Background
When I started at NewRetirement, one of the first things we wanted to adjust was onboarding. The existing flow was dated, didn’t work on mobile, was overwhelming, and didn’t give users the best introduction to the product. I saw this as an area of high impact, and advocated to prioritize it on our roadmap. The team agreed.
Proposed Direction
I proposed a flow that had more steps but less data density on each screen, inspired by TurboTax. This would keep users focused on the task at hand without overwhelming them, and also enabled us to optimize for mobile.
Features
User Testing
I used unmoderated user testing to iteratively refine the flow. Both flows tested well and helped identify areas of confusion. The biggest question users had while going through these flows was about capturing edge cases that factored into their financial plans. For example, a couple of users wanted to know where to input passive income streams.
Rather than add more steps to the flow, we added tips and clarifying copy throughout the experience to help anticipate these questions and let users know they could enter additional information once their initial financial plan was created.
Over 5 days of testing with 60 participants, 95% of participants found the flow to use easy to use. This testing helped the stakeholder team gain the confidence needed to move forward with A/B testing the new flow.
A/B Testing
We tested both the quick and comprehensive paths against each other, measuring onboarding rate, and the plus upgrade rate to determine success.
Round 1
Quick vs. Comprehensive
The upgrade rate for the original onboarding was around 80%. With our A/B test we compared the comprehensive path to the quick path.
Despite the comprehensive path winning out, the stakeholder team was hesitant to to commit to using only that. They wanted to find a way to merge the two paths.
Round 2
Path Picker
To solve this, I proposed a path picker solution. Users could select which path they wanted to take (quick or comprehensive). Once chosen, the user would proceed through their selected flow.
We started a new A/B test comparing the path picker to the comprehensive path only. That test revealed the following:
When we dug deeper and looked at the 60% of users who selected comprehensive in the path picker their upgrade rate was 3.7% (higher than comprehensive only).
Round 3
Recommended Badge
With the knowledge that the comprehensive path of path picker was yielding a much higher upgrade rate we wanted to optimize for this. I added a simple “Recommended” badge to the comprehensive path option in the path picker, then we tested again.
The results were positive:
The team fully switched over to the Path picker flow.
Our onboarding flow has largely stayed the same, we’ve updated it aesthetically when we went through both phases of our rebranding and the success metrics have held firm.