A retention loop for a food-sharing app

Phase 1 · School team

Phase 1 enabled food sharing but did not build lasting food habits.

The Phase 1 survey included 18 UBC students. Coursework often displaced cooking, groceries went unused, and community fridges were inconvenient to reach.

With 3 other designers, I helped shape a marketplace for listing extra food and arranging nearby pickups.

Once the food was gone, the experience ended.

Phase 2 · Independent

Reducing food waste meant helping students before food became urgent.

I returned in November 2024 and interviewed 6 students who lived on campus and cooked for themselves.

Key finding: Students thought of food sharing only when groceries were already going bad. Between those moments, the app disappeared from their routine.

I began asking what would make students return to LastBite and build better food habits between exchanges.

Prototype testing

I tested whether the loop stayed useful between pickups.

I tested the original and redesigned prototypes with 6 students who cooked for themselves. I varied which version they tried first and gave them two scenarios.

Result: 5 of 6 students found a useful next action in Phase 2 without prompting and explained how they would use LastBite between pickups.

Limit: These sessions captured perceived value, not repeat use over time.

Decision 1

Turn a lesson backlog into a library students could explore.

The initial screen showed every lesson in one long list. Testers read it as a backlog they were expected to complete in order.

I replaced the list with swipeable cards that show one topic at a time, so students could explore without seeing the entire library as unfinished work.

5 of 6 testers preferred the cards.

Decision 2

Put the pickup action where people could find it.

In think-aloud testing, 3 of 6 participants could not find the pickup request. They also wanted more information about the person listing the food.

I made the pickup CTA sticky and moved profile details higher in the hierarchy. In the later test, 5 of 6 participants completed the sharing flow faster.

Profile details can help someone assess a listing. A real marketplace would still need food-safety rules, trust systems, and operational support.

Decision 3

Tie the reward to the food.

I made the virtual fridge the destination for proposed coin rewards. The coins turn into decorations, giving sharing and lessons a visible payoff.

The metaphor keeps the reward tied to food. Because the fridge is optional, someone can still share groceries without entering the reward loop.

Design System

One system across sharing, lessons, and rewards.

I built reusable components and custom illustrations across listing, lessons, and rewards. The system keeps core actions clear while giving the reward screens more personality.

Reflection

Define what a meaningful return looks like.

LastBite taught me to question what I count as success. More app visits would not necessarily mean less food waste. If students stored food better and had less to give away, fewer pickups could be a good outcome. I would define the food habits I wanted to support before choosing how to measure engagement.

I also learned to be more selective about what I test together. Lessons, coins, and the fridge formed one concept, so interest in the whole experience could hide which parts were useful. Next time, I would test the lessons on their own before deciding what the rewards add.