My guest today is Stephen Winn. He's the Chief Operating Officer of StockX. Stephen, great to have you on.

Amazing to be here. Thanks.

Tell us about where we are here in Detroit.

We're at our verification center here. This is the first real verification center at StockX — we had a couple smaller ones before, but this is where every product that comes into Detroit to be verified is processed.

Products arrive here on a truck. Walk us through what happens.

Before that truck even pulls up, we run every single order through a machine learning model. That looks at order characteristics, seller behavior, different product risk factors, whether we've seen verification issues with products before. As the truck pulls up, we separate items into different lanes basically by experience level. As they go down those lanes, we send them through a number of different checks. One of those is RFID scanning. We look at RFID as a tell — often a counterfeiter who hasn't encoded the RFID properly will be easier to catch that way. A lot of major manufacturers like Adidas, Nike, and Jordan will embed RFID or NFC chips into their products. It can also help us detect whether a product is what we expect — sometimes we find out the code is wrong, it's the wrong size, it's the wrong product. There are a lot of innocuous issues that come up and get resolved that way too. From there it goes to a verification expert, and that's where a lot of the magic happens. They do a physical inspection of the product and can identify different issues with the box, the product itself. Sometimes they'll even smell the product — a very strong glue smell can be a tell. For certain high-value products, we take them through our innovation lab here, where we have a Lumafield CT scanner. We image the product, and that gives us better insights into how it's structured internally and whether those characteristics match a real product.

So StockX is a marketplace where people can buy and sell high-value and collectible shoes and other products. Give us a quick rundown of how the whole thing works.

The way StockX works — or worked back in 2016 when we founded it — is a seller somewhere in the world, across over 170 countries, lists an ask on an item. A buyer sees that price and matches the ask with a bid. The seller puts the item in a box and ships it to a StockX verification center. We verify the item and then ship it to the buyer wherever they are in the world. Over time we've built out different offerings — Flex Fulfillment, also known as our Express Ship offering, where we warehouse the seller's inventory at one of our verification centers, and our Verified Seller Program, where we focus heavily on verifying the seller and the quality of their inventory so we can allow them to ship directly to the buyer.

So compared to other online marketplaces, what you're really offering is the authentication step — the assurance that the customer is getting an authentic product.

Trust is absolutely critical to what we do. We verify products in all kinds of different ways. It's a defense-in-depth strategy.

These counterfeits are becoming more sophisticated. What percentage would you say are indistinguishable from the outside and actually need to go deeper?

I don't think I could give a specific percentage that are indistinguishable. One pattern I've observed from being close to this over the years is that when a new item releases, the replicas won't be as high quality, but as time goes on they improve. One thing we know as a pattern is that a lot of counterfeiters focus their efforts on what the customer can see — what a buyer can determine on their own. What the CT scanner allows us to do — and we use it on all products over $3,000 in value, except for some apparel, where you can't get great signals from a shirt — is tell us whether the supports are there, if there are air pockets, if components are bonded together in a certain way. We can get a really clear read on that using a CT scan.

Tell me about the typical StockX seller. Are these professionals managing their own inventories of collectible shoes, or individual sneakerheads who love to get new products and move the old ones to new buyers?

There's a bit of a Pareto principle involved. About 80% of the sales volume comes from roughly 20% of the sellers — a more mature base who might be selling across multiple marketplaces, selling new and used goods, using StockX as a channel for their new deadstock product. There are also plenty of consumer sellers, though they're a smaller percentage — the folks who win a raffle and want to move what they got.

How much of your authentication takes place in software versus truly physically inspecting the product?

That's a good question. What we're really trying to do is augment human process with better technology. I can't say this percentage is technology-driven and this percentage is human. It's more about thinking of the human as someone with enormous knowledge and judgment at the moment they're standing at the pod trying to determine if something looks legitimate. From there we give them training guides, knowledge guides, things like RFID scan data and CT scan data, to make them even more powerful. It's the combination of software, hardware, and the human element — that's the special sauce.

Give us a sense of the scale. How many verifications have you done?

Since 2016, our process has held up $750 million of products that haven't met our standards for one reason or another. That includes over $100 million of suspected counterfeit footwear — just to give you a sense of the scale.

What do you do with the counterfeits once you find them?

It depends. Not everything we stop is counterfeit — sometimes an accessory is missing, it's the wrong size, there's a blemish that shouldn't be there. Then of course there are clear-cut cases where we say this is a counterfeit product. In that case, we might hold and destroy it, work with law enforcement if we think we can aid their investigations, or hold it as a known replica that we can compare genuine products to. Which path we take depends on the frequency, severity, and confidence in the signal.

Do the original brands participate in helping you catch counterfeits, or do they try to keep this at arm's length? Do they tell you what the tells are?

It varies. There's a different level of openness across brands. We don't have established partnerships that we talk about publicly, but we do have different conversations with brands when we have questions.

