• Depop's new gen-AI feature creates item descriptions based on photos that users upload.
• The tool has boosted the number of listings on the company's website and saves sellers time.
• This article is part of "CXO AI Playbook" — straight talk from business leaders on how they're testing and using AI.

Depop is an online fashion marketplace where users can buy and sell secondhand clothing, accessories, and other products. Founded in 2011, the company is headquartered in London and has 35 million registered users. It was acquired by Etsy, an online marketplace, in 2021.

Situation analysis: What problem was the company trying to solve?

Depop's business model encourages consumers to "participate in the circular economy rather than buying new," Rafe Colburn, its chief product and technology officer, told Business Insider. However, listing items to sell on the website and finding products to buy take time and effort, which he said can be a barrier to using Depop.

"By reducing that effort, we can make resale more accessible to busy people," he said.

To improve user experience, Depop has unveiled several features powered by artificial intelligence and machine learning, including pricing guidance to help sellers list items more quickly and personalized algorithms to help buyers identify trends and receive product recommendations.

In September, Depop launched a description-generation feature using image recognition and generative AI. The tool automatically creates a description for an item once sellers upload a product image to the platform.

"What we've tried to do is make it so that once people have photographed and uploaded their items, very little effort is required to complete their listing," Colburn said. He added that the AI description generator is especially useful for new sellers who aren't as familiar with listing on Depop.

Key staff and stakeholders

The AI description-generation feature was developed in-house by Depop's data science team, which trained large language models to create it. The team worked closely with product managers.

Colburn said that in 2022, the company moved its data science team from the engineering group to the product side of the business, which has enabled Depop to release features more quickly.

AI in action

To use the description generator, sellers upload an image of the item they want to list to the Depop platform and click a "generate description" button. Using image recognition and gen AI, the system generates a product description and populates item-attribute fields on the listing page, including category, subcategory, color, and brand.

The technology incorporates relevant hashtags and colloquial language to appeal to buyers, Colburn said. "We've done a lot of prompt engineering and fine-tuning to make sure that the tone and style of the descriptions that are generated really fit the norms of Depop," he added.

Sellers can use the generated description as is or adjust it. Even if they modify descriptions, sellers still save time compared to starting with "an empty box to work with," Colburn said.

Did it work, and how did leaders know?

Depop has about 180,000 new listings every day. Since rolling out the AI-powered description generation in September, the company has seen "a real uplift in listings created, listing time, and completeness of listings," Colburn said. However, as the tool was launched recently, a company spokesperson said that specific data was not yet available.

"Aside from the direct user benefits in terms of efficiency and listing quality, we have also really demonstrated to ourselves that users value features that use generative AI to reduce effort on their end," Colburn said.

Ultimately, Depop wants sellers to list more items, and the company's goal is to make it easier to do so, he added. Automating the process with AI means sellers can list items quicker, which Colburn said would create a more robust inventory on the platform, lead to more sales, and boost the secondhand market.

What's next?

Colburn said Depop continues to look for ways to apply AI to address users' needs.

For example, taking high-quality photos of items is another challenge for sellers. It's labor-intensive but important, as listings with multiple high-quality photos of garments are more likely to sell. He said Depop was exploring ways to make this easier and enhance image quality with AI.

A challenge for buyers is sometimes finding items that fit. Depop is also looking into how AI can help shoppers feel more confident that the clothing they purchase will fit so that their overall satisfaction with the platform will be enhanced, Colburn said.

Evan Duby had been a real-estate broker for just a few years when he decided to try an unusual method of selling a home. Five buyers were offering to pay similar amounts for one of his listings, a one-bedroom co-op unit in a leafy Brooklyn neighborhood with an asking price of $485,000. Rather than instructing the house hunters to submit their best offers and cross their fingers — as is customary in the US — Duby, with his client's permission, convened an auction. The buyers gathered on a conference call, where they signaled their willingness to pay as Duby raised the price in $5,000 increments. The home sold for $505,000. The sellers walked away satisfied, as did the winning buyers; the losing bidders were disappointed, but at least they knew where they stood. For Duby, it was a revelation.

"I don't know what possessed me," Duby tells me. "I was just sort of trying to see, is there a better way to do this?"

Trying to buy a house can feel like playing a game of poker in which one player holds all the cards. When the seller's agent tells you they're weighing another bid, or even 30 other bids, there's no way to tell if their claim is bluster or fact. When you lose, you may not know whether another $10,000 would have sealed the deal or if your insistence on an inspection tanked your chance. The nagging uncertainty isn't limited to buyers. Even in a hot market, a seller may leave the closing table unfulfilled. Did their request for "highest and best" offers actually yield the highest and best? It's hard to say.

