Dupe.com CEO Bobby Ghoshal: 18M shoppers, approaching $100M GMV, and going #1 in the App Store

company doing welcome to the show.

What's up? Thank you. We made it guys. We made it.

We made it finally. Uh great great setup here. You got you're just daily driving that for Zoom.

That's it. Yeah. I mean,

fantastic.

It's comfortable. It's cozy.

Introduce the company. Uh I remember when uh you launched, I remember seeing a whole bunch of posts go viral, but uh how are you describing the company today? Yeah. So, so Dupe is used by millions of people to find the products that they want or similar products for less.

Uh, it's a search engine. It's a generative marketplace. I could throw a whole bunch of buzzwords. It's AIdriven this and that, but ultimately it's a shopping companion. It's a shopping app that helps give people a confidence pill, a little bit of a boost to know that what they're buying is at the best price and is the right product.

Mhm. Agentic Commerce, OpenAI is talking a big game about it. Uh, is that a headwind for you?

Yeah. You know, it's first of all, uh, nothing's set in stone yet. Don't don't count us out. We're also talking to them. Sure.

Um, we're we're building we're building an app, uh, that would be native to ChatGpt. Okay. uh you know it's like the the thing I like to say about chat GBT Google all these tools is that

they can do everything

and that is the problem.

Um it's like you can if you're a chef you can use a Swiss Army knife to to cut a tomato but why would you? I think verticalized AI tools are are very powerful because they are single purpose uh and they solve a problem in a in a in a in a more precise way.

Yeah. I mean I it's odd. I have like no doubt that there's people out there that are making tens of millions of dollars on uh Studio Gibli apps, right? Or, you know, uh different animation apps, different uh different sub apps that maybe could be done in chat GPT if you know the right prompt, but because it's lower tier on like the prompt ladder, you kind of got to know to go there to get it. Uh it's not coaching you through it. it's not actually giving you uh all the different uh harnesses and and wrapping uh that you need to do it.

How hard has it been to actually build the ex like the product experience? because I'm assuming you're effectively like browsing the web on behalf of, you know, if I upload a a product or I paste in a link and then you guys are browsing the web or you already have and you're and you're uh sorting products to find uh products that look the same and or maybe uh from a different manufacturer or the same uh end manufacturer, different brand. But uh how hard has that product experience been? because I've been continuously disappointed by uh computer use and just web browsing from a lot of different LLMs at this point.

When you say computer use and you're talking about things like uh the browser is doing the browsing on your behalf basically. Yeah.

Yeah. Go out and find this thing. The the example that he gives is

I use it for cars. That's my

find me a GT3 RS. It's a rare car. It's a specific. It's a great car. Uh, but it's not it's not uh like find me someone who sells Diet Coke. Like Diet Coke is a commodity. It's sold everywhere. A GT3 RS uh there might not be one for sale right now. Like it's possible that there's just none for sale. And so

Well, yeah. GT3S quite a bit more common, but there are certain cars you look for

and it'll say I found you this one. And it's like it sold two years ago.

Yeah. And so and so on that eval of like go hunt around on the internet, find this particular thing that's not particularly funible, not particularly commoditized. Uh at least agent mode, these agent these a uh these agentic commerce products have not been uh up to Jord's standards at least.

I take it you haven't bought the the GT3 RS yet. Like

not yet.

Not yet. And the only reason is uh just the AI is not good enough.

Yeah. Yeah. Of course. Of course. That's the only reason. Come on, guys. Uh so so so first of all I think it's I I too have been fairly disappointed with with these like browsers that that browse on your behalf. That is ultimately not what we're trying to do with dupe. AI has some flaws in like the shopping universe and maybe that you know it it speaks to some of the pain points you're finding with you know finding your perfect car which is they don't have uh live inventory, they don't have live price, they don't have product cataloges, they don't have customer preferences, they don't have customer behavior. There's a lot of data that is missing in the shopping journey. So ultimately what you're left with is the best guess. So to answer your question, how hard has it been? It's hard because the way you do it, the way you do it right, is you got to do it the old school way, all right, in the shopping world, you got to go press palms, kiss babies, hug people, take them out to, you know, to a drink, meet them at conferences. You got to know their names, you got to know their kids' names. It's like a lot of trust building in person. It's not as simple as just kind of plugging it in because when you build that trust, they then trust you with their product cataloges and they give you a commission for sales because you know you turn out to be a partner that drives a lot of sales for them. So it has been it has been pretty hard but the the tech that we have is product ingestion from these partnerships but also live scraping. So it's you know in a sense it is browsing

