Ecopart — Trust-First Coliving Platform
A trust-first coliving platform built from scratch, with a 6-mechanism trust architecture and intentional friction to lift match quality.
Ecopart is a trust-first coliving platform built from scratch. People were being asked to share a home with strangers, and the existing tools treated that decision like booking a hotel room. I designed the end-to-end product: the trust architecture, the profile and matching logic, the request flow, and the visual system that holds it together.
Coliving is not a transaction. It is a decision about who you live with, and it carries real risk on both sides. Yet almost every platform in the space optimises for speed: swipe, book, move in. Speed is exactly what makes people anxious, and anxious people abandon the flow at the moment of commitment.
So the core question was not "how do we make matching faster?" but "how do we make two strangers feel safe enough to say yes?" Everything I designed answers that question, including the parts that deliberately slow the product down.
“I left the day I deleted my account — hunting for rooms in social groups was never safe anyway. What kept me on the platform I switched to was simple: someone guides the process and guarantees a fair deal.”
The questions people actually ask at the moment of commitment:
- “How can I be sure this isn't a scam?”
- “Why do they need my data before I've signed anything?”
- “What if my roommates turn out to be untrustworthy?”
- “What if I need to cancel?”
Coliving is two-sided — renters looking for a home, owners deciding whether to trust strangers with one. I interviewed 11 people (8 renters, 3 owners) about one thing: what makes them hesitate.
Every barrier traced back to one root: a lack of trust.
- SAFETY CONCERNS
- LACK OF TRANSPARENCY
- UNCERTAINTY ABOUT LIVING CONDITIONS
- ABSENCE OF PLATFORM ACCOUNTABILITY
Emotional risk beats financial upside — people pay more for a studio over the wrong flatmates.
“Why do they need my data before I've signed?”
Safety, shown before you commit.“What if my roommates can't be trusted?”
A viewing you can't skip.“How does this even work — did I miss a step?”
One path, one step at a time.“When's my payment due, and how much?”
Everything in one place.“Do I have to message a manager for everything?”
Answers before a human.“Verification's the scary part — that's where I'd quit.”
A guide through the hard parts.Safety, shown before you commit
Identity verification and clear rules — before you commit, not to browse. You know who you're living with; they know you.
A viewing you can't skip
Every rental starts with a viewing. I made it mandatory on purpose: an informed choice beats a hopeful one.
One path, one step at a time
Application, contract, move-in — in order. You can only apply once the viewing is marked “Completed”. The order is the reassurance.
Everything in one place
Address, dates, payments, documents, support — one dashboard, for renters and owners. Nothing buried in email.
Answers before a human
FAQs, reports, emergencies up front. The manager is the last resort, not the first tap.
A guide through the hard parts
A mascot for the stressful moments — honest when things break, present for the boring bits. Reassurance, not decoration.
Soft geometry, everywhere. Rounded cards, pill buttons, gentle corners — nothing sharp. The shapes themselves feel safe, so committing to a decision feels less like a risk.
Pastel accents that calm, not sell. One muted violet, one soft green, and nothing louder. Colour is used sparingly and with meaning — to guide, and to mark what's safe or confirmed — never to push.
A mascot for the hard moments. A friendly cat appears exactly where anxiety peaks — verification, errors, the wait, the confirmation. It's the one warm, human touch that turns tense steps into reassurance.
The lesson I keep from Ecopart is that friction is a design material. I spent the first weeks trying to remove steps out of habit, and the product only started working when I began asking which steps deserved to exist. Slowness, placed correctly, reads as care.
What I would do differently: test the six trust mechanisms separately before shipping them together. I validated the system as a whole, which told me it worked but not which parts were carrying the weight. This project sits in my portfolio because it is the clearest example of designing a system rather than a set of screens.