I'm mic'd? Can you hear me? All right, great. Just, my name is Dylan Carden. I am the Analyst here that covers ThredUp. With me, I have James Reinhart, CEO, Sean Sobers, CFO of the company, and yeah. Oh, there are disclosures that are relevant to this conversation on our website. We still can't get that language exactly right, but that gets you the thrust of it. Thank you.
Yeah, thanks for having us.
I know we're going to make this more interactive, but there is a steeper learning curve, I think, for resale, for your business in particular. Can you just give us the elevator pitch with a slant towards your competitive advantages and -
Sure.
- the industry more broadly?
Yeah. ThredUp is a marketplace for second-hand clothing, women's and kids'. Company is about 15 years old. The core insight was that we all have closets full of clothes that we don't wear, and we thought that there was an innovative way to figure out how to reprocess clothing at some scale and make that available for consumers. We started as an asset-light business. It was just connecting buyers and sellers. I think the core insight was if you could really lock down the supply chain, you could really change the dynamics in the marketplace. In 2012, we started building infrastructure to process clothing at some significant scale. The competitive advantage that we were starting to develop was this unique supply chain for taking back vast amounts of clothing.
The way the business works is we send you a prepaid bag, or you can use a box that you have. You send us back all your clothing, we process it, we put it online, and it's effectively now the world's largest online thrift store. We invented this supply chain that was built purposely for that. Along the way, what we realized is there's no barcodes on clothing. To do this at some scale and efficiently, you need to build a data set. We started cataloging vast amounts of data, 35,000 brands, 100 categories, all in service of building this price arbitrage model. That's the second pillar. The third piece was building a consumer brand, right, that connects buyers and sellers, drives high liquidity. We often talk about building a unique supply chain, building a data business, building a marketplace.
Each one of those is actually quite hard in and of themselves, and the power in the ThredUp business is we built all three. I think what you see today now is ThredUp as a managed marketplace. We're really the only game in town for goods at this price point. We're continuing to invest in the business and it's been good.
I'll jump a question then, because you've spoken to investing $400, $500 -
Yep.
- million.
Yep.
Your market cap has been.
$400 million, $500 million.
$400 million, $500 million.
Yeah.
I'm just curious, how you think about the value you've created as it relates to a moat.
Yeah.
Right? As far as what is replicable, what is proprietary, and what it really allows you to do that others can't.
I think if you go back to when we started the business, there was no playbook for this. The eBays of the world, right, have been around, eBay's been around 30 years now at this point, right? It's connecting buyers and sellers. They take a transaction fee. That's sort of how it works. There was no playbook for managed resale, right? Of like, what does it look like to take all the product back? I think the first probably five or six years of the business, it was, can this work? I think it was a classic R&D experiment of can you build infrastructure? Can you take hundreds of millions of pieces of clothing, right, and turn this into an economic engine?
It really wasn't until, honestly, 2020 was probably when we started to build out Atlanta, that you were able to do this at sufficient scale. I think the first seven, eight years was like, can you do this on the internet, right? I think people have sort of sometimes lose the plot on that of like, we actually spent a lot of years figuring this out because it's so darn hard, and now we figured it out and we have an economic engine, and it's nearly impossible now to compete with ThredUp, right? It's not truly impossible, right? I mean, but yeah, close to $500 million of capital investment, consumer brand software. There's no real great vector to compete with us.
It's like if you think about the classic ways that somebody would come in and try and compete with ThredUp, it's like, "Oh, we're going to offer a better selection." It's like, "No, we have better selection." "We're going to offer lower prices." "No, we've got tremendous" "We're going to compete on logistics and speed." Like, "No." "Well, we're going to compete with a great brand." It's like, "Mm." Right? You run down the list. I think what people often tend to miss in this is just how defensible this business is that we've built. Now we have an economic engine that's free cash flow positive, business grew 20% last year. I think the building blocks are there for this to be quite an exceptional business over time.
