Second-hand clothing stores are booming, but struggle to grow online. Can AI help?

A photo of Athiya Rastogi.
Athiya Rastogi, who graduated from the quantitative finance stream in the statistics program, is the CEO and co-founder of SnapWrite.

Alexa Battler

Give a picture of a t-shirt to Athiya Rastogi’s AI-powered software and within seconds it writes a detailed web page designed to get that shirt in an online shopper’s cart.

The software detects and describes the t-shirt’s size, brand, colour, materials, condition, usually gleaning 15 to 17 features from a single image. It creates a title and writes a product description in full sentences, strategically using the trending keywords it predicts people and search engines will most likely use (a tactic called search engine optimization or SEO).  

Rastogi’s start-up, SnapWrite, developed that AI and other software to help second-hand clothing stores make a name for themselves in the world of online shopping.

“While in school I ran my own resale store and met a woman who ran a pre-loved clothing store,” says Rastogi (BSc 2020 UTSC). “When I looked her up on Instagram, there was this endless feed of other accounts that were doing the same thing. This is a market with challenges that have not been fully addressed yet.”

Demand for second-hand clothing is exploding, and sellers are adapting to a customer base that increasingly prefers to shop online. But because each pre-loved product is unique, every posting made to sell them must be too. Rastogi says many second-hand businesses struggle to expand online when the need for laborious data entry grows exponentially every time their inventory does. 

SnapWrite’s users, about 50 resellers and counting, including one that has more than 35 stores, can upload hundreds of photos at once and give each item its detailed digital identity in about 15 seconds, a task Rastogi says usually takes a human five to fifteen minutes. So far, the company has digitized 25,000 items and generated more than 700,000 product attributes, saving its users upwards of 6,000 hours. 

“As long as the human eye can see a product’s feature, the AI can see it,” she says. “Doesn’t matter if it’s a low-quality photo or there’s a busy background.”

Aryaman (left) and Athiya Rastogi are siblings and co-founders of SnapWrite.
Aryaman (left) and Athiya Rastogi are siblings and co-founders of SnapWrite. 

To shave even more time, SnapWrite integrates with several major website platforms, including Shopify, Wix and Magento, to automatically sync the clothes’ attributes with websites, so garments can be posted automatically and appear in the right place when a shopper is suggested similar products, without any extra human labour. 

Rastogi knew a substantial chunk of sellers in the second-hand market weren’t businesses, but regular people selling their own clothes through social media, particularly Instagram. That was her chance to modernize the consignment store. Both people and brands can now upload garment photos to SnapWrite’s platform and have them digitized. They can then browse, choose the items they want to resell, send and accept offers, and even put the posting right on their site without having the item in-hand.  

While the AI is also helpful for traditional clothing companies, SnapWrite has managed to get brands that sell new clothes involved in the second-hand cycle. The start-up has an initiative in which partner brands ask customers to bring their used products back to their stores; those items are then run through the AI and — as with the consignment model — offered to second-hand stores to resell. They’ve already saved 25,000 pieces of clothing from ending up in landfills, with no sign of slowing down. 

This could be the start of a kind of digital passport system for used clothing, particularly because the AI generates a unique inventory (or SKU) number to help businesses stay organized. That t-shirt, for example, may get digitized by SnapWrite, sold by a thrift store, then re-donated months later back to another second-hand store that also uses SnapWrite, or uploaded to its consignment platform. Even more time could be saved should the AI recognize the shirt and be able to pull up its data.

Rastogi initially grew her start-up out of The Hub, one of U of T Scarborough’s entrepreneurial incubators, and the company has spent the last year picking up steam. SnapWrite placed first or second in almost every pitch competition it entered, winning around $40,000 in funding (none of which required them to give up equity) from various competitions across U of T and beyond, including ones at The Hub and The BRIDGE. The start-up has also attracted funding from multiple big-league investors, including FounderFuel - Real VenturesInovia Capital and Panache Ventures.

Rastogi is a seasoned data scientist and built the AI technology with her brother, Aryaman — SnapWrite’s co-founder and the person behind the software side of the business — after taking courses in machine learning at U of T Scarborough. Now, they’re focused on keeping the momentum going. 

“There’s no software in the Canadian market solving this problem in the resale market right now, so we decided to do it,” she says.