Online resale is a notoriously difficult business to make money in. Platforms must court both buyers and sellers, and they have to invest in complex infrastructure that enables them to monetise vast, unpredictable inventory that ranges in quality, size and authenticity. That’s on top of the cultural context that pushes up against fast fashion and price sensitivity, plus resistance from luxury brands, whose key objective — selling new pieces at full price — is in contrast to resale’s business model.
Could new technologies be the key to finally cracking it?
Recent advances, particularly in artificial intelligence, are contributing to signs of traction in the space. They’re making it easier for platforms and customers to sell, list and find items. Business models are adapting, seeding new partnerships and social strategies. The hope is that more and better inventory, coupled with lower costs of doing business, will overcome the profit hurdle. Those in the industry also hope this will encourage firsthand spenders to invest in higher quality goods, which have higher resale value, in place of buying limited-use fast fashion.
“We are still in the early days of this, but there is no doubt that AI is helping to make resale more efficient,” says Globaldata retail analyst Neil Saunders. “The big problem with resale is the number of stock-keeping units and the fact the offer is not uniform. AI can help with inputting and with search.” In its October earnings call, Ebay CEO Jamie Iannone said that AI-enhanced listings have generated “several billion dollars” of cumulative gross merchandise volume on the platform.
While it’s still early days, he cautioned, it’s already “having a major impact at scale”. Earlier this month, The RealReal CEO Rati Levesque said that its AI and other tech investments are improving speed and reducing human intervention: “The work we’ve undertaken in 2024 is reshaping how we process items.”
With a new CEO and rising revenues, The RealReal is at a turning point. Analysts weigh in on the company’s next phase as it works towards profitability.
Resale platforms are investing in these types of technology — a bet they’re willing to make even as profitability has been hard to achieve. The RealReal is using AI to predict which items need closer authentication inspections, which items are most in demand, and to predict the right prices — with positive results. The company beat expectations in its third quarter this year, with revenues rising 11 per cent, and losses of $18 million (compared to $23 million the year prior).
Ebay introduced a luxury consignment service last year and a suite of tools that use AI to make listings easier. While recent earnings missed expectations, the fashion category is growing, the company said. Thredup is investing heavily in AI, and recently hired its first-ever head of social commerce. Depop, which was acquired by Etsy 2021, grew 30 per cent year-on-year. In September, Depop introduced an AI tool that can automatically generate a listing description and item info such as category, colour, sub-category and brand, based on a single photo.
By 2028, the global secondhand market is predicted to reach $350 billion, and last year it grew 18 per cent, according to a March report from Thredup, in partnership with retail analytics firm Globaldata. “Overall, the fashion resale market is positive. It has not been unaffected by the wider consumer slowdown, but it has weathered the storm much better thanks to a number of very positive underlying factors such as people searching for value, an influx of new customers and the sustainability trend,” Saunders says.
“What we all know is there are more consumers wishing to participate in the resale space,” says Hannon Comazzetto, founder and CEO of resale tech startup Airrobe, which partners with brands to take the hard work out of listing items for resale. “A part of it is the innovators’ dilemma for retailers: how do they approach this space? Often the solution that will serve the brand today might be quite different from what serves the brands tomorrow. I think it’s really thinking about, what is it that makes sense in five years’ time?”
Acquisition of clothes and customers
Supply of desirable inventory has become a core focus for resale sites. “We have learnt that for marketplaces in the secondary market to prosper, the breadth and depth of inventory becomes important,” says Comazzetto. It’s such an opportunity that LePrix — a business-to-business company that sources luxury goods for secondhand retailers and marketplaces, with clients spanning hotel boutiques, cruise ship stores and TikTok resellers — which started as a consumer-focused business during the pandemic, pivoted to B2B after noticing that sourcing good supply was the biggest challenge, LePrix CMO and co-founder Emily Erkel says.
To improve its acquisitions of valuable inventory, The RealReal has developed a “smart sales tool” that identifies which brands and categories are in high demand, in addition to other data points, to better predict the clients that are most likely to consign at a given time and to help the company’s luxury managers become more efficient in pursuing those items. (The RealReal has recently focused more on consignment, rather than its “direct” business, which acquired unsold merchandise directly from brands.)
This mimics similar tools created by conglomerates Kering and LVMH, each designed to help their sales associates in terms of customer outreach. This led to increased orders and larger average basket sizes compared to the prior year period, and active buyers increased, according to The RealReal’s Levesque, who was recently appointed to the CEO role after 13 years at the luxury resale company. After items come in, The RealReal also uses automation and algorithms to make the process more efficient, having reduced processing time by more than 10 per cent. Using its years of data, it is now able to better determine which items are most likely to need human authentication and to help in copywriting.
Both Thredup and The RealReal use pricing algorithms to determine the best price to list items, with the majority of The RealReal’s items now priced in this way — based on historical sales data, fashion and pricing trends.
Additionally, Thredup is preparing to introduce 360-degree high-definition photos for new items, in addition to automated digital measurements and AI-based flaw detection. “While these seem routine, photo quality, flaw detection and measurement accuracy are critical friction points in shopping secondhand online,” said Thredup CEO James Reinhart, during the company’s earnings call earlier this month. “We should see this translate into improved conversion, lower returns and increased customer retention.”
Thredup also uses AI to help it acquire and retain customers. “This technology is quickly becoming the foundation for all of our onsite merchandising, email, advertising tech and marketing campaigns,” Reinhart added.