Yeah. Dylan, I'd probably add two things, that the $400 million-$500 million that we invested, we have now infrastructure to support a business that can easily do $500 million in revenue, right?
Yeah.
We're not there. We did $310 million last year. I think that's all the investments. We have some leveraging that we have and the infrastructure that we've already built. I think what's happened from an economic standpoint on unit economics is it was around IPO time, our contribution margin was around 27%, and today it's in like the mid-40s. We've been able to show that you can do this, not only do it online, but you can do it online profitably, which I think was a big question for a long time, and I think last year was the year where we were able to be like, you can look to the bottom line, see positive cash flow, you can see positive earnings. It's also not, I think, the other thing, the misnomer is it's not a fashion company.
It's not a brand, right? It's a logistics infrastructure company, and we compete at this layer. If you don't believe resale is going to continue to grow, you have to believe we're not going to wear clothes in the future.
You stole my line.
Right. Yeah. Somehow people turn around one day and they're like, "Nah. I'm only going to buy new." There's nothing to suggest that that is the arc that we're on. I love doing the hard work at the infrastructure layer.
Well, that's exactly how I see it. Yeah, my line is typically.
Sorry, yeah.
No one wants to talk to a consumer apparel analyst, but I say, "I know nothing about you, but I know tomorrow you're not going to go outside naked." Right. That's my relevance, our relevance.
Yeah.
No, I view resale as one of the major structural shifts in apparel.
Yeah.
We've moved online. We're now half online, right?
Yeah.
Off-price has had a fantastic run as far as value is concerned. It makes sense to me that you are inheriting sort of a certain off-price vector of good value, dynamic shopping experience.
Yes.
Tends to be that there need to be catalysts for that type of a shift. I would call your business and the investments you've made a catalyst. I know you do kind of a certain amount of industry analysis.
Yeah.
Where are you now as far as a closet industry penetration rate, resale in total? How do the dynamics play out between kind of mature Goodwill, eBays? How do you see the sort of landscape shifting?
Yeah, we publish a resale report every year. You can find it at thredUp.com/resale. It talks about the growth of the industry. We expect it to be $80 billion or so by the end of the decade. It's growing four or five times faster than traditional retail. What's so interesting about the data is every generation of young people is adopting resale at higher rates. When you kind of extrapolate those curves out.
Yeah.
You either have to believe that all these habits that people develop when they're younger, they just literally wake up one day and they're like, "Never mind." The penetration of resale in people's closets, if you just kind of think about this stacked bar, is going to be pretty substantial over time. I used to think maybe the resale estimates for 2030 were high. I think they might be high for 2030, but they're way too low for 2040.
Yeah.
I just think this will be the fabric of what people do over time. I think off-price is a great place to point to. Off-price, TJ Maxx, Ross, Burlington, these guys got started in the early 1980s.
'70s for some.
Right. The businesses didn't hit these inflection points until the financial crisis -
Yeah.
- in 2008. They're building stores, they're figuring out the merchandising assortment. They're waiting for this customer shift, and now, these companies are massive. I think resale, it's not an exact match, but I think has some of the same similarities for how consumers will adopt this over time.
When they were shut out of inventory -
Yeah.
- for a long time. I think that's a similar story. They're stigmatized from a brand standpoint.
Yeah.
All but Ralph Lauren. I think, to the destigmatization of resale, but it does speak to inventory availability.
Yeah.
Is the bottleneck access to goods? It was sort of notable in your last earnings that you're now going after sellers.
Yeah.
Right? Kind of speak to unlocking.
Yeah, I think, over the past 10 years, we've never spent a single dollar acquiring sellers. We've just had an inordinate amount of stuff. We've been focused on the buyer side. I think what has changed now in the last probably years, we launched our Premium Kits at the beginning of 2025, and we saw a real opportunity there.
Explain what that is.
