Splitwise Teardown

problem

The Problem

Splitwise loses users after they settle up

Splitwise dominates expense splitting, but the product lifecycle effectively ends when debts are settled. There's no reason to come back. D30 retention drops by 60% once a group becomes inactive.

• 60% D30 retention drop post-settlement
• Product lifecycle ends at transaction completion
• No group continuity features exist
• Users see Splitwise as a 'utility' not a 'companion'

research

The Research

Mapping the full user journey

I mapped the complete user journey from group creation to settlement and beyond. The critical gap: Splitwise captures the 'splitting' job perfectly but completely abandons users after the core job is done. The group context — the most valuable asset — is discarded.

• Mapped group creation → expense → settlement → churn funnel
• Identified that group context (who owes whom, shared history) is the real asset
• Benchmarked against Venmo, PayPal, and shared finance tools
• Found that 'household' groups have 3x higher lifetime value than one-off groups

decisions

Key Decisions

Reframing the North Star

The existing North Star (settlement frequency) was measuring the wrong thing — it optimized for closing transactions, not retaining relationships. I proposed Household Continuity Rate: the percentage of groups that remain active beyond settlement cycles.

• Rejected settlement frequency as the North Star — it measures endings, not engagement
• Proposed Household Continuity Rate — measures ongoing group value
• Recommended shared budgets and recurring expense tracking as retention features
• Designed a 'group memory' concept — expenses become shared financial history

solution

The Proposal

From transactional to relational

Shift Splitwise from a settlement tool to a household-first shared-finance companion. Key features: shared budgets, recurring expense tracking, group financial goals, and a 'group memory' that keeps context alive between settlements.

• Shared budgets for ongoing group expenses (rent, groceries)
• Recurring expense tracking with automatic splitting
• Group financial goals ('save for trip', 'monthly budget')
• 'Group memory' — expenses become shared financial history, not just debts

impact

Projected Impact

Retention-first metrics

The proposal targets the 60% D30 retention drop directly by giving users a reason to return beyond settlement. Household Continuity Rate replaces settlement frequency as the metric that drives product decisions.

reflections

Reflections

The biggest insight wasn't a feature idea — it was realizing Splitwise was measuring the wrong thing. Sometimes the most impactful PM work is redefining what success means, not building more features.

• Sometimes the best product insight is 'you're measuring the wrong thing'
• Retention is a context problem, not a feature problem
• Group context is the most underutilized asset in P2P finance
• This teardown changed how I think about North Star metrics forever