Kettlefund · shared household money
Case file / 0
1
Making shared budgets less awkward
Saga
Role
Lead product designer at Latchfield
Team
2 designers, 4 engineers, 1 PM
Tools
Figma, Maze, Dovetail
Discipline
Product design
Pinned summary
The 30-second version
Couples checked their shared budget less often because each visit felt like an audit. I reframed the app around planning together, tested a calmer monthly view and shipped it in stages. Weekly active households rose by 18%.
1
+18%
weekly active households
−26%
“Where did this go?” messages
4.6/5
planning-flow confidence
Illustrative demo metrics.
Problem frame
Challenge
Weekly check-ins dropped because the app felt like an audit. Transaction lists led the home screen, warnings appeared before context, and one partner often became the unofficial administrator.
Across twelve interviews, people said they opened the app when something felt wrong, not when they wanted to plan. The challenge was to increase useful weekly check-ins without hiding overspend or turning money into a game.
Evidence → direction
Approach
Research pointed us towards shared planning before visual polish. I mapped the moments that prompted a money conversation, paired interview themes with usage data, and invited six households to annotate their monthly view.
Three low-fidelity directions tested a transaction feed, a flexible plan and a weekly prompt. The plan performed best because both partners could set context before seeing where spending differed from the plan.
Decision log
Decisions & trade-offs
We chose calmer context over density, blame or daily rewards.
01
Lead with the shared plan
We moved transactions one level deeper. People gained context, but power users lost immediate density.
02
Describe change, not blame
We wrote “£42 above the plan” instead of “You overspent.” Friction fell, but partners still needed to talk about why.
03
Keep one monthly rhythm
We resisted streaks. Visits were less frequent, but each had a clearer purpose.
Gallery / Final flow
The solution in three moments
Each screen supports a conversation rather than scoring a person.
After release
Outcome
The calmer monthly view improved participation and reduced support demand. A staged release began with a 20% cohort. Over six weeks, weekly active households increased by an illustrative 18%, while related support messages fell by 26%.
Participants described the app as a place to “get on the same page”, rather than a record used to prove a point.
Next iteration
Learnings
Sequence and tone can make the same financial data safer to discuss. I would next study households with irregular income and test whether a rolling four-week view creates better context.
AR
“Saga kept the product honest without letting the numbers set the tone of the relationship.”
Sample
Ansel Reed, product manager at Kettlefund
Next project ·
Ferrylane
Finding the right crossing, faster
Remix this template ↗