OAKE
A secondhand garment loses its history at the donation bin. OAKE was an NFC tag that let the person giving an item away record where it had been, so the next person could read it by tapping their phone. Three stores had agreed to pilot it when retail closed in March 2020.
The problem
A secondhand garment is worth more when it has a past. That is most of what separates a vintage sweater from a cheap one, and every thrifter we interviewed said some version of it unprompted.
The donation bin destroys it. An item arrives in a bag with thirty others, gets sorted, priced, and hung, and reaches the floor carrying nothing but a size. Shoppers compensate by inventing a history or by skipping to the story of the find. Our guiding question was how to improve product transparency in a secondhand store, which stayed broad enough to be unusable until the interviews narrowed it. The missing information was not condition or origin. It was who had owned the thing, which turned the question from how to document an item into how to make handling one mean something.
Research
Ten interviews, from occasional shoppers to people who thrift most weekends.
The first finding was that thrifters were already storytellers. They described the hunt and the anecdote at the end of it rather than the purchase.
We were not adding a story to a garment. We were replacing an invented one with the true one, for people who would tell a story either way.
The second was a constraint. People will consume but will not go out of their way to do it. No app download, no account, no URL typed off a hangtag. Anything we built had to work in the seconds a shopper already had the item in hand.
Four of the ten interviews. Amina’s objection to sustainability messaging and Stephen’s attachment to an item’s history pulled the project in opposite directions.
Story over sustainability
The environmental version of this product was on the affinity map and we did not build it: a sustainability score, an estimated carbon footprint, an ethically sourced badge, geographic history. It was the version we arrived expecting to build.
One participant closed it. Zero waste, she said, gets pushed onto consumers when it belongs to the companies producing the waste. She already knew secondhand was the better choice, which was why she was standing in the store. Being congratulated for it at the point of purchase was an irritation rather than a reward.
Provenance had the opposite posture. Offered rather than asked, and wanted for the shopper’s own reasons rather than ours.
The medium column held roughly fifteen options. The environmental cluster on the left is the product we did not build.
The tag
Fifteen or so delivery mechanisms went on the board: QR codes, magnetic strips, a text-in number, a mini pamphlet, wand readers, a website, an AR app. We took NFC.
The argument was about what buying should feel like. Online purchase had become effortless to the point of detachment, and in-store shopping had picked up the same flatness. A tap keeps a hand on the garment. A QR code sends the shopper backwards into a viewfinder, framing a square instead of looking at the thing they are considering. If the premise was that buying a physical object should feel physical, contact was the consistent choice.
NFC was also the newer technology in 2020, and its novelty was part of the appeal. That was a real input at the time, not a rationale I would construct now.
Tags were disposable and single-use, which suited a store attaching them by hand to donated inventory at intake. The donor writes at drop-off and the shopper taps weeks later. The two never meet, and the garment carries the message between them.
The product
The name field
Intake opens by asking the donor to name the garment, ahead of category and condition. The screen states that the name is what future owners will know it as.
Naming an object is the cheapest available form of investment. A donor who has decided a sweater is called Ayershire has already agreed it is a specific thing with a history worth recording, which was worth more to us than the field itself. The plural was deliberate. The item was meant to outlast this donation.
Three passes at the labels
The first version was written the way we wanted the product to sound. Sections were called It carries a connection, From a person, and Who wants you to know. A fourth, Its family carries a past, sat above a classical painting of women working a bolt of silk.
Testing removed it. A participant asked what the pictures had to do with the sweater, and there was no answer. The image was atmosphere standing in for provenance. Another said she was intrigued but could not tell what the screen meant. A third asked for something as ordinary as a learn more link.
The second pass renamed the sections A Memory, Meet the Owner, and You Should Know. The third made them plainer still: Backstory, Previous Owner, What to Know. Same content, three passes, each less written than the one before it.
The shipped cards, swiped through.
The intake form
The first form was a single scroll: name, memory, notes for the future owner, and the donor’s own taste, all four fields visible at once. Four empty boxes are a statement about duration, handed to someone standing at a counter with a bag at their feet.
The shipped flow asked one question per screen, name first, so the donor faced a question rather than a remaining quantity.
Intake, from naming the item through the memory prompt.
The supply problem
Donors completed the flow, but only for items they cared about. Across two rounds, more than twenty participants between moderated sessions and in-store volunteers, the pattern held: someone donating eight things wrote for one or two.
Workable at three stores, not at three hundred. Most items on a rack would carry no story, and a tag that returns nothing twice teaches a shopper to stop tapping. The interface could not correct it. The shortfall was on the supply side and we never solved it.
The pilot that never opened
Three stores had agreed to run it, upscale secondhand shops where donations carried real value. That was the right first environment. An item worth eighty dollars is one someone might write about. An item worth four is not.
Hardware was built, intake software was working, counter materials were printed, and staff training was planned. The pilot was set for spring 2020.
Retail closed first.
A store-facing layer had been on the whiteboard throughout: scan data telling shops what customers picked up and what they ignored. After the pilot fell through, that half became the whole plan, and it drifted toward online shopping analytics, further from the store with each revision. Nobody built it. A product designed to make a physical object worth handling had turned into a way to collect data about people shopping online, which was no longer worth building. The four of us moved on to other work.