and it's scraping this data for you. We have vectorized millions of products or tens of millions of products at this point. So we can do kind of a live assessment of what products look like each other, what products should go together, stuff like this. Um, ultimately the way to solve it, you can't, it's AI is not sort of a panacea. You have to pick your battles. So Dupe, for example, for the last year and a half has been focused solely on furniture. Why furniture? Well, turns out unlike the GT3 RS, a lot of furniture is made in the same few factories, sold back to brands under different names. the brands, then upload them to their website under different names, different prices, but it's the same furniture. And I know this because I got scammed uh with buying furniture online, and it really pissed me off because I found the same product I wanted for about 40% less. Exact same same spec, same color, everything, exactly the same. And so that ultimately is like the lore of what spawn dupe. Uh and I've seen these factories in person. I've visited China. How do you think about uh incorporating human feedback? That feels some that feels like something that would be a really valuable resource. I was thinking about like on Instagram you see a lot of ads for a bunch of different stuff. Some of the stuff is is like it really grabs your attention. People just buy it, but then it's really low quality. And if you could actually have some sort of, you know, okay, feedback. Hey, we we we you know, we we shipped you this dupe, this product. um we think we got you the lowest price, but like tell us how the product actually performed because that better informs our our our our decision-m uh are do you think that'll be like a compounding advantage? Do you think that's valuable over the long term?

It's so it's so uh it's so valuable.

Uh and we don't even need to ask the customer. First of all, we have access to all these reviews. We also have access to the checkout data and we have access to the return data because if if someone returns something, we don't get paid. Yeah.

All right. So these brands pay us four months after the sales happens, three months after the sales happen, after that return window is set. So we know which products get return the most. So absolutely that's a compounding effect.

Yeah, that makes a ton of sense, Jordy.

Uh you get a lot of hate mail

from from certain brands that uh that that uh kind of realize like you're providing a service that consumers love and then certain brands out there that that are sourcing products and setting whatever price they feel like is relevant. And then you're actually in

blowing up my spot. I had an I had 99% margin on this couch.

Yeah, we basically like killed shop drop shippers like for sure. Like the drop shipping industry is a is a total scam. Um and and they hate us. But ultimately our brands the part our partners love us. We drive millions of dollars in sales for them every every single month.

And so why is that? Well, let's say like let's take a partner like Walmart. So some people think Walmart has thousands of products. They don't. They have millions of products, right? Millions of products. They're not always in store because a lot of those products are drop shipped. So, what we do is, you know, let's say you guys are looking for, we'll take a piece of furniture, for example, that chair right there. I mean, that's that's not being sold at Walmart, but let's say you want something like that, and you upload an image to dupe, or you tell Dupe like, "Hey, that's a Herman Miller. Find me something like that for less." Walmart has 15 versions of that chair. You don't know that. You would probably never even think to look at Walmart for that chair, but Walmart has a great version of that chair and they love us because we merchandise their products at that exact high intent moment where the customer knows what they want. They just want a great deal on it.

We surface the partners up. Now, there are some brands, for example, we were sued uh last year and

Yeah. Yeah. We we we were sued. I'm not going to name the brand. It's very easy to find the brand. They are the largest largest furniture brand now. the lawsuit essentially like it tried to kill our company. It was it was like a pretty tough uh a pretty tough lawsuit that we that we had to fight. We settled it uh to both parties content. Like it it it's fine now. Um but

what was the what was the claim?

