./loop

resale, minus the listing pain.

role:

founding designer · 0→1

team:

1 pd · 3 eng · 2 founders

timeline:

2018 — launch 2019

platform:

ios + web · brand + product

[ loop — photo-first listing flow, 4 screens ]

$ cat ./context

the problem

64% of sellers quit before finishing their first listing.

every resale app treated listing like a tax form: 14 fields, category trees, condition matrices. the average first listing took 11 minutes — when it got finished at all.

loop’s bet: the supply side wins the marketplace. if listing is effortless, inventory follows, and buyers follow inventory.

the brief >

“make listing feel like texting a friend, not filing taxes.”

3 min

median first listing (was 11 min)

64→22%

listing abandonment after photo-first flow

+28%

sell-through rate with smart pricing

3.1×

gmv growth year one post-launch

$ cat ./process

research → flows → system

(a) research

we watched 14 people list a jacket. the photo was the wall.

in-home sessions with 14 sellers, phones in hand. nobody struggled with price or description — they stalled at taking an acceptable photo, retook it 5–6 times, then gave up. the form wasn’t the problem. the blank camera was.

[ in-home session stills — photo attempts ]

[ photo-first flow — camera opens first ]

(b) flows

camera first. everything else is a suggestion.

the flow starts in the camera with a framing guide. category, title and price arrive pre-filled from the photo — sellers confirm instead of compose. price suggestions come from comparable sold items, shown as a range, never a demand.

a design system from zero, sized for a team of three.

as founding designer i also owned brand, tokens and the component library — deliberately small: 9 components, 2 type styles, 1 accent. boring on purpose, so shipping stayed fast.

[ loop design system — components sheet ]

$ open ./solution

what shipped

[ photo-first listing — final ui, camera to live in 3 steps ]

[ smart pricing — suggested range ]

price shown as a range from comparable solds — sellers stay in control, listings stay realistic.

[ offer inbox — accept / counter / decline ]

offers handled like messages: accept, counter or decline in one tap. no negotiation dread.

$ tail ./outcome

after launch

loop launched in 2019 and tripled gmv in its first year. the photo-first pattern got copied by two bigger competitors — the sincerest form of product validation.

“listing used to be sunday homework. now i do it in the queue at the supermarket.”

— seller #9, in-home session follow-up

// back to the start — 01/04

./payflow →

back to index ↑