A better way to save food, a kinder way to share it

Outcome

Research that changed the product, not just the deck

I ran a survey and interviews to test whether peer-to-peer food sharing was actually wanted, or just a nice idea. The results were decisive enough to reshape the core of the app: they killed my original "share your home-cooked meals" concept, and they turned trust from a feature into the backbone of the product.

The Problem

Two problems sitting next to each other

Some people cannot reliably afford to eat: international students who are not allowed to work, people living alone without a safety net. At the same time, their neighbors are throwing away food that is still good. The two sides are often on the same street and never meet. Existing options do not close that gap: second-hand marketplaces are not active for food, and food banks ask for proof and paperwork that many people do not have.

Where existing apps stop

I studied three products already in this space. Olio handles neighbor-to-neighbor surplus but leaves listing slow and pickup socially exposed. MealConnect solves food rescue at business scale, not for individuals. Share the Meal makes donating money effortless but has no local, physical exchange. Each does one part well, and none covers a hungry student at 9pm two blocks away.

User Research

What people actually told me

I ran a survey plus open-ended questions covering food access, past reliance on help, willingness to give and receive, motivations, and blockers. I analyzed the responses across three dashboards.

Food insecurity is closer than it looks

People are more open than I assumed

Trust is the real barrier

What I did with it

Open answers were blunt and useful: people wanted verified donors and review-based credibility, anonymous profiles and contact-free pickup, in-app chat and public meeting spots, and visible proof of impact. Several also said they would share ingredients even when they could not share cooked food.

Design Foundation

Designing for Charlie

Charlie is 24, an international graduate student living alone in a rented room, working part-time and stretching ingredients across days. He is independent and introverted, and he will not ask for help face to face. His line became the design brief: "I don't want to ask for help directly, but if there's a quiet way to give and receive, I'd be all in."

Journey map

Mapping Charlie's arc from hunger to ongoing use exposed exactly where the design has to carry him: hesitation at sign-up about handing over personal details, and anxiety right before the pickup. Those two moments are where contact-free options and the trust score do the most work.

Design Process

From flow to screens

I mapped the full flow first, then worked from sketches through a three-column grid into low and high resolution wireframes, so the structure was settled before any visual polish.

The Product

Four screens that carry the idea

Home

One place for shared food nearby, the global donation feature, and community events, so giving money and giving food live side by side.

Explore

A map of what is available near you, with distance and pickup window on every listing, and dietary filters for Halal, vegan, and allergen needs so nobody has to ask an awkward question.

Post

Sharing is as simple as posting a photo, adding a category and an available-until date, and choosing a handoff. The pickup option lets a giver skip scheduling and the awkward handoff entirely.

Profile and the KindScore

KindScore is the trust system the research demanded. Everyone starts at 600 and needs 700 to begin sharing food. Showing up and sharing safely raises it; no-shows and expired items lower it; below 300 carries a penalty. It gives strangers a reason to trust each other without forcing anyone to reveal who they are.

Evaluation

Heuristic evaluation

I ran the finished flows against a heuristic checklist to catch the usual failures before they reached a user: system status, error prevention, recognition over recall, and consistency across the tabs.

Reflection

The research earned its keep

The most valuable thing this project did was prove me wrong early. I started out designing a home-cooked meal-sharing app and assumed the emotional barrier would be pride. Research said the opposite: pride was not the issue, safety and trust were, and cooked food was a non-starter. Rebuilding around ingredients, anonymity, and a visible trust score made the concept viable. If I took it further, the next step is testing whether the KindScore thresholds actually feel fair to new users, since a 700 gate could quietly lock out the people who need it most.