Retailgentic, presented by the ReFiBuy Media Network and hosted by Scot Wingo, is where Agentic Commerce meets retail innovation. We help retailers and brands prepare for the future of agent-driven shopping with actionable news, expert insights, dxis, forward-looking predictions, and practical guidance on Agentic Commerce Optimization (ACO).
Welcome to the Retailgentic podcast where we explore the intersection of retail, ecommerce, and AI agents, which we call Retailgentic. Our mission is simple. We are shining a bright light on the new agentic AI shopping megatrend which will disrupt the trillion dollar digital retail market. The stakes are high. Google, OpenAI, Microsoft, Perplexity, Amazon, Apple, Meta, and Anthropic are all fighting to be the Consumer AI Shopping Agent winner.
Speaker 1:There are gonna be some tough decisions for online brands and retailers to make. What happens to checkout? How should payments work? Should you continue to invest in retail media networks? What does the world look like when over half of your traffic is not from humans but from AI agents?
Speaker 1:These existential questions are the tip of the iceberg. On this podcast, we're talking to industry experts, analysts, and influencers to get answers to these questions, identify trends, and best practices. This is Retailgentic, your podcast partner in navigating our AI shopping agent future. And now here's your host, Wingo.
Speaker 2:Welcome to this episode of Retailgentic. Today, we have an international podcast. Today, I'm joined live from India with Naveen Tewari. He is Founder and CEO of InMobi and Glance. Naveen, how are you?
Speaker 3:Very good. Scot, how are you doing?
Speaker 2:I'm doing great. Give us a what's the situation in India these days? Is this monsoon season or are we is that coming up later? I think it's kind of later in the year, right? Or is it spring?
Speaker 3:We're we're in the seasons. The monsoon is late this year. So we're going to head into towards the winters in the next month or so. But yeah, otherwise things are very good on this side of the ocean.
Speaker 2:Awesome. Well, appreciate you taking time to come on the pod. Very excited. You guys can I just just all the news about Agent Commerce and I swear every week Glance has a big new announcement? So I've been really excited to and I've been using using the software and the app and playing with it.
Speaker 2:It's amazing. So wanted to do this for a long time and I'm glad that we're able to make it happen.
Speaker 3:Same here, man.
Speaker 2:Let's start with your background. How did you, you know, how long have you been in technology and what led you to be interested in Agent to Commerce?
Speaker 3:I started my first company in Mobi in 2008. So we just finished, what do I say, eighteen, nineteen years. And that was in advertising. InMobi is doing amazing. It continues.
Speaker 3:And then about few years ago, built Glance, which is our foray into agentic commerce. And our belief is that, you know, and, you know, who am I saying this to? We believe, you know, in the next few years, half the population on the planet would probably be using an agent to essentially shop. And so we are at it and we're building some of the most advanced agents for shopping out there and pretty kicked about that. So I run two companies, both InMobi and Glance.
Speaker 3:And, you know, I also therefore have a very strong take on where advertising is headed. And I believe every ad in the world will be an agent, you know, eventually, you know, or a store, Agentic store. And that's the future both for advertising and for commerce. And I just feel very kicked about being at the cutting edge of both of these things.
Speaker 2:Got it. And in Mobi's side, is that a ad tech business or is it an agency or is it kind of both or is it Oh, none of
Speaker 3:the We are an advertising technology business. We we have a, you know, we are a we are a suite of programmatic platforms across the mobile SSP, DSP. We are a very large we're on a pretty large exchange. So, yeah, so that's the... That's on the on the InMobi side.
Speaker 2:Do you get involved in the retail media network kind of part of advertising that a lot of my friends are excited about?
Speaker 3:Oh, yeah. Yeah. We we are pretty excited about the Retail Media Network. It is something that we, you know, we go after both from the InMobi side and from the Glance side. And therefore, it's something which is where we are we are headed into that path.
Speaker 3:It was not the frontier that we went after as purely as InMobi. But because of what we are doing with Glance, we are actually starting to look at Retail Media differently. And yeah, we'll see how that evolves because I think the retail media networks have to essentially be far more, you know, somewhat different than what they are today.
Speaker 2:The advertising business, I don't have to tell you, for the audience is just brutal. I could never build a company in that space. It's just there's a thousand competitors. You've got the two hulks of Facebook and Google. They're kind of with a duopoly of some sort.
Speaker 2:And it always surprises me Amazon kind of came out of nowhere and kind of added themselves to the equation for a triopoly, I guess we call it. Yeah. How did you, you know, so just surviving if, you know, not only thriving in that space is amazing. What is it about what's InMobi's edge? Like, what do you guys do different that has allowed you to survive that long?
Speaker 2:Is it a special tool for a merchant? Is it a certain category? Give me the angle that has got you this far?
