← All work

Case Study 01 of 03

Redesigning Zelle troubleshooting for Wells Fargo

Audience
B2C
Role
Senior Content Designer
Team
1 Product Manager, 1 Product Designer, 5 Developers
Timeline
2025 to 2026
Tools
Figma, FigJam

Context

Zelle issues were a top driver of Wells Fargo support calls: 650,000 calls per year, with 85% of customers engaging with Fargo, the bank's virtual assistant, first. But every Zelle related question routed to one generic response with static tips, regardless of the actual problem. Users defaulted to calling an agent.

Challenge

Users with completely different problems, a failed send, a missing payment, an enrollment issue, all got the same non-specific answer.

Solution

Led the end-to-end redesign of the in-app Zelle troubleshooting experience for Fargo, turning one static list into a branching, diagnostic conversation.


Research

Analyzed call driver data with Support to find the most common Zelle failure points, and how customers actually described each one.

Spreadsheet of Zelle-related call driver utterances grouped into categories like New Intents, Zelle Down, Zelle Issue, Why Can't Use, Zelle Eligible, and Zelle Tokens, each with frequency bar charts.
Information-first design

Grouped root causes into categories, then used those to define branching error types, from account restrictions and insufficient funds to failed and missing payments.

Wide flowchart mapping eight branching Zelle error types, unregistered account, restricted account, insufficient funds, exceeded limit, incorrect recipient info, money not sent, money not received, and cancel transaction, each with its own decision logic and in-app response screens.
Usability testing

Tested flow options with Product, Design, and Engineering. The structure that performed best surfaced specific problem categories up front, then closed with a checked-issues summary.

Two side-by-side variants of the Zelle troubleshooting conversation flow being tested, each showing the search entry, category selection, and final troubleshooting summary screen.
Solution and impact

A new troubleshooting flow that checks for common issues automatically. Each error type provides targeted, relevant guidance instead of a generic list, and closes with a transparent "here's everything I checked" summary, so users can see exactly what the assistant already ruled out.

After

After: the assistant asks what specific problem the user is having, offers targeted options, and ends with a transparent summary of everything it already checked. Zelle is up and running, service is activated, enrolled, payment profile set up.

Before

Before: the same entry point as After (welcome screen, typing the question), but every question routes to one static troubleshooting tips response, regardless of the actual problem.

Outcome

Reduced Support calls by 29% and saved the bank $205,000 annually.