Our Basic Kits is you would send us all of your stuff. There's a basic service fee. We pay you out. We did a bunch of testing in 2024 that suggested that consumers, and again, we just do women's and kids, so for the men in the audience, think of your wives, your children, your sisters, and your mothers. That they wanted an experience where they had more control.
Right.
More control over the product experience. If I sent you a bag, I'd have opportunities to reclaim items more efficiently. I can control pricing. I could have a longer consignment window. Effectively, think about it as like, I'm willing to pay for a little bit more control.
Yeah.
That business was up and to the right out of the gate. It went from 0% of our mix to single digits and then low double digits in just a few quarters. That indicated to us that there was opportunity to capture more of that market. What we see in the broader market with RealReal as an example, they're trying to continue to elevate the brand, which I totally understand, and probably moving away from the pocket of inventory that we want.
Yeah.
We're going out to acquire sellers, not for the core basic business, but for that premium mix. I think the prices of those items are double what we see in the business. I'll just take it a step further. We launched direct listings, which maybe you will ask about, but just to give you the continuum. As part of the learning with the launch of our premium product, customers said, "This is great, but there's also a few items that I don't send you guys. I want total control over those items. I want to sell them on eBay. I want to sell them on Poshmark or something." It's like, you have these two platform, isn't that kind of annoying? Everyone's like, "Yeah.
That's the industry.
That's the industry. We launched our direct listings business as a direct reflection of that, which is, okay, send us your basic stuff, but if you want, send us your premium stuff if you want. If you've got those few special items, that Reformation dress or whatever it is, and you want total control over that item, you can list it yourself. The through line of all of this, to wrap up to where you started is, we want to dominate the supply chain. The whole vision of the business from way back in the day was controlling supply in this market, and I think we're making the steps each year to do that. I think if you roll the book forward three or four years, I think you'll see that we can dominate a large swath of this.
What does that look like buying supply? I get sort of what's really strange about it too, to me and to you, I think as well, is that you start charging fees.
Yeah,
You start getting more, better inventory. The RealReal speaks to their biggest competitor is laziness, human laziness.
Agreed. I agree with that. Yeah.
What is it between turning your business into a service that sort of unlocked deeper pockets?
I think somehow people, the psychology of when something is free, we don't value it. Right.
Google Calendar. Yep.
I think you're in this position that when you start to charge a fee, people are like, "Oh, this is actually saving me a bunch of time." It's doing good in the world. Yeah, there's a bunch of economic research around this value equation. Yeah, by charging it, we improved. Yeah. We started charging fees, we got better stuff, more stuff faster, right? If we had one of those three, we thought would happen. Yeah. Right? I think that acquiring premium sellers, I think it just makes sense to go out there and help people understand what we do, what the value proposition is.
Mm-hmm.
The whole strategy is just, hey, we're here to provide a compelling service for you because it's crazy. As we feel like the business has been around for a decade, it's still only low double-digit penetration.
Yeah.
Right? It's a good and growing business with huge upside, right? You've got millions of college graduates every year who become another big opportunity for us.
Taking share from who?
Yeah.
Everyone always punches on the department store.
Yeah.
Get that one out of the way.
Oh, I know I can't even make fun of the mall anymore because malls are back.
Yeah. Right. That's true.
I think it is generally taking money away from just department stores. Yeah.
You still have 15% of the industry.
Yeah. There's still a lot of product out there, but I think you're seeing off-price and resale just continue to eat up some of that.
Right. Is there a point where this competes more directly with off-price?
When we survey customers still, I think they are shopping at off-price, they're shopping at ThredUp. What everybody's looking for is a great deal. Right? I think the thing that off-price provides you is an in-store experience. Yeah. The thing that ThredUp provides you is not an in-store experience. Right? One you can do during your day, and one you can do at night. I think there's a nice segmentation there. It's probably best not to compete directly with TJX, too. No.
You mentioned peer-to-peer.
Yeah go.
It would make sense that that would increase supply, but maybe stickiness, cross-selling between buyers. You can share about what that has done for the platform?