While the original innovation of resale platforms was to bring secondhand finds online, the category has found considerable traction in live video-based social commerce, with specialist influencers (focusing on, for example, bags or streetwear) attracting niche audiences on platforms such as TikTok and Whatnot. In October, Ebay expanded its live-stream shopping tool to clothing, and for its Secondhand Sunday event, Poshmark is encouraging its sellers to use its live-shows format. The tech enables people to click and buy items in real time.
Social media and live shopping are increasingly important to secondhand, LePrix co-founder and CEO Elise Whang adds. This summer, Ebay announced a Pre-Loved Fashion Week during New York and London Fashion Week that included a live, shoppable runway show of pre-owned fashion (pictured at top). This January, Thredup’s first head of social commerce, Danielle Vermeer, will start looking at how the company can sell secondhand on social media and through other new formats. This is in addition to new social commerce features from Thredup that will enable creators to curate and share shoppable looks from the marketplace’s four million available items. Similarly, The RealReal has started engaging potential customers — both buyers and sellers — on TikTok and instagram, with a focus on content that explains complex topics such as authentication and sustainability.
Making it easier for people to buy and sell
Thredup and The RealReal are marketplaces that hold inventory; but peer-to-peer companies, which rely on consumers to list their items, are also benefiting from tech-enabled efficiencies. This is especially valuable for the resale industry, as listing individual items for sale can be daunting to sellers, while sorting through those items in search of treasure can be equally overwhelming to shoppers.
These updates are overdue, Comazzetto says. “There has been quite little innovation in the space, and still today, consumers are being asked to catalogue their products and take and edit photos and upload those photos and fill in these listing forms from scratch… When I first started speaking to customers in the space, there were lots of discussions of, how can you make it easier and faster? But what they wanted to do was cut out the listing process entirely,” she says. To that end, Airrobe integrates with a primary retailer’s e-commerce site to display a potential resale value as customers are shopping. Once they have bought the item, it is added to their universal digital wardrobe, which the customer can then use to quickly list the time for sale on participating resale sites, or on the original brand’s resale marketplace.
Poshmark recently began beta testing a feature called Smart List AI, which automates and streamlines the task of creating listings. Poshmark has also recently begun partnering with brands to offer instant resale listings; in September, it linked up with Coach’s Coachtopia line, which uses Eon digital product passports (DPPs), enabling owners of Coachtopia products to scan the item and semi-automatically list the item for sale on Poshmark. It followed that up with a similar partnership with Reformation. On average, listings through these “Resell on Posh” partnerships were up 40 per cent week-on-week through October, according to the company. Expect similar partnerships with additional resale sites soon. (The RealReal has also partnered with Eon, paving the way for any brand using Eon’s DPPs to unlock resale on The RealReal.)
Ebay, meanwhile, is heavily invested in tech that enables fast-tracked listings. A year ago, it began introducing a series of AI tools designed to decrease the time and effort it takes for sellers to list items, including intelligent pricing recommendations, AI-generated descriptions (such as a suggested title, category and description) and photo enhancements. Ebay’s Iannone said that already, these capabilities, which are still in beta mode, are “making a major impact at scale”, having “notably decreased time to list, improved listing completion rates, increased conversion, sold items and GMV per listing attempt, and elevated adoption of promoted listings”. Already, more than 10 million sellers and 100 million listings have used these “magic listing” tools, with AI-enhanced listings generating “several billion dollars” of cumulative gross merchandise volume. “We’re still continuing to ramp up, improve the models and expose more and more people to Magical Listings — so I feel really good about it,” he said.
Vestiaire Collective is working on a price recommendation tool, which will look at the age and the wear and tear of a specific item, plus the number of other comparable items and the desirability of the specific brand or designer at the time, CEO Maximilian Bittner recently told Vogue Business.
Buyers are also getting a boost in efficiency, with new tools that help them sort through the millions of individual product listings. In October, Vestiaire Collective — which is currently undergoing a crowdfunding campaign as it eyes a potential initial public offering — enhanced its search functionality with computer vision that can better read images and match them to the words in searches. Image search is also in the pipeline.
Similarly, Thredup’s new tools include the ability to use natural language, spanning multiple details or specific scenarios, in their searches. They can also upload images to find matching items for sale. It’s added a chatbot that can help people create outfits by generating a head-to-toe look based on a natural language prompt, which people can subsequently refine by details such as colour or style.
“The adoption of image search has been rapid as customers have made it one of their go-to tools to find what they want on Thredup,” Reinhart said, adding that generative AI is now able to curate items based on thematic moments, emerging trends or niche demand (such as “ugly holiday sweaters”). This capability, he added, wouldn’t have been possible as little as a year ago. The rapid rise in sophisticated generative artificial intelligence is to thank, he said. “We continue to believe that AI disproportionately benefits our marketplace relative to other marketplaces and retailers,” he added.
While resale platforms are still chasing consistent profitability, they are hoping that AI will prove to be the magic technology that enables them to scale without dramatically increasing costs. Still, it will take time to pay off — even if the benefits of AI are immediately obvious, as buyers and sellers will be the key barometer for success. Reinhart is convinced it will help secondhand blend in. “Generative AI can significantly enhance the secondhand shopping experience. For years, our dream was to build an experience that was indistinguishable from shopping new. Advancements in generative AI are quickly making that a reality.”
Sign up to receive the Vogue Business newsletter for the latest luxury news and insights, plus exclusive membership discounts.
Comments, questions or feedback? Email us at feedback@voguebusiness.com.
