A redesigned onboarding flow cut bounce rate 36.4%

A redesigned onboarding flow cut bounce rate 36.4%

A redesigned onboarding flow cut bounce rate 36.4%

04
RESPONSIBILITIES
UX Design
UI Design
Design System
UI Design
UI Design
INTRO

TeamPassword is a cloud password manager for teams. You store login credentials in one place, share them securely, and reach them from web, desktop or mobile. It's useful as soon as it's set up, and that was the problem. Too many people left during sign-up and onboarding, before they got to the part that works.

I led the design and UX on a rebuild of that flow. We started from the data. Google Analytics showed us where people dropped out, and the client had A/B test feedback on top of that, so we redesigned the path from the first screen to a working account around the points that were losing people.

KEY CHALLENGES
Redesign the sign-up and onboarding flow around where users actually dropped out

Redesign the sign-up and onboarding flow around where users actually dropped out

Base the changes on Google Analytics data and A/B test feedback

Base the changes on Google Analytics data and A/B test feedback

Build a material design library on the brand guidelines so screens stay consistent

Build a material design library on the brand guidelines so screens stay consistent

That work turned up a second problem. The experience wasn't consistent from one screen to the next, so I built a material design library aligned to the TeamPassword brand. After that, every step of onboarding and everything past it looked and behaved the same.

THE OUTCOME

After the new flow shipped, bounce rate through sign-up and onboarding dropped 36.4%. The design library came out of the same project. Once the flow was consistent, the components behind that consistency became the system the rest of the product could build on.

THE THINKING

A SaaS product tends to lose people in onboarding for unglamorous reasons: one step too many, a screen that doesn't match the one before it, a beat of hesitation about what to do next. You can't see any of that by designing on instinct. It sits in the analytics, which is where we started.

Reading the drop-off data first meant the redesign fixed the steps that were costing users, instead of the ones that happened to look worst in a design review. The design library did the less visible part of the job. A flow can convert perfectly well and still feel shaky if every screen is a little different, and with a password manager, that consistency is part of whether people trust you with their credentials.