Yeah. Look, it's very early on direct listings, we've been really methodical because it's not like something where Is it open to everyone? It's not open to everyone. It'll be open to everyone at some point soon. I've been really focused on the product experience for customers. I don't want to scale anything that doesn't feel like it's meaningfully better than what's out there. We've been watching the cohort data and it's good.
In terms of?
In terms of repeat usage of sellers and as a founder, cohort data is my jam. I am really looking at those early curves, and it suggests that people, once they start using it, they really like it. We're going to start to roll it out to more sellers and again, I think it's sort of fitting the thesis of we want to be a one-stop shop for all your selling needs.
Dylan, as you walk through the offerings that we have, we have the standard offering, we talked about the premium offering, and then the peer-to-peer direct listings. I think it's good to understand where ASPs are in that model because the standard offering is around $25, $27. You go to the premium offering, James already mentioned, it's about double. If you go to what we're seeing in direct listings, it's something like 4x the standard start. There's a little nuance that's interesting. It's 4x listed and it sells for 3x.
Yeah.
Kind of goes to the point of when you do your own work, you overvalue your clothes prices. Yes. It still sells at 3x, which is kind of an interesting piece.
Yeah. Just to give you math, it's like call it $25, $50, $75. Yeah. Right? To me, that is like a disciplined offering for what people are trying to do. Yeah. If all the direct listings were $25, wouldn't matter. I'm like, oh, what's the point of that? Right.
You could be my wife into this? User experience.
Yeah, user experience. This is actually one of the only companies I feel comfortable talking about AI with or about. There's lots of different ways to kind of cut this. Let's start with systems. I think it'd be helpful too, what are the real problems that you've solved for in that kind of $400 million-$500 million investment? Where are there sort of tweaks, you mentioned that you can kind of ramp into the infrastructure that you've built, what's still available and how has sort of the new technology, without using buzzwords, kind of helped in that?
Yeah. Just to take you back, I think we were very early in this AI adoption moment. There's some restructurings out there, tech companies. We did that in March of 2024. More than two years ago, we saw the promise of the technology and restructured the organization to be AI first.
I've been very public about saying that I think ThredUp disproportionately benefits from these AI investments relative to others because of the nature of the problem we're trying to solve, which is this long tail skew. You don't know what's coming. You need to attribute items in dynamic ways. The way that we have thought about the problem is, number one is just rebuilding the search infrastructure. At the end of the day, if you're going to process this year, we'll put online, I don't know, 25 million-30 million items, something like that. If you're going to put on that number of items, attributing it so people can find what they want is critical. We used to add, call it 10 attributes to an item so that it could be searchable.
We have over 100, and that's just using an AI, sort of OpenAI backend and some stuff that we've built that allows us to sort of tag those items. What the AI can do is tag stuff in ways that you would never imagine an item being tagged.
Which is all image-based.
It's all image-based. Like, we're in Chicago, right? Let's say we're getting a bunch of stuff for the Bears in, right? We've got operators and we've got some basic visual stuff that's tagging the logo for the Chicago Bears. We don't catch everything, right? The image search can now tag everything exactly the way you would want.
Also when a customer's now searching for Bears stuff, they see some stuff, and then they're like, the technology's good enough for them to be like, maybe they're shopping for Cubs stuff. Maybe they're shopping for White Sox stuff. It's spring in Chicago. What else? The way that the technology now can pivot what you might be looking for is vastly superior to what old algorithmic work could do. It's rapidly changing how the consumer shops, and it's all built on its ability to tag and drive inference. That's kind of one. The second piece is in our distribution centers, just improving imagery, measurement, just everything about how we process items has been exceptionally much faster than I thought. We're lowering our inbound costs, our outbound costs, improving the quality of our imagery.
Think about these as two foundational things, and now we're just building on top. We rebuilt the search thing and now we're just layering products one after another on top of it. That's pretty exciting.