It doesn't have to be this way. In Australia, about a third of homes sell via auctions that wrap up in a matter of minutes. Sellers get to see exactly how far buyers are willing to go to nab their dream home; buyers gain a clear picture of what it takes to win in the market. The openness and simplicity stand in sharp contrast to America's system, in which buyers write blind offers and then pray theirs meets the mark.

In the US, open auctions are usually reserved for swanky mansions or, more often, distressed properties facing foreclosure or extensive repairs. Mention an auction and people are likely to ask what's wrong with the house, or how lavishly expensive it is — or, simply, why? Real-estate agents haven't been too keen on making the process more transparent: Conventional wisdom says that asking buyers to submit their best and final offers will elicit the highest price. A FOMO-filled buyer, the thinking goes, may unknowingly blow the competition out of the water and deliver a windfall for the seller.

Despite all that, the idea of open auctions is more tantalizing than ever. Buyers and sellers are exhausted from years of opaque bidding wars — even these days, with the market substantially cooler than it was a couple of years ago, a lack of inventory means homes may still draw multiple offers. The real-estate industry is notoriously resistant to change, but recent class-action lawsuits have rewritten the rules about how buyers pay their agents and opened the door for more overhauls down the line.

Capitalizing on this feeling, a small cadre of companies are trying to bring versions of the Australian model to the States. They face an uphill battle. Duby, who recently started GoEx, a venture-capital firm focused on real-estate tech, is squarely in the ripe-for-change camp, but even he'll admit it's hard to shift the status quo. For roughly a decade after that first auction, he continued to broker deals the conventional way — for American sellers, Aussie-style auctions were a tough sell. But just because they haven't caught on in the States doesn't mean they're destined to remain a pipe dream. Duby imagines a not-so-distant future in which prequalified buyers bid for homes online as if they were picking up a rare watch on eBay.

"I don't see why we don't do that," Duby tells me. "I don't see how that doesn't help."

The man could be mistaken for a pastor: crisp gray suit, arms stretched toward the heavens, a crowd gathered before him. But he's here to sell real estate, not religion.

"Reflect on this absolutely fantastic opportunity in front of you," he booms in a distinctly Aussie drawl. "Not me — the house!"

With that, the auction begins. The property at stake is a quaint one-story home surrounded by a white picket fence in a suburb of Melbourne. The bidding starts at 1.26 million Australian dollars, but the price climbs as the auctioneer needles buyers to dig deeper: "You know you want to!" In the end, it sells for more than 1.5 million Australian dollars, or about $980,000. The whole thing takes less than half an hour.

I watched all this unfold on TikTok, where the account @AuctionReporters maintains a steady stream of these strangely addictive dispatches from Australia's real-estate market. Every now and then a video like this will go viral among Americans who balk at the ritual. An open auction is so vastly different from the secretive practice here in the States that it can break our brains — in a reply to a similar video on X, someone posted, "this is real???"

Real-estate auctions are a time-honored tradition in Australia, dating back more than 200 years to its days as a British colony. Their popularity varies based on location and the strength of the market: In boom times for home sales, more sellers turn to the gavel — a study by Kenneth Lusht, a professor at Pennsylvania State University, found that in some pockets of Melbourne, auctions accounted for as much as 80% of home sales during periods of particularly strong demand. A later study of sales from 2011 to 2019 in the states of New South Wales and Victoria, home to more than half of the country's population, found that 30% to 40% of listings went to auction during that period.

Auctions are risky, to be sure. Homeowners publicly disclose the terms and privately set a reserve price, or the minimum amount they'd accept — if the bidding doesn't reach that figure, the house doesn't sell. A study published in 2022 described such a house as carrying a "stench of failure" and found that it was more likely to sell at a discount later. An auction is basically impossible to stop once it's in motion, and sellers may not always be pleased with the results. But in times of healthy demand for homes, auctions can deliver benefits for sellers looking to ride out the frenzy. The study on the risks of auctions also found significant upsides: Successful sales tended to achieve prices that were 1.2% higher than comparable "private treaty" sales, in which sellers set an asking price and then wait for bids to roll in. A separate 2010 study of a broad swath of Australian home sales also found that auctions tended to yield higher selling prices than the alternative.