Outside of shoes, what else do you sell on StockX?

Apparel, collectibles, electronics, watches, bags. Basically any product that meets a certain supply and demand profile can move well on StockX.

Where are you seeing counterfeiting pop up in your non-shoe markets?

It's changed over time. Right after we launched, counterfeit product was really consolidated into high-value, high-premium products, which made it easier to know what to look at and what to detect. Over time, counterfeiting has moved into more general release product. That's where the ML risk scoring and the defense-in-depth strategy really come into play, because you don't know exactly where it's going to hit you.

So this used to be something you'd see in extremely expensive shoes, and now it's mainstream accessible products getting counterfeited a lot?

Correct. Products commanding a premium will continue to see this type of attack, but if something is really popular — even if it's not a very high average order value — that product is also a risk.

We at Lumafield have bought a lot of counterfeit AirPods and scanned them on our systems. Putting myself in the shoes of the counterfeiters, I'm always amazed by the sophistication and the amount of work they put into it. We just bought ten pairs of AirPod Pros that we knew were probably counterfeits from an online marketplace and found ten different counterfeit designs. There are dozens or hundreds of factories out there coming up with their own very convincing AirPods, each with their own unique approach. At some level it's astonishing that they can make the economics work — something that retails authentically for a couple hundred dollars, they're able to sell for considerably less and put into the secondhand market.

We've used our CT scanner on AirPods as well, on different pairs of sneakers, even on boxes of trading cards to figure out if someone has tampered with them. For something like an AirPod, it's a great example of how the external characteristics can be faked. But as soon as you scan the item, you can see the speakers aren't quite where they ought to be, or wires are out of place compared to the genuine product.

Some of the counterfeit AirPods are even able to pair with an iPhone using the same channel as the real thing — which you'd think would be the last thing they'd figure out. It really means the only way to distinguish them is with a CT scan or by listening for a long time and feeling disappointed in the quality.

Right.

Do you find that counterfeits are, for the most part, truly lower quality? Would the wearer be disappointed in the experience, or are many of them very similar?

There are buyers out there who are comfortable knowingly buying counterfeit products — it looks close enough, the price is lower, they're okay with that. Our customers want us to be verifying the product. That's part of our value proposition, and that's what we stand for.

Let's go deeper into the manual verification process. Right through the door next to us is a big room where you have dozens of trained verifiers taking shoes out, looking at the stitching, smelling them, examining the packaging. Tell us about the training process. How do you become a verifier?

You don't get a college degree in verifying sneakers. A lot of our verification experts have been resellers for years, worked at a Foot Locker or a Finish Line, run their own consignment shop. They have a real passion for the products they're verifying every day. When new products release, we get our hands on them, CT scan them, RFID scan them, and do everything we can to understand what a real product looks like. From that we build knowledge guides where verification experts can study, take quizzes, and advance through different verification levels. On the screen in front of them at the station, we give them additional detail — accessories that should be in the box, a 360 view of the sneaker they can spin around to understand what they're looking at.

How many different types of verification have you added to the process over time? Do you add new steps regularly?

We do add new steps. When we're thinking about adding a new technology or a new step, we really look at three things. The first — and maybe this seems obvious — is does it actually do anything? There are a lot of highly experimental technologies out there, software and hardware, and if they can't reliably detect a counterfeit, we don't want to add that step. The second is we don't want additional steps to balloon the cost, because we'd have to pass that cost somewhere, and our buyers don't want it to take longer or cost more. We don't want to double or triple the time spent at a verification station. And finally, we want to make sure these things can stand the test of time. RFID, for example, is a relatively durable signal for us — it's at least an indicator that a product could be legitimate. And the CT scanner from Lumafield is a great example of that: being able to see the internal aspects of a product is critical, and it's one of the hardest things to fake — and one of the last things most counterfeiters seek to fake.

Do you ever notice evidence that counterfeiters are learning from you? That they watch what you're catching and adjust accordingly?

I'm flattered if they're learning from us. What I'd say is counterfeiters are committed, and over time, as new products release or as a particular product sustains in popularity, they simply continue to iterate.

What's the most surprising thing you've seen in a CT scan of a counterfeit?

One of the coolest by far was we CT scanned a pair of Wizard of Oz Dunks — the Ruby Reds golden tickets — and what we identified was that under the tongue, it said the shoe was numbered out of 85. We knew they were limited, but we didn't realize that until we CT scanned it.

So you actually figured out how many of these were made by doing that?

That's right.

That's incredible. Do you ever go back to the shoe brands and say, hey, I found this fun secret you put in?

We don't go back to them directly, but when we put that social post out about the Ruby Reds, I think they noticed. And when we publish things like our Big Facts reports — indicating what's trending from a sales and counterfeit perspective — we know the brands are picking up that information.

You were employee number six here at StockX, right? Joined way back at the beginning. How have you seen the technology stack evolve over the time you've been here?