Sean, as you think about the contribution margin per unit -
Yep
- you haven't put a target out there, but it would seem that you kind of continue to push that up or?
Yeah, no, it should improve across the board. You think about the backend infrastructure, James talking about the AI help is, when I first started ThredUp seven years ago, there was a person that opened the bag and looked to see was it okay, the item. There was another person that put in the 6 to 10 items. Those people don't exist anymore. The attributes. The attributes.
Yeah. Yeah.
Which were the only things you could search on.
Yeah. The only things you could search, trust me. Size, color, type of item. Yeah. A lot of time you would search and you'd come back with nothing. Right. Yeah. We had it. I think that the AI is one of the best things it's done from the search perspective is now you can find whatever it is you're looking for. It's rare you're going to get a response that isn't 1,000 items. It's something that will drive automation at the DC, it'll drive engagement at the front end. It'll improve the financials across the board. It's definitely a game changer. I think the thing that I'm also really focusing on, though, it's very easy, I think, for all of us in the room to see how AI can reduce costs. Yeah. Right?
It's very clear what it can take out of the P&L. In fact, we had a leadership offsite last week. We had about 30 VPs together. One of the questions I asked them was, what percent of your team's current workflows are likely to be automated by AI in the next couple of years? The average was 55%.
Wow.
There's no question the cost leverage, right, that this is going to provide. I think the harder challenge for founders and product companies is how is it going to meaningfully change the customer experience? Yeah. Right? I think I'm spending all of my energy thinking about how does this technology really delight buyers and sellers over time. For sellers, it might be instant recognition of direct listing items, right? It might be voice. We're working on a glasses app, right?
If you believe people eventually will wear Warby's glasses or-
Some variant the Ray-- Right?
You just have glasses on in your closet and you're just picking up clothing and there's a camera running, right? Stuff like that, where I think that is the most exciting thing because not only then will you combine the cost savings, but you'll combine real customer delight, and I think it'll feel like magic. Like the movies.
That'll be like a, hey, you could sell this for X amount kind of thing?
Yeah. There's just something running in the background that's always giving you style advice. It's plugged into your calendar, right? There's a lot of these ways it can integrate to help you make your life better. I think secondhand can do that in ways because we sell so many brands, we can do outfitting in ways that is harder for a traditional merchant.
On the logistics side of it, too, because then can you bypass the DC? I guess because you're famous for your kind of the conveyor belts.
Yeah, yeah.
Right? Some of the conversation's now going to do you store it in a more convenient way?
Yep.
I guess where are those efficiencies in the model? Photography's always been a bottleneck.
Yeah. There's lots of efficiencies. We do everything on hanger. There's lots of conversations, do we need to do anything on hanger anymore, right? Right. You can change. You could get more automation and robotic without the hanger. So there's opportunities there. I think photography is something that takes a lot of time.
You take the item out, you put it on the hanger, it goes to photography, it gets put on a mannequin, the picture gets taken, it gets taken off the mannequin, put back on a hanger. What if we didn't have to take that type of photo? Can we take a lay flat? Could AI simulate it on a mannequin? If you do that's probably a third of inbound processing from the time we open the bag.
Yep. Huge.
I think there's lots of opportunities there, and I think we've messed around with AI imagery already. We were putting them on AI models, like fake humans. We're talking about putting it on a mannequin. That's way easier. How do you do it in a way that it makes sure it flows and drapes the right way? I think it's very possible, but I'm not the tech guy.
Yeah, again, I think it just speaks to this infrastructure layer, right? We're just improving the processing costs and the throughput. At the end of the day, you have to have the infrastructure to bring all the stuff in, right? I think that, again, is I think what people kind of miss is like We're improving all the efficiencies in the distribution center, but we still play at the most important layer.
Yeah.