It's hard to imagine regular homes in the US trading hands this way. But a handful of companies have proposed a middle ground between the public spectacle of Australia's auctions and America's behind-closed-doors strategy. Final Offer, in Massachusetts, is one online marketplace that mediates auction-ish sales. A real-estate agent can list their seller's property on the platform, specify their asking price and their terms, and input a "final offer price" and specific terms of sale, or the amount a buyer can agree to pay in order to stop the bidding and win outright (similar to eBay's "Buy It Now" feature for auctioned items). When a buyer makes a qualifying offer, the clock starts ticking: The seller can choose to reveal the price and terms of any offer in contention, and interested buyers can try to exceed the bids before the window closes.

Here's a real example: Late this summer, the owner of 5818 Ipswich Road, a two-bedroom home built in 1951 in Bethesda, Maryland, listed it on Final Offer for $650,000. Buyer 1 submitted an offer of $658,125, and the seller agreed to take it if no better offers came in over the next three days. Other buyers soon entered the picture: Buyer 2 bid $661,500. Buyer 1 responded by going $3,000 higher. Over the next two days, Buyers 3, 4, and 5 threw their hats in the ring, and the price climbed above $800,000. At the eleventh hour, Buyer 6 emerged with a bid of $810,573. Then came the kicker: Buyer 1 made the "final offer" of $850,000, ending the bidding process. The entire saga is available for anyone to see on Final Offer's website.

In real estate, it turns out, a little transparency goes a long way. "You're giving buyers information they've never had before about what the seller really wants," says Tim Quirk, who cofounded Final Offer and serves as its chief strategy officer. In a typical sale, spurned buyers rarely walk away knowing what would have won — maybe with that knowledge they would have been willing to up their price or adjust their terms, perhaps waiving an inspection and agreeing to buy the home as is. And sellers, even when they take a deal, never quite know if someone might have gone even higher if they knew what they were up against. "What ultimately ends up happening is you get remorse on both sides of the table," Quirk tells me.

Final Offer is still small — Quirk said that a little more than 1,000 homes had sold on the platform in the two years since it started. Sellers on the site don't have to disclose the prices and terms of offers that come in and can opt to let buyers see only that other offers have been made. But Quirk tells me more than 80% of sellers choose to make the bids public.

SparkOffer, which last year was acquired by Auction.com, is another platform that aims to give buyers a sense of their competition. The site shows buyers how many offers they're competing with and is beta testing a new feature that assigns each bid a score based on its price and proposed terms; when other bids come in, buyers get to see how their score stacks up. They don't know the exact details, but they'll at least have a sense of where they stand and what it might take to climb the ranks. Sellers get to outsource the messy back-and-forth of negotiations to a platform that prods buyers to sweeten their offers — while preserving the possibility that a buyer may unwittingly overbid.

Neither Quirk nor Mike Russo, the founder of SparkOffer, thinks of his platform as an auction exactly. For one, home sales aren't just about price. When a buyer makes an offer, they can propose contingencies, or conditions that need to be met in order for the sale to close. They might request an inspection and ask the seller to make repairs if something comes up. They could retain the right to back out of the deal if an appraiser values the home lower than expected. The buyer could also ask for concessions, or some money back from the seller to cover stuff like closing costs. At the height of the pandemic-era homebuying frenzy, some desperate buyers threw caution to the wind and waived these contingencies — they were simply tired of losing. Bottom line, not all buyers are equal in the eyes of a seller.

To reach widespread acceptance in the US, an auction-esque model would have to let sellers choose the terms that work best for them. Most important, new marketplaces would have to convince American sellers — and their agents — to turn away from tradition. That kind of cultural shift is a tall order. "I don't think the American consumer is ready for auctions yet," says Rob Hahn, the CEO of Decentre Labs, a company he started in 2022 to bring online real-estate auctions to the masses.

But it's unclear who really benefits from the system we have. Buyers rarely get feedback that could help them make stronger offers in the future. Maybe as a seller you'll get lucky and a buyer with blinders will overpay for your house — but that approach is shortsighted, since most sellers have to turn around and purchase another home. The biggest beneficiaries, Duby tells me, may just be real-estate agents: When buyers and sellers are shuffled through this murky process, they'll look for a professional guide.

Australia isn't a perfect analog for the US. The country doesn't have a system of local databases where most homes are listed — as we do in the States with the multiple listing services — which makes it tougher for buyers to find homes. And while most sellers are represented by agents, most buyers are not. But still, hardly anyone is proposing we copy and paste this Australian tradition. The country's real-estate market does tell us, however, that another way is possible.

American consumers may not be ready for auctions yet, Hahn told me. "But it doesn't mean that it's not going to happen at some point."

James Rodriguez is a senior reporter on Business Insider's Discourse team.