotional states of members at each touchpoint. It became the strategic foundation for the product work that followed.
A conversational experience gathering member insights about purchasing goals and financial constraints, generating the data needed to build the calculator.
Helped car buyers assess which vehicles fit their actual purchasing and total ownership budget, including financing, insurance, maintenance, and depreciation.
Simplified a complex insurance marketplace using personalized member data. Featured a personalized card component surfacing the specific data points influencing the member's insurance rates while shopping, plus social comparison components showing where they stood relative to members in similar financial situations.
Used LexisNexis data to connect the member's actual vehicle to their Credit Karma account, enabling budget tracking, value monitoring, mileage milestone predictions, and optimal trade-in timing recommendations.
An independent telematics app helping drivers with poor credit access lower insurance rates based on driving quality. It led telematics in the auto insurance industry.
Within 6 months the Autos team became the most prolific design team in the Credit Karma org — shipping more successful products, faster, than any other vertical.
Canceling projects that failed in research wasn't a failure — it was the process working. Every concept that didn't ship was an engineering investment that didn't get wasted. Engineers went from feeling unheard to contributing to problem definition at kickoff. Designers went from executing requirements to leading discovery. PMs went from designing in PRDs to collaborating in shared problem spaces.
The principles workshop and XFN kickoff format became a model for what design process could look like across the org. The Diamond Design Process we developed during this period laid out a tiered approach to design attention based on project type, a diverge/converge weekly rhythm keeping designers and partners in sync, and a reframe of the full XFN team as a force multiplier rather than coordination overhead.
The insight at the center of it: designing the right thing and designing the thing right are two different problems. Conflating them — moving straight from assumption to execution — is how teams stay busy without making progress.