I think you tie it into like where are we today? We're selling items that are $25- $28 ASP, and we have a contribution margin in the mid-40s. Doing all this stuff is going to make that better. I think it's really important to understand, these aren't high-value items, and we have lots of room from a contribution margin perspective. They're lower value items, and we've been able to crack the code to be profitable on a $25 item.
Am I making it up that it was the low 40s a couple of years ago?
Contribution margin?
Yeah.
It was, yeah.
Like 42-ish?
Ish. Yep.
Now I'm going to read into that 45-ish.
You remember numbers.
What about on-.
It's all off the record, right?
Yeah, right. You're not public or anything. What are we doing? Okay, good. On the marketing side, I think similarly, there's some sort of powerful engagement tools. We can talk about Agentic only in that I think they're sort of a there, here as well.
Yeah.
Yeah, open-ended for now.
Yeah, I think on the marketing side, I think once we got rid of the European business end of 2024, we went all in in 2025, in two areas. One, inbound processing, but great quality, and then investments in marketing. I think the mix, I think was more Google. We said sort of publicly last time we were switching to more Meta and Pinterest, which I think are proving to have higher LTVs. This apparently hit Google's radar since they called us. They're like, "Wait a minute.
I had another company actually say, I'm surprised that they said that publicly.
Yeah.
I thought you were probably too small to be on the radar, but I guess not.
Apparently not.
Yeah.
They called with.
Is that going to be a problem?
No, they called with a, "How can we make this better?
Yeah.
They sort of acknowledged.
Great partnership.
Yeah, it was a very good partnership conversation, right?
It's like cancel your cable and they.
Yeah, exactly. Yeah. They sold us DirecTV. It was weird. Yeah. I think we're continuing to evolve the mix. It's been very bottom of the funnel. We know we need to be more mid-funnel, more top of mind for customers. I think we've been doing stuff, I think that's elevating the brand.
You need to build the brand, which is what a lot of DTC companies do.
Yeah, exactly. We did a big rebrand at the end of last year. We're sort of leveraging that. We actually just did our first pop-up in a while in New York City, this past weekend for wedding season, and the line was two blocks long, right? In SoHo. It was like a home run event. I think there's going to be opportunities you'll see over the next year, more activations, things like that do more storytelling now that we've built the foundation for direct response. Because I think you got to have both.
Yeah.
Right? Companies that build brand and then they have no engine, those things flame out, right?
You spend $1 to make $1.
Yeah. I think, but we're well-positioned, I think, for what the next few years look like.
There's three things I kind of want to touch on. One, macro.
Yeah.
You kind of mentioned on your last call that you're seeing some sort of slippage in AOV.
Yep.
Conversion, albeit sequentially, not year-over-year.
Correct.
Just sort of state of the consumer, your consumer.
Yeah. I think the mistake we made on the call was that we said, "Hey, prices were off 3%, conversion rate was off 5%." What we failed to say is they're still way up year- over- year and up sequentially. The reason why we wanted to say this, is we just wanted to point out that not everything's peachy out there. We wanted to acknowledge that there are segments of the consumer that are struggling more. It felt valid to have a bunch of companies then come out after us and be like, "Yeah, it's not all peachy," right? Whether it's Gap.
It's Walmart. I do think we just have to acknowledge that, and I think we are more resistant to it than ever. I'd be lying if I didn't say we don't have some segment of customers that are really struggling with gas prices. That's just we wanted to acknowledge that, but the business continues to be resilient.
Sorry.
Yeah.
Better than it was, call it 2021, 2022, simply in that your customer mix is different.
Oh, for sure. Yeah, I mean, back in 2022, what was that? What's it 70%?
Yeah, 70% of our customers in 2022 had household incomes below $100,000, and now it's the reverse of that. 70% of our customers have household incomes above $100,000. As you've improved the quality of the inventory-
Yes. Yeah, you basically had to really shift the customer demand and customer supply up in 2022, we did that, I think, very successfully. We're benefiting from that now. There's still some segment of customers that have some pressure.
30%.
Yeah.
