A neighborhood social app that helps people borrow a ladder, find a dog-sitter, or just learn the names of the folks two doors down.
✦ Client
Capstone · Google UX Design Certificate
✦ Role
Lead UX Designer
✦ Duration
10 weeks
✦ Platform
Mobile-first iOS / Android
✦ The shape of it
StayNTouch is a hyperlocal social network that flips the script on traditional social media: instead of showing you the world, it shows you your block. The MVP focuses on small, low-stakes asks between neighbors and a friendly directory of who lives where.
Problem
Hyperlocal apps fail one of two ways — they get hijacked by complaint threads, or they get so empty they feel like ghost towns. We wanted to figure out what shape a positive, useful community app would take if we designed for the boring 80% of neighborhood life: borrowing things, asking quick questions, organizing a Saturday cleanup.
What I did
A trust-first onboarding flow that verifies addresses before content, three primary feed lanes (Asks, Offers, Happenings), and a deliberate absence of public comment threads on profiles. The whole product is shaped around 'the thing my neighbor would actually do.'
✦ Outcomes
92%
task success rate on the borrow-an-item flow
4.6/5
average usability rating across 12 testers
−40%
fewer steps than the closest competitor
2
design awards from the certificate cohort
01 — The brief
Our team was given a vague prompt: 'design something that brings communities closer.' We narrowed in on hyperlocal — the apartment building, the cul-de-sac, the literal next door — because the existing options felt either toxic (rant-y forums) or transactional (paid handyman marketplaces).
02 — Research
We ran 14 interviews across three cities with people in apartments, suburbs, and one small town. We expected loneliness to be the big theme. It wasn't. The actual theme was friction: people wanted to ask their neighbor for jumper cables but had no socially acceptable way to do it without knocking on a stranger's door.
03 — Big ideas
Two principles shaped the rest of the project. First: low-stakes by default. The most-used flows had to be one-tap and feel like a Post-It on the fridge, not a Facebook status. Second: trust shows, content follows. We sequenced onboarding so verification happened before posting, not after — a small step that quietly removed a class of bad actors.
04 — Design
We tested three feed structures (chronological, lane-based, AI-curated) and lane-based won decisively. The 'Asks / Offers / Happenings' split gave people a mental model that mapped onto real neighborhood behavior, and let us cap each lane to ~5 visible items so the app never felt like a doomscroll.
Visually we leaned into a friendly, residential palette — mustard, sage, terracotta — and rounded everything aggressively. The brand is closer to a community fridge than a tech product, and that was on purpose.
05 — Outcomes
The borrow-an-item flow tested at 92% task success in moderated sessions, with an average completion time under 35 seconds. Two of the three feed iterations were retired entirely after testing — a decision I'm proud of, because earlier in the project I would have tried to keep them in 'just in case.'
06 — Reflection
If I shipped this for real, I'd invest more in moderation primitives before turning on any kind of public posting. Trust is the product, and the design has to make breaking it feel awkward. The MVP gets the shape right, but a real launch needs a lot more on the safety side.
✦ Keep going