Speaker 3:So on the InMobi side, I think there are a few things that we do really well. One is, you know, we ensure that we have our, you know, in order to build out our network, our publishing relationships, we ensured that we had our own SDK distributed across the apps. So, therefore we have control and dataset access on the environments in which we essentially render advertising. So, that's one part of it. The second part of it is what I would say, our ability to essentially run neural net algorithms, we are really good at that piece.
Speaker 3:And we are able to essentially deliver return on investment for a return on ad spends for advertisers because of the quality of the neural nets that we run. And the third piece, which is, you know, kind of really important is the fact that we are able to be on both sides of the equation, you know, on the supply side and on the demand side. So therefore we are able to compute enough data sets to be able to deliver that value. And I think where a lot of, as you rightly said, where a lot of advertising technology companies get started but have a hard time scaling, which is the issue with any advertising technology is to be able to harness enough data sets to essentially deliver that value at scale across hundreds of millions of consumers on a real time basis. And I think that is a technological challenge at scale and we kind of do a great job at it.
Speaker 3:Now add to that the fact that we have our first party dataset and our first party consumer property really gives us a huge advantage.
Speaker 2:I have a thousand questions. I'm a fellow entrepreneur, so this is always a fun part of the conversation. Sometimes as an entrepreneur, you make a bet on something you're So right or you bet on neural networks. You call it a neural network, which means you were early than 2022, which was in hindsight was, you made the right bet. Did that feel like kind of being out there on the edge of, right on the hairy edge?
Speaker 2:I'm sure you were way, a lot of people were not doing that, than having it pay off so much with the evolution of of... You know, that ended up being the thing. And and you were... Therefore, you were early, which gives you a bit of an advantage.
Speaker 3:Yeah. Look. As an entrepreneur, you you have to play a few shots to win. And and and you gotta choose those shots that that gets you to win. If you get them wrong, you look look really bad because then you're like, you know, oh god.
Speaker 3:How could you be so stupid? It was it was It was written on the wall that you couldn't do this. So there were a few decisions, including the decision to, you know, bet on neural nets, the decision to essentially bet on the fact that in this AI world, the value would actually be attributed to the application layer was a big bet we made back in 2022 and 2023. We made that huge bet then. We went off to saying you've got to essentially be in this vertical as against trying to build anything horizontal.
Speaker 3:The world was clearly going after the frontier models and we said, hey, we should actually be very, very focused on building for shopping and kind of going deep into that. So, these have we have made multiple big bets. Getting some of them wrong early could have really impacted us. But getting most of them right, as had the reverse gain to say, oh, wait a minute, like, you know, these guys are onto something really big. Yes, we are.
Speaker 3:And, you know, those bets have paid off and we feel pretty good about those big bets that we made at that point of time.
Speaker 2:And if you make a bad bet, you got to at least pinch it off, learn from it and move on. You can't just soul search forever. The consumer piece you're talking about that is Glance, right?
Speaker 3:Yep. That's Glance. Yeah. Okay.
Speaker 2:Yeah. So it's interesting because Amazon surprised everyone in the ad space because no one thought you could even get up to the scale or even approach the scale of the two big guys of Google and Meta. And then they've done it relatively fast. And what was their tailwind was the first party data and that simultaneously with the changes in privacy. So I think your strategy is going to be very good.
Speaker 2:And I think my theory is that's what's driving a lot of people say to me, Google's tried this something marketplaces and now it's e commerce for a long time and failed many times and not really done it. I think they're especially motivated because I think they've seen what Amazon's done and they want first party data too. Yes. I think that's what's driving Meta. I just listened to the we just had the big Meta earnings.
Speaker 2:I'm sure you track this as closely And as I Mark Zuckerberg's practically giddy about this muse opportunity that he has. And I think the reason they're giddy, two advertising companies, is they have caught on to something you caught on a while ago as well. And they're kind of realizing this is going to give us the holy grail of the ad holy trinity, which is first party data.
Speaker 3:You're bang on, man. You're bang on, Scot. The idea that the most valuable data set is the first party data set of, you know, consumer transactions and you know, why they are thinking about those transactions is something that, you know, necessarily Google didn't have, but they of course had the intent data. Facebook apart from the social leader didn't have it. So, they're both going after it.
Speaker 3:Amazon showed the value of that. That's become very visible. We bet ourselves on that exact thing by building out Glance and saying, hey, if I in the world of agents, can I actually understand not just the transaction but also the context behind that transaction and build out my context graph? I fundamentally believe that's more powerful than the first party transaction datasets and that's what we are building which is true for any agentic world that look, I build a context commerce graph? And our view is that that is the true asset we're trying to build, connect that into our advertising infrastructure and see it convert value for our merchants, our advertisers and everybody.