Night and day. In 2016, we took an e-commerce platform and said we need to make this into a stock market. We took the concepts of orders and shipments and added bids, asks, a matcher, inbound shipments. The whole process was very straightforward. What's changed is this is no longer a Detroit-based business — it's a global business operating in 170 countries across 5,000 trade lanes with over 71 million lifetime sales. To support that scale, almost every line of code I wrote has been replaced.

That's a good outcome if you're an engineer — to see the company so successful that everything gets rolled over.

I love it. It's a compliment for sure.

You're a Michigan native, right?

Actually, I was born in Akron, Ohio.

Oh, well — we won't talk about Ohio. What does it mean to you to be able to build a company in Detroit?

Detroit's a blast. We've had access to great talent and capital here. It's not like Silicon Valley where there are a million startups. StockX stands out here in Southeast Michigan, and that helps with retention and attracting talent. This has been an amazing place to call home for both me and for StockX.

And everyone who works here that I talk to is a big sneakerhead.

One of our goals at StockX is to fuel people's passions for the products they love, and this company through and through has team members who are pretty passionate about what we do.

You've just announced you're going to launch a product this summer called Live Shopping. Can you tell us what that is?

Live Shopping is the ability for buyers and sellers to come together, have direct interactions, with sellers on screen participating in live auctions and giveaways. It's an indicator of where customers are going — they want to be part of an interactive experience, part of a community. That's a major change for StockX. But what's always been core to us is wanting to participate in culture, in what's cool and what's happening. Live Shopping is there. Something else that's always been central to StockX is trust. On day one of Live Shopping, sellers will ship directly to buyers, but shortly after we want to add optional verification. Buyers can say, I want to use StockX's verification network and technology solutions to make sure the goods I'm getting are real. We're going to apply our policies — things like the StockX Buyer Promise — if there are issues, to make sure buyers can be made whole. And of course we're going to be moderating content, making sure sellers behave appropriately on camera, ship on time, all the things that go into running a marketplace.

Even before Live Shopping, did you have sellers who had their own followings on the platform — where people would say, I like to buy from this particular seller?

Today the platform is completely anonymous. This is an exciting new model we're adding. What's important to know is that for people who love the core verified model, that's not going away. It's just the ability to interact directly with a seller that'll be new.

And to have that live, continuous experience. This is the grand transition happening in all of commerce right now — everything has to be an experience.

There's just something to it.

Excellent. I'm going to wrap up by asking you the question we ask every guest. This podcast is called Go/No-Go. Can you tell us about a go/no-go moment you've faced at some point?

The very first one that comes to mind is launch day for StockX. Go or no-go? And right after saying go, pushing out the press release — suddenly we're committed to whatever happens next, all the problems we hadn't anticipated. But to be specific: around 2018 and 2019 we were expanding globally, adding verification centers in London and the Netherlands. One thing we wanted to think through was how to improve the experience for customers all over the world, and there were competing points of view on how to do it. One option was a StockX Europe marketplace and a StockX US marketplace with local supply and demand in each. What we decided was that one of the great powers of this platform is having one single global marketplace. So we invented — and actually patented — what we call the all-in ask. What happens on the StockX platform today is that if a customer in Germany is looking at a listing from a US seller, they see that price marked up to include import costs and additional shipping. That flattens out the marketplace so the buyer in Germany still has access to all the supply in the world, but with the US product marked up a little, German sellers with the product can create more of a local marketplace for that buyer. It's a complex piece of technology — a pricing algorithm with all kinds of tax implications and regulatory considerations to work through — but I think it's a real differentiator and one of the most hotly debated implementations we've done.

Is there a directionality to your global trade? Do Germans buy more shoes from Americans than Americans from Germans, and do you pick up patterns like that?

We do see what's popular in different areas and different countries. We see different import and export behaviors. The United States is the biggest market, but there's a really powerful European market as well.

I've heard that in the used car market there's a flow of cars from the Southwest to the Northeast because cars in the Southwest don't drive on salted roads. I'm wondering if there's something like that with shoes — people preferring to buy from certain regions.

Buyers don't actually pick where the item is coming from. But there are definitely differences in preferences. One thing I noticed years ago is that in Germany you might see a preference for blacks, blues, and whites — flatter color patterns. Then in LA maybe you have brighter colors. We actually just did an analysis on top sellers across all the different US states.

So you can read the character of places in terms of their preferences for shoes.

That's right.

I love it. Stephen Winn, thank you so much for coming on. I encourage everyone to check out StockX — it's an incredible platform and a really amazing operation you have here.

Thank you so much. It's been a blast.

Go/No-Go is brought to you by Lumafield, which was founded to upgrade manufacturing. You can learn more at lumafield.com. Go/No-Go is produced by Austin Carder and edited by Brian Tran, with additional assistance from Eric Petralia.