Our original flow required all new users to open an invest account first. That was a product decision made early in the company's history, but it no longer reflected how users were actually engaging with the platform. Users who wanted to add a second product later faced friction: re-entering information, navigating flows that weren't designed for their context, losing momentum.

The fix was straightforward in concept: give users a choice upfront. The compliance requirements around that choice made it significantly more complex. Each account type carried its own KYC path, which meant a decision on screen one cascaded through 18 legally-required screens downstream.

Early in the project I made a mistake I see a lot of designers make, including myself before this. I tried to design for everyone. I built flows that would accommodate every possible user type, every edge case, every path through our account types. The result was a sprawling decision tree that didn't do anything well.The insight that changed the approach came from the data. Roughly 80% of M1's revenue came from the top 20% of users, and that segment skewed toward financially literate people who were already comfortable with investment apps. They weren't looking for a tutorial. They were looking for a fast, confident path to getting their account open and their money moving.

That narrowed our focus significantly. Rather than designing for the user who had never opened a brokerage account before, we optimized for the user who had. That decision shaped every subsequent call on copy, screen count, and information density.

We decided to lead with a combined invest and spend account option. The complication was that invest and spend were different account types under the hood, each requiring their own KYC path. A user choosing both effectively needed to complete two compliance flows. The design challenge was building a single, coherent onboarding experience that could accommodate either selection without making the user feel like they were starting over.

The Social Security number collection screen had the highest drop-off rate in the entire flow. Our first instinct was to address it by introducing trust builders and social proof, either on the screen or just before it. We tested the hypothesis and found the opposite was true: any additional screen added friction. Users who had made it that far through KYC didn't need more convincing. They needed to get through it quickly.

What did move the needle was transparency on the screen itself. We added explanatory copy directly on the SSN screen explaining exactly why we needed the number and how it was used, rather than hiding that information behind an info icon. Making the reasoning visible by default reduced hesitation without adding steps.

The core trust builder ended up being speed. Getting users to that screen faster, with less buildup, performed better than any reassurance we put around it.