Speaker 2:It's interesting. I was reading the Wall Street reports after the meta earnings, and Mark Mahaney, one of my favorites, he he he called this. He said, I'm gonna brand this thing that we... That's out there. He he he branded what you're calling agentic intent intent.
Speaker 2:I like that.
Speaker 3:Yeah. Yeah.
Speaker 2:You know, kinda like there's first party data and then it's kinda like first party data on steroids, is capturing not only the the little the little crumbs we got in the the old era, but now we're getting like the whole sandwich.
Speaker 3:Yeah, exactly. Exactly. The why behind the what is really powerful.
Speaker 2:Yeah. So when did you start Glance? And it seems like kind of two things converge. There's kind of like, wow, it should be sure would be nice to have some first party data to really enhance our whole system. And then disbelief in Agent Commerce.
Speaker 2:When did you guys decide to pull the trigger and jump into this?
Speaker 3:Yeah. You said it really well. Look, the idea was to say, hey, what could what could we be building that has a agentic that has a first party data set? And so, you know, we were the first version of Glance was actually trying to do, you know, a lot of personalized content. We were building, you know, to begin with, with Glance, we're building the personalized content platform to realize as we saw the generative AI kick on and we realized that, wait a second, there is this whole new opportunity of thinking about, you know, how in the world of generative AI shopping would change and what could a potentially agentic shopping experience look like?
Speaker 3:And so, we set out to say, you know, let's change the way the world shops. And our belief was if AI is going to disrupt and change everything, one of the big elements it's going to change is shopping. And so, we took a take on, you know, Agentic shopping. And our take was not driven from just automating, you know, shopping workflows that are standardized, like ticket buying, buying milk and all of those things. That's the easy part of the agent if you ask me.
Speaker 3:We actually went to the other extreme and we said, Hey, the true value add for an agent needs to be on the inspiration side. And can I actually fundamentally add intelligence on the part where inspiration has to be driven because that's the really hard part of shopping? And so, what we did with Glance was to essentially think about how do I do a few things? One, add intelligence to categories like fashion, accessories, to travel, to home furnishing, pets, categories that require there is no standardised way of making a decision. You make a decision through a certain amount of logic, but a lot to do with emotion.
Speaker 3:And therefore, identifying that is what we've really focused on to begin with. And therefore, the way we led it is not through text. And we said text is okay, but you know, I don't know when was the last time I went and bought like something basis text. I got to see stuff. Of course, I don't need to go see the milk, but for most other things, I got to see that.
Speaker 3:And so, we went out there and created a whole the whole of glass is actually based on images. What we do is that we try to essentially use the agent to generate the potential output of a product that a consumer should consume. And not just the product, but the product in the context of the user and in the context of the environment. So for example, if I take fashion, which is one of the most complex categories in the world, and you want to buy a jacket, we can actually fit the agent first figures out basis, who you are, how you look, where you're going to use the jacket to kind of shortlist few few jackets for you. But in order for you to make this decision, that's not good enough.
Speaker 3:It's going to actually show the jacket on you. And when it puts the jacket on you, you you go like, oh, now I know. Like, that's a great decision because it just simplified my decision making process. And then once the selection is done, it goes to stage two of it, which is to say, Okay, Lon, let me just help you buy it. Because now I could go in there and figure out what is the best place to buy this from?
Speaker 3:What are the best payment mechanisms to use in order for you to not pay the least amount of money for that product? That to me is how we thought about agentic shopping. And so therefore, we went after the categories that were the hardest categories out there. The beauty of that has been the amount of context we learn about consumers is insane because consumers actually, when they are buying things that have an emotional decision criteria, a lot about themselves because they want the agent to really understand who they are. Because if I'm buying like, something that's not emotional, I will be very logical in my price or attributes and whatnot.
Speaker 3:But here I'm actually expressing why I need something different. So therefore, the amount of context we understand about consumers is amazing and therefore the agent moves up the ladder very, very rapidly. So that's what we do. We fundamentally believe we are trying to change the way world shops and we led off from inspiration. Lead through images and not text.
Speaker 3:And of course, we then do all the right things that are required to complete the buy.
Speaker 2:That's super interesting because, you know, the name of our company is, short for research find to buy. And, a set of... The early set of Agent at Commerce shopping is gonna be very tactical. It's kinda like my my other podcast buddy, Jason Goldberg, calls it spearfishing. So Yeah.
Speaker 2:I need some milk and a dog treat, and that, you know, I'm not really looking for inspiration if I'm buying those two things. I'm... I I already know the milk I like, and I'm just gonna reorder the dog treat.
Speaker 3:So so when we thought about Glance, we knew the efficiency part would be automated by everybody. Like, that's just
Speaker 2:gonna be the whole field is. Right? So so if we look at answer engines, they've all kinda gone that direction. If we look at the new raft of personal shopping agents like Metas Muse and and so forth, Claude just came out with a whole another one. So there's...
