Stand up a modern investment product on legacy tech stack. Plug in multiple vendors.
Led standup for product and content strategy to launch across mobile app/web, tablet, and desktop.
$500B+ in account funding, aggregate increase in digital account enrollment.
• AI/ML analysis of millions of user bank accounts for trailing 6-month period to determine target segments
• Run Monte Carlo simulation to predict money growth based on weak, median, and strong markets
• 'Nudge' users from banking accounts i.e. "See how much your money can grow over time"
• Lead users to 'madlibs' experience and $ growth chart:
• Aggregate hundreds of research sessions with JPMC financial advisors
• Simplify complex financial products into conversation form, maintain JPMorgan Chase voice standards
• Write questions and answers that consider OKRs
➡️ Generate account opens, focus on retirement type:
• Retirement account type journeys were being abandoned at higher rate than others
• Determine reason(s) for session abandons (red text):
• Guardrails solution: remove user ability to choose options that trigger error message
• Greyed-out helper text informs rationale
📈Result
26% increase in retirement account opens
After questionnaire completion, users see projections:
JPMorgan compare Mutual Funds and ETFs
• Users can exclude funds from portfolio
Managed cross-product journey, helper content, account open origination
Top Self-Directed Investor Satisfaction experience awarded by J.D. Power
Work with scrum team to strategize and design end-to-end user flow of adding external accounts via Plaid (and afterwards) Yodlee/Envestnet.
PLAID - Auth
API reference for Auth endpoints and webhooks:
Retrieve bank data to set up EFTs, e.g. ACH payments in the US, EFT payments in Canada, BACS payments in the UK, and IBAN / SIC payments in the EU.
PLAID - Balance
Balance API reference for Balance endpoints and webhooks.
Ensure language describes third party participation, what user is permitting to share with Chase.
After advisors inquired about 3rd party account access, I created the 'advisor toggle.' This went through extensive LRC!
Program language and back-end model for virtual assistant (VA) content design
Plugin learnings from natural language search (NLS)
Create conversation flows and write language, include deep links for efficient user navigation, connection to live agent
.svg)
