Personalized Goal-Driven Onboarding

My role

I led a team of three designers across this initiative, end to end - discovery, strategy, and delivery. I set the design direction (segment by intent, deliver value early, defer friction), ran weekly critiques to keep the team aligned as the flow evolved, and served as the primary design voice in front of PMs, engineers, content, and executive stakeholders. I coached the two designers on my team through a high-visibility initiative - giving direction on flow structure while pushing them to own the rationale behind their decisions when presenting to stakeholders, not just the screens.

I stayed hands-on in the critical flows myself, particularly the initial goal-selection moment and the branching logic that followed it, because that's where the whole strategy either worked or didn't.

Designing within a new technical constraint

We were building this on server-driven UI (SDUI), a development approach that was new to our engineering org at the time. As the component libraries were still being built out, I made the call to adapt parts of the design to what the new SDUI libraries could actually support, rather than hold out for pixel-perfect fidelity. That meant simplifying some branching interactions and working closely with engineering earlier than I normally would, so the design system and the new technical foundation matured together instead of the design outrunning what could ship.

Deferring friction against business pressure

The instinct on the business side was to front-load data collection and product education early, while attention was highest. I pushed to defer that friction instead - get users to their stated goal first, and layer in credit education and account setup after they'd already found value. That was a bet against the standard onboarding playbook, and it meant defending a slower path to some business metrics in exchange for a stronger activation signal downstream.

Spending political capital on onboarding as a growth lever

Onboarding wasn't seen as a strategic investment before this - it was a solved problem nobody revisited. I made the case to executive stakeholders that it was actually the highest-leverage place in the product to invest, and framed the trade-offs so the team could ship with real conviction instead of shipping around competing opinions.

The personalization layer

This wasn't a static flow with a few branches bolted on - it was goal-driven by design. During discovery, we identified distinct user motivation profiles (a user optimizing a credit score behaves very differently from one shopping for a loan) through journey mapping and research, and mapped the onboarding experience to those profiles rather than to a single generic path.

The experience adapted dynamically to each user's stated goal at intake, surfacing relevant features and content and skipping what wasn't relevant to them - reducing cognitive load while increasing activation for each segment. It's rule-based personalization rather than machine-learned, but the underlying discipline is the same: use what you know about a user's intent to decide what they see next.

I designed the goal taxonomy and the branching logic to be extensible on purpose, not as an afterthought - a foundation that could eventually support algorithmic routing as usage data matured, rather than a one-off flow that would need to be rebuilt from scratch to get there.