In a world of rising costs and shifting social norms, traditional financial tools fall short of supporting modern relationships. Money is emotional, especially when it involves others. Our initial insight was simple but powerful: We started studying the landscape. We read through countless reviews on reddit about existing solutions like Venmo and Cash App, and Zelle—tools that offer joint mobility but lack the visibility couples need. We also learned these tools create massive bottle necks around payment frequency.
60% of divorces cite money as a core reason.
75% of Gen Z and Millennials prefer to keep finances separate.
But the tools available — like Venmo, Zelle, or even joint bank accounts — created bottlenecks, misunderstandings, and friction. We saw a gap: no product truly supported the collaborative financial journeys of modern relationships.
We started with digital ethnography — combing Reddit, App Store reviews, Twitter, and online forums to understand how people use and complain about money tools. Patterns emerged around:
Frustration with mental math: “Didn’t I pay the last bill?”
Lack of visibility: “I can’t see what we spend together.”
Tense negotiations: “We split things 50/50 but our incomes are different.”
Then we hit the streets. Literally.
We conducted hundreds of interviews — with friends, family, and even strangers at coffee shops and bars. We interrupted dates. We asked roommates. We poked into co-parenting arrangements.
We learned that:
People don’t always label their setup as “joint finances.” But collaboration is happening.
Tools that felt too “official” (e.g. opening a joint account) created discomfort or resistance.
People craved flexibility, fairness, and a sense of control over their money.
Each persona gave us insight into why they manage money with others and what functionality they actually need.
Onboarding + KYC
Frictionless but compliant. Two people, one pod, separate accounts.
Account Linking
Plaid integration for visibility and ACH for movement.
Visibility Tools
Joint views, personal views, custom control over what’s seen.
Mobility
Splitting bills. Funding a shared pod. Setting financial responsibilities.
This tree helped us prioritize ruthlessly: we picked features that hit emotional resonance with users while delivering quantifiable value.
Couples needed to see where their money goes — without feeling like they were losing control.
Dual ledger showing who paid for what
Filters for shared vs. personal expenses
Customizable views for privacy
“I finally understand our bills without feeling like I’m being watched.”
Splitting finances had to feel fluid and fair.
Split pods: Create shared spaces for bills, rent, or recurring categories
Ratio-based splits: 50/50 or income-adjusted
Approval flows: Either party can propose a change
ACH-based funding: Real joint power, separate accounts
We introduced PenelopeAI — an embedded conversational agent that helps users:
Understand their spending habits
Answer financial questions
Make suggestions for fairness (e.g., “Your partner paid 70% this month”)
Translate behavioral data into relationship-sensitive insights
From the beginning, we baked in tight feedback loops:
In-app feedback prompts after key actions
Weekly interviews with power users and churned users
UserTesting.com scripts for onboarding and AI interactions
Surveys on emotional trust and friction
Switched reminder language from “Pay Now” to “Just a Heads Up” after testing tone sensitivity
Removed auto-splits on first use — too jarring; added preview before execution
Simplified AI chat with less finance jargon after testing comprehension rates
14,000 users on waitlist onboarded in 9 weeks
65% week-over-week retention
10,000+ of 1:1 convos used to shape features
Users report feeling less stressed, more “in sync” financially
$1M in venture capital raised to support the project
Accepted into Google for Startups
Accepted into NVIDIA’s AI Program
Peas didn’t just need to be a good product it had to be trustworthy, invisible, and emotionally calibrated to manage joint money journeys. That set the bar higher than most consumer fintech products. Here’s how I anchored the product strategy blow past that bar.
A fundamental decision we made early was to decouple our product from the outdated metaphor of joint banking. Most existing solutions, from Chase joint accounts to Venmo, assume shared finances; which means shared accounts. But 75% of our ICP actively avoids those tools due to complexity and bottle necks. What they wanted wasn’t shared money, it was shared understanding of that money.
So we defined our architecture around “collaborative autonomy”:
Shared visibility, but separate accounts
Mutual decisions, but no forced merges
Joint control, but granular privacy
This was a strategic design wedge. It let us serve Gen Z and Millennial users who live together, raise kids together, and split lives together without forcing them to change how they feel about money. Even applying to secondary and tertiary personas like Roommates and Families.
The product’s edge wasn’t that we built better features. It was that we built for a different emotional premise.
From the beginning, we designed Peas to grow through usage. Marketing is great, but real retention comes from viral loops and users needed to share our product with others.
Value Creation: Provide clarity, fairness, and emotional ease from day one (i.e., “This helped us finally talk about money without arguing.”)
Value Delivery: Fast onboarding, seamless linking (Plaid, Finicity, MX), and just-in-time nudges from PenelopeAI
Value Capture: Introduce meaningful subscription layers only after we deliver a real “money relationship OS”
By focusing first on how often a couple interacts with the product, we drove strong retention and had low churn during our first launch. For a consumer subscription app, hitting 65% week-over-week retention was a rare sign that we were in the right emotional lane.
Instead of tracking just DAUs or MAUs, we tracked shared moments per week per pod as well as frequency of conversation. This was our north star for relational stickiness and helped us prove that using natural language to manage money was the proper direction.
Traditional fintech often forgets this.. but language is part of money. How we communicate about it, how we learn about it, even how we plan with it.
We wanted to test every word for emotional friction. From call-to-actions (“Pay Now” → “Just a Heads Up”) to AI phrasing (“You spent more” → “Your contribution looks higher than usual — want to rebalance?”).
This was to create a system of trust. Money is inherently stressful and confusing. Every surface of the product, text, timing, and tone, became a lever that shapes whether couples could trust us and more importantly, just our AI.
This also impacted how we scoped the chatbot. PenelopeAI’s suggestions were designed to offer insights without accusation and at the right times in the money journey.
We treated emotional trust as a tier-one KPI. That’s what made Peas feel like a friend, not a finance tool. Something the best experiences are doing well in the new age of AI powered tools.
Most products try to do too much too fast. We deliberately sequenced Peas as a trust-ladder:
Step 1: Visibility – show joint data without judgment
Step 2: Mobility – enable equitable, flexible movement (splits, pods, joint approvals)
Step 3: Intelligence – introduce PenelopeAI as a liaison to the couple
We thought AI should enter a relationship like a good therapist; by listening first, providing value second, and offering solutions later.
Peas walks a delicate line between financial tooling and emotional safety. We intentionally rejected overbuilt features that sounded “smart” but felt too intrusive. Some of the tradeoffs:
No auto-split on first use (felt aggressive)
Preview + confirmation flows (earned trust first)
No pushy subscription walls (we actually removed the paywall after initially seeing friction on activation)
Delayed monetization in favor of building usage-based value
We asked ourselves often: Would this make a user feel more in sync or more managed?
That became our internal litmus test for feature decision making.