The other thing I wanted to make sure we touched on is just the how you're balancing growth with profitability, which are now both important, right?
Yeah.
You're kind of yo-yo-ing between that, and it seems like you're kind of picking a low double-digit threshold in which you're able to still show kind of simple improvement per year. Is that kind of the right way to think about it? Is that the baseline?
No, I think we're talking like teens growth with 150 basis points of expansion and EBITDA. Think about that for the not too distant future. We're going to go that way for a while, with our goal to exceed that. I just really like setting people's expectation on where we're headed, and I think we grew 20% last year, so we have the opportunity to do better than that.
That allows you to keep investing in marketing, keep investing in process, and keep investing in engineering.
Yeah. All the leverage, I think the 150 basis points a year, a bunch of that's SG&A. I don't think it's crazy to imagine setting aside credit card fees that you're in SG&A to have SG&A be flat over the next three, four, or five years.
Wow.
You can really leverage that, and then, in our OP&T, product and technology line, there'll be some give and take because engineers will be able to do more, but they'll also use more tokens, and so I think that'll leverage, but I don't think immediately.
Improving the contribution margin being dependent on you.
Yeah. Exactly. I think we feel pretty good about what the engine looks like, and just got to execute.
Awesome. I really hope there's not, but there might be. No, there's not
What?
Agentic commerce.
Okay.
It's such a huge shift change in consumer behavior that -
Yeah.
- it's not going to be the next two or five years, it's going to be the next 10 to 20 years.
Sure.
Just how are you positioned for that? It would make sense that with all the inventory, all the SKUs you have, that if there's a crawler out there on the web, you're sort of uniquely positioned.
Yeah.
it's brand agnostic, so just what's the.
Yeah. Who is using agents to shop?
Yeah.
Anyone?
Good answer.
Has anybody built a proper agent, like in coworking? You've got to ask this question every year, because when we come back next year.
It'll be four.
Those numbers are going to be 4x. The year after that it's going to be half the room. Then the year after that it's going to be everyone. The answer is, agents are coming for everything. If you're not building them for your current workflows, you should be. Because there's so much that agents can do on our behalf to make our lives better. Sean's heard me tell the story recently, but I was on a plane from Salt Lake City to Oakland last Thursday, and I was on Delta's new Starlink, like fast Wi-Fi. I had four agents running in the background. The agents were working, I was working. Agents with complete work, I would pivot to them.
I probably did 25 or 30 hours of what pre-CEO James AI would've looked like a year ago. It is a frontier of productivity that I think will be extraordinary, but you need to invest in the J curve -
Yeah.
- of what does it take to build those agents. I say all of that because that's going to happen in the consumerization of these things. Our ability to build, we're already building agents on your behalf, and I think it will usher in this sort of real-time personalization. Agents out on the internet, agents looking at ThredUp inventory. Imagine if there's somebody who's literally, every time items are getting listed or returned or whatever, and they're like, "Hey. Hey, Dylan, I found some more stuff." Right? It becomes like a style advisor, it becomes a shopping assistant, and then eventually you give it permission to do stuff on your behalf.
Yeah.
I remember when my mom was like, "I don't shop on the internet. Where do they keep my credit card?" Now, it's a funny thing, but for a while, people were like, "I don't know if I store my credit card on the internet." I think that's where we are in agentic commerce, is permission for people to do stuff. That's changed.
Next five years?
Oh, within that.
Oh, okay.
Yeah.
Damn.
Oh, yeah.
I'm glad.
We didn't even talk about this during the last World Cup.
Right.
Right? If you think about it, the last time the World Cup was played, AI was not something that any of us talked about.
Sure.
The cycle, the acceleration is extraordinary.
Were you buying khakis on the airplane?
I was not buying khakis.
His agent-
Not yet. My agent was.
We're well over. Thanks for bearing with us. The breakout is Jenny A. It's in one of those rooms that overlooks the atrium.
Thank you.