The there are a couple claims. So copyright infringement. They were pissed that we were using or letting our customers use images from their website to find similar products. Uh they were they were pissed that like they were part of we went super viral when we came out. I did all these like post-it note ad campaigns and stuff that went super they were pissed that they were in some of those some of those uh

social kind of Tik Tok videos. They weren't even ads. They were literally just social content talking about our our website. They were they were really really upset about that and ultimately they were pissed that uh we were

unfairly competing essentially. Uh, I obviously, you know, we we we have a a public statement on how we felt about all of these claims. Um, and so they basically said we were falsely advertising, you know, that they were white labeling stuff that we that we were unfairly competing and also infringing on their copyright. So ultimately, you know, cold water under the bridge. We're we're here. Dupe is alive. It's well, it's working exactly as we wanted to and it's crushing. And so there are some of those partners, but those are very few and far between. And it tends to be like the almost the monopoly type partners, you know, uh that that are that are upset. Everyone else Everyone [snorts] else and most of the Fortune 500 that we work, they are thrilled that Dupe exists.

That makes sense. How'd you get the domain?

It's absolutely fire.

We So the company was called Carrot before it was called Dupe. We uh had three months of runway before we pivoted the company to Dupe. We worked with Nikita Beer on it. Uh he was a total goat. Helped us find

You spent your last your last 20 grand on intro.co. [laughter]

Well, we had we had more than 20 grand. We had about three months of runway. But yes, we definitely spent more than 20.

Speaking of numbers, yeah,

just with Nikita. Probably the best money probably the best money I've ever spent. Saved our company. Let's be completely honest.

That's actually crazy. I love it.

It it it like to total bananas moment for us. But um the the domain was a very last minute decision. We we knew we needed a great name to stand out in the in the kind of shopping space. I did a bunch of Google research on keywords. One of the words was dupe and I saw dupe was on on Google Trends 100 rated Google Trend keyword and growing.

Mhm.

And I saw that no one used no one was using dupe.com. We reached out to a broker that was sitting on that name. They wanted just the most insane price. We then negotiated

a a far lower price for dupes.com

went back to the owner of dupe.com and said if you don't sell this domain to us in like 5 minutes with this price which I thought was a fair price. It was all of our money and we gave them all of our money. Wow.

And you know that we're going to pull the trigger on dupes.com and they sold it to us like on the spot. Basically said okay fine go for it. Smart.

That's great.

That's how we that's how we negotiated it. [laughter]

I love that. That's great. Thank you. That's a good story.

Very cool. Yeah, that that was like a bet the like a literal bet the company moment. And then Nikita did the most mench thing ever. He put out a tweet about us and instantly blew us up.

It was crazy. I remember that tweet. He he he framed it very very well. And then you can imagine this just works on social.

What other categories do you think are most ripe other than furniture?

Yeah. So we do fashion now. Handbag, shoes,

clothes. Uh really big one that we don't do very well yet because AI doesn't have a sense of smell. You know, it's looking at kind of vision. Perfumes, perfumes are huge in the

Yeah, it's basically I look at as anything without a logo. So, like furniture like there's iconic shapes and silhouettes and and uh you know, materials that get used along that, but ultimately people don't have, you know, you're not like you don't have a couch that has like a restoration hardware big logo printed on it. The challenge with fashion is that people want like the association with a specific like brand and that can come through a logo or a or a silhouette that actually can't be uh implemented without getting um you know uh kind of

you're right. You're right. Except we're going through this like bizarre moment in shopping where brandless is a thing and not showing logos is a thing. like this this kind of quiet luxury movement like people people want something to feel good and that butresses

like even the brands that you love

they're white labeling guys like they are they are none of this stuff is like custom made specifically for a single brand maybe if you're going into the couture world of course that stuff is is custom made but ultimately you know when someone sees Skiim's leggings like yeah they they want the look

people shop people shop the like that's what they're shopping ultimately and it's it's less about like you know people often misconrue what dupe.com is like we're not trying to sell you replicas we actually ban replicas across the board

and all those sites

we are trying to get you that look that garment you know that furniture for less and and most people shop for the look now especially in this market.

Yeah, that makes sense. Uh what uh how do you are do you think you'll have enough leverage at some point to get to get uh the underlying brands to pay you at the time of sale?

