Can AI stitch up a profitable future for Resale-as-a-Service platform providers like thredUP?

Stuart Lauchlan Profile picture for user slauchlan March 5, 2024
Generative AI is going to be a game changer, according to thredUP CEO James Reinhart.


On the gen AI stuff, I remain very bullish, on its ability to improve our business really disproportionately compared to others. 

That’s the sort of comment that frankly we’re all expecting every CEO to be coming out with this year as standard. In this case, it’s James Reinhart of Resale-as-a-Service (RaaS) platform provider, thredUP. As noted previously, the RaaS space is a hugely lucrative one - in theory.  

thredUP’s own research posits that this year some 10% of the global apparel sector will be made up of secondhand items, with a projected market size of $351 billion by 2027. But the harsh reality is that companies such as thredUP remain loss-making. 

So is AI a ‘silver bullet’ waiting to be fired? Reinhart is all in at any rate: 

Given the long tail of product, the constantly changing nature of our product, we really rely on sort of the dynamic nature of the technology to do a lot of work that would otherwise be done by inferior algorithms. So, I am very bullish on its ability to delight the customer on the front end. And I think we are working on a number of things that will start to materialize this year that I think will really change how consumers shop resale. 

He adds: 

On the operations side, we have been employing AI in a number of ways in our DCs (Distribution Centers) for years. But I think just in the last 12 months, you have seen the step function change in what the technology can do and I think it has real implications for how productive our operations can be and what the margin profile can ultimately look like.

OK, so that’s the theory. How’s that working out in practice? In terms of improving the customer experience, Reinhart points to the debut of an an AI-powered search experience that is pitched as making it easy and intuitive to find any secondhand item on thredUP: 

Our vast selection of inventory is one of our biggest assets, but it also creates challenges for buyers, as they shop up to 4 million unique secondhand items at any given time. This new search functionality significantly enhances the secondhand shopping experience in our marketplace by combining visual language with personal style [and] by enabling buyers to curate style inspirations effortlessly. 

Whether it’s by searching for a popular item, like a satin cocktail dress, or a descriptive trend or look, like Sunday brunch, or a phrase that evokes a notion, like Academy Award, thredUp can help shoppers find exactly what they want. It’s not only fun to use, but it also has that sense of magic to it. Sometimes you just can’t believe how good the technology is at delivering relevant results. Early indicators show an increase in searches precession, a higher advice conversion of items from search and higher click-through for individual product pages.

Next up comes - inevitably - generative AI. Reinhart says: 

We have begun to leverage generative AI technology that will soon give customers the ability to create outputs they love using just a text description. 

He cites the example of a someone looking for an outfit to wear on an upcoming trip to Hawaii. Using natural language prompts the generative AI tool comes up with a floral crop top, a white maxi skirt, paired with highly embellished sandals: 

We want to create an outfit from popular magazines or style influencers or runway trends. We can now easily do that while delivering shoppable secondhand product up to 70% off what a consumer might pay new. The list of outfits that can be generated through this tool is endless, restricted only by the imagination of our buyers. We’ll be leading these style inspiration touch points throughout the product experience over the year ahead and look forward to sharing more soon.

This can provide significant competitive advantage, he argues: 

I want to emphasize that AI is an enormous leap forward for us in bringing a motion and storytelling to the millions of unique shopping journeys that regularly happen on thredUP. Given the breadth of our offering and the limitation of not having own model photography in our core product experience, we believe generative AI technology disproportionately benefits [us] to manage a marketplace like thredUP compared to other apparel or peer-to-peer marketplaces.


Behind the scenes, AI is being implemented across more operations in the distribution center network to improve throughput and productivity, Reinhart explains: 

Once the garment has been photographed, we employ advanced AI technologies to extract a wide range of detailed characteristics of the item from its image. This capability not only enriches our inventory database, but also streamlines the organization and processing of items. This has improved operational efficiency and the accuracy of our product listings, resulting in better search and personalization in our marketplace.

We see near-term opportunities for generative AI to improve visual merchandising and add more engaging content to the shopping journey without us having to use expensive on model photography. Eventually, you can imagine a role where AI not only supplements manual photography, but replaces it. 

All told, generative AI is going to be game-changer, he concludes: 

These AI-driven initiatives have enabled us to set new standards for efficiency and accuracy and are paving the way for continuous innovation and potential margin expansion. So much of what we believe we could achieve over the next few years through our own software and industrial engineering development has now become readily available - and it’s cheaper and faster than we imagined.

My take

Sound in theory and all makes sense. But will it take firms like thredUP into the realms of profitability? Time will tell. 

A grey colored placeholder image