Digital Banking for Startups

Digital Banking for Startups

In late 2021, I joined JPMorgan Chase's innovation team as the lead designer on a new product: a banking platform built specifically for startup founders. The brief was simple on the surface — make it easier for founders to open a business account. What we actually had to solve was harder: one of the world's largest banks trying to become relevant to a customer it had never really designed for.

By the time I joined, the team had already gathered stories from dozens of founders. I added another 20 or 30 conversations to the mix. As we listened to nearly 70 founders, a pattern began to take shape.

What the Research Revealed

Most business banking products seem to assume that founders want to actively manage their banking. What we heard, again and again, was the opposite.

Founders, it turns out, operate in two modes. There's the work that energizes them: building the product, talking to customers, closing deals. That's the core. Then there's everything else that comes with running a company: payroll, compliance, expense tracking, moving money. That's the chore.

Banking, almost entirely, lives in the chore category. Founders don't want a richer banking experience. They want banking to get out of the way.

This became the organizing principle for everything we designed. The question stopped being "how do we make banking better for founders?" and became "how do we make banking invisible to them?"

The Problem

JPMorgan Chase's existing business banking product had been built for large enterprises. That's who the bank knew how to serve, and the product reflected it — bundled features, complex onboarding, processes designed for companies with dedicated finance teams and months to spare.

Startup founders are the opposite. They're moving fast, wearing every hat, and have zero tolerance for friction they didn't sign up for. Opening a business account shouldn't take two weeks and forty questions. But that's what the existing process required.

The competitive pressure was real. Mercury had built its entire identity around being the bank that actually understood founders. SVB, before its collapse, had that relationship too. JPMorgan was late to this segment and knew it. The question wasn't whether to fix the experience — it was whether the organization was willing to change the underlying processes that made the experience bad in the first place.

That second question was harder than the first.

My Role

This wasn't a blank-slate startup project. It was an effort to shift how one of the world's largest banks thought about a group it had long overlooked. The design work was just one piece of a much bigger puzzle.

I worked side by side with the Chief Innovation Officer, a product lead, and the middle office compliance teams, with analysts helping us dig into research and synthesis. We also brought in an external KYC technology partner. To get our onboarding flow from concept to reality, we had to persuade compliance teams to rethink processes they'd relied on for years — not just accept design recommendations, but actually change how they worked.

I was the only designer on the product for most of its development.

What We Did

The core vs. chore insight became the organizing principle for every decision we made. If banking was fundamentally a chore for founders, our job was to make it as invisible as possible. That meant attacking friction at every layer — not just in the interface, but deep in the processes that created it.

Redesigning compliance, not just the form

The 40-question KYC process wasn't a regulatory requirement. It was accumulated process debt — questions added over the years, never revisited, never challenged. Working directly with middle office compliance teams, we mapped every question to its corresponding regulatory requirement and removed everything that wasn't mandatory.

Forty questions became fourteen. But the bigger shift was what happened next: we automated every question that remained. By partnering with a KYC technology provider, we swapped out manual document review for real-time API validation. Founders no longer had to upload documents and wait for someone to check them. The system did it instantly.

E-consent was the last piece. The manual email-based consent process got replaced with digital agreements embedded directly in the flow, removing an entire back-and-forth cycle that had added days to the timeline.

The interaction design decision

Even with just fourteen questions, how you ask them makes all the difference. I explored two paths:

A long single-page form — everything visible at once, fastest path for someone who knows what they're doing, but risks feeling overwhelming upfront.

A guided wizard — one step at a time, conversational in tone, lower cognitive load.

I chose the wizard — not because it was simpler for the average founder, but because of the edge cases. Founders with complex ownership structures or international backgrounds needed a different path, and a long form only surfaced those issues at the very end, after they'd already invested time. The wizard let us spot those branches early and guide founders through them smoothly.

Designing for the edge case made the entire experience sturdier for everyone.

Getting the organization aligned

Fast onboarding took more than thoughtful design. It needed executive commitment to change compliance processes that had been in place for years. To help make the case, I built a vision storyboard that mapped out the founder's journey: discovering the product, navigating onboarding, and finally landing on a working account, ready to go.

The storyboard wasn't built for users — it was for leadership. By showing the emotional arc of the founder experience, not just the UI, we gave the Chief Innovation Officer something tangible to rally the team around. Startup banking shifted from an innovation experiment to a strategic priority.

Results

Our team's role was to show what was possible. The forty-question, fifteen-day onboarding process wasn't inevitable — it was a choice, and we proved it could be different. That demonstration of the art-of-the-possible became the new target for the product teams.

Getting there took more than a year of implementation after our handoff. But the direction was set, and executive commitment was in place. Startup banking became a major strategic priority at JPMorgan Chase, with the bank making a substantial investment to build out the platform we had prototyped.

The research, the core vs. chore framework, and the design principles we established in late 2021 and early 2022 became the foundation for the digital banking platform that JPMorgan Chase will launch to the world in summer 2026.

The SVB collapse in 2023 made the stakes real. Suddenly, thousands of founders needed a new banking home. The urgency we'd been designing for became tangible, and the groundwork our team had laid gave JPMorgan Chase a real path to serve them.

This project led directly to my promotion to Executive Design Director.

What I Learned

Compliance is a design problem. The forty-question KYC form wasn't required by regulation — it was process debt that had piled up over time. No one had challenged it because no one had mapped it against what the rules actually called for. That's where design comes in: spotting the gap between what exists and what's truly needed, and making that gap visible to the people who can change it.

The measure of success wasn't the fidelity of the prototype — it was whether the organization changed direction because of it. The storyboard, the research synthesis, the wizard flow — those weren't just deliverables. They were tools for getting a large organization to commit to a direction it otherwise wouldn't have taken. That's a different definition of success than shipping features, and it requires a different way of working.

Design for the edge case first. The decision to use a wizard came from asking what happens to the founder who doesn't fit the standard profile. By designing for that person first, we made the whole experience sturdier. This lesson goes beyond onboarding — the edge case almost always reveals the hidden cracks in any system.