Is that even

It's actually very important. I'll tell you why it's important. If we were growing 100% month over month and we weren't seeing our cash flow for 3 months, it stifles our marketing budget. We would have to then either raise credit or go raise a massive round just to pay down the marketing. This is actually like the biggest risk uh that I that I think about. So one of the things we're partnering with the largest payment network to do uh Aentic checkout global checkout um and you know we're not announcing the the name yet but we are working through that tech and the idea being

you know most of these brands are really upset with a gentic checkout because they they don't get the customer information but we do it in a way where the checkout would happen on dupe but the brand retains

the customer information and the customer relationship and the the only reason we want to do that is so that we can you know go from net 90 payments to net zero net one you know instant payment uh back to to our company.

Uh so yeah this is this is very top of mind for us and it's it's definitely a risk that I think a lot about.

How how big uh how big do you want do you want dupe to get?

I think dupe uh we we often talk about a road to a billion shoppers. The way that we're going to do this is we have an MCP. Uh so we any LLM can plug us in. you know, the the largest LLMs and the longtail LLMs, they can plug us in. Ultimately, we want to be the canonical answer engine for how do I get this for less or find me something similar to this. We want to be that answer engine for these websites. The beauty is because they don't have access to the behavioral data, the shopping data, the product cataloges, the product pricing. The moment they plug us in, it's basically zero effort. It's just, you know, a few lines of code. the moment they plug us in, any click going out from our answers that gets monetized, we split that commission back with the LLM. So, it's like instantly monetizing their user base in an entirely uh entirely new way.

Um, so that's that's something that we're very excited about. When you think about how like the the growth potential of that and the tailwinds for any LLM to really consider partnering with us uh for our links and for our expertise and the ability to monetize instantly, I think we can get to hundreds of millions of shoppers on our infra. They might not be on dupe.com, but they'll be

you're still monetizing them.

Yeah,

that's right. That's right.

That's very exciting.

That's right. And in the last 18 months, we, you know, we're right around sub 20 million users. Uh, between Let's go. Oh my god. Do I get the con? Let's

go. Not even sure if you're going to give us a number,

but we got

Yeah. Yeah. No, no, no, no, no. So, so I'll give you a few numbers. I mean, we're we're rounding a corner. We're going to come up on 100 million in GMV sold through dupe.com.

18 million shoppers. Fantastic. uh between one and two monthly uh one and two million monthly shoppers which we just launched our app last week. Went number one in the app store by the way super stoked about. Please everyone listening go download dupe dup look for it in the app store. Leave us leave us a review and rating. Got to put in the plug. Uh went number one in the app store and of course we had a lot of fun with that. We've been putting a lot of fun content out.

How did how did you go number one? Was it content or or some type of like viral incentive? We have a lot of we have a lot of users like these people they want our app like they you know they use our website to shop and have used our website to shop for for many months. Um they want they wanted our app and you know the way the app store works is on a 24-hour clock. So the the number of installs you get in a 24-hour period if it's more than any other app you rock it up to the top. I didn't think it would happen on day one. I actually had a whole I was convinced that we were going to go number one. First of all I called this shot in August on Twitter. I said we are going to un unseat dethrone chat GBT when we launch our app in November. So we we shot our shot and we we kind of netted it but it happened within a couple hours of us going live. I didn't expect that. I expected it to be about 48 hours. So it happened very very quickly.

Wow.

Very cool.

We still have a lot we still have a lot of work left. I mean you know obviously we're at the beginning of our journey.

Call your shot again. 100 million in GMV. When are you going to uh or when when are you going to hit it?

We're a couple months away from that.

Okay. Well, we got the gong ready for you.

Amazing.

You're in LA, right?

Uh New York.

Oh, okay. Okay.

Well, well,

we we should we we should be able to get to that. I mean, this this shopping season. Let's see how Black Friday goes. Make sure none of our systems go down. But that that would be pretty dope.

Yeah. Really, this is this is your Super Bowl.

Very cool. We'll talk to you soon.

Great hanging, dude. Congrats on all the progress. Congratulations. We'll talk to you soon. Uh, let me tell you about wander.com. Book a wonder with inspiring views, hotel, great amenities, dreamy beds, top tier cleaning, 247 concier service. It's a vacation home but better. And that's our show for today. That's this week um in TVPN. Thank you for tuning in. Thank you for listening. Thanks for hanging out in the chat if you've been chatting this week. Uh we love you and we'll see you on Monday.

Have a fantastic weekend.

Have a great weekend. Goodbye.