Speaker 2:It's like every company now is just checking the box. Google's the only one that doesn't have one. And then even the ones that are hosted on the retailer site, the the Rufus slash AFS and Sparky, very, very tactical. Just help me find the thing. You know, I I know what I'm looking for.
Speaker 2:Get it and put it in the cart and buy it. Yeah. And then, you know, the only other company that kind of lives in the inspiration bucket is Pinterest and they just can't seem to get, they've had like a lot of leadership changes and they can't really, they've never been able to, even in my era of ChannelAdvisor, they've never been able to connect the visual part to actually like buying something. And it's like been very hard due to legacy and execution reasons. So I think there's a wide open lane there, which is great.
Speaker 3:Yep. Yeah, exactly. And I think the possibilities, by the way, are just crazy. Like, I'll give you an example. You know, you know, was using Glance for buying a shoe for myself and, you know, I wanted a certain specific shoe.
Speaker 3:I have a, you know, middle distance runner. And I was having some pain in my heel for the last few months and I went online and said, Hey, can you get me a shoe? Oh, by the way, I seem to have a pain on my right heel. So can you kind of take that into account? Of course, it did.
Speaker 3:It kind of started showing me shoes which you know, OCA shoes which were a little more had a bigger heel. And then I kind of went away and came back later on and they of course showed me the shoe how it would look on me and whatnot and therefore I bought it. And next like few weeks later I was buying something for you know for a wedding that I was going to go to and I said hey can you just get me some some shoes and it basically went in there showed me the shoes that would look great on me on my on my you know on the theme of wedding but actually went back and said okay this is the shoe I think you should get. Know it's not the best option for you it's the second best option but it actually the heel has a higher cushion. And I was like, okay, it just did the exact thing I wanted it to do.
Speaker 3:And that's the idea. It is using inspiration, the context combined together and then helps me buy it. The buying process is also amazing because if you think about the buying process today, I have like bunch of credit cards, I'm members of a bunch of these programs and whatnot. And I think each one of us probably have like a thousand to 2,000 offers running on us at any point of time. We never use them.
Speaker 2:Yeah.
Speaker 3:And once I started to connect and that product is in beta for us right now, but the minute we started to connect all my credit cards, etcetera, it was computing its way through and giving me the output to the exact place to buy this from by saving like 10%, 15%, 20%. I would have never done that. And suddenly, inspiration with efficiency, not just in efficiency of time, but efficiency in terms of pricing is the power of what we really try to do here. So, I actually think the second part of efficiency is useful, very useful. But once you connect it to the inspirational piece, it just becomes an insane Yeah.
Speaker 2:It's like, not only did you help me find something interesting, saved me money on it. That's how you build loyalty. I'm sure you see it in your data. Yeah. Talk a little bit about that, the piece in beta.
Speaker 2:So it sounds like you're going to go through to completing the transaction. You started at the discovery layer and now you're getting through to the whole transaction. Is that gonna
Speaker 3:be we just did a
Speaker 2:Did you guys go to the direction of building your own wallet or are you gonna do connect to UCP? It'd be kind of a universal connector. How how is that gonna work?
Speaker 3:Yeah. Yeah. I I... You know, our idea is to essentially, you know, we just did a partnership with PayPal and we're very excited about that partnership. We did a partnership with Mastercard.
Speaker 3:We are effectively saying, we want to ease the process of buying. The agent should be able to buy. The challenge with today's experience in our platform in itself today, as we stand today, is that I have to take the user away from the agent has to essentially stop at a point and send off the user onto some other website to close the transaction. But that's not efficient because the agent should be able to do this. So, with our partnership with PayPal, what's going to happen within like four weeks is it is just once you decide what you need to buy, it is just automated completely.
Speaker 3:Are not as of right now, we're not thinking about having our own wallet, but we just to connect agentically to every payment system out there and make sure that the agent chooses the payment system that's best for the consumer and then basically completes the transaction. And what I am hoping that we start applying on top of it again in few weeks would be all the offers that every payment system ends up giving to you and users don't use them. If you ask me, there's billions and billions of dollars of unused offers that consumers sit on. Not that all those offers should be used, but when applicable, they should get used and therefore we're going to make that automated also completely.
Speaker 2:Let's kind of peel the onion on Glance. So, one layer, building a consumer product's hard and you're kind of building a marketplace. It's not a marketplace business model, but the hard thing about building a marketplace, and you probably know this from ad tech companies in many ways are a marketplace because you've got advertisers, publishers, and then the people, know, the tech in the middle. And then you also have a network you have to connect to. There's like a lot of constituents in there.
Speaker 2:Here, you've gotta have, obviously, consumers. So distribution of a... As a B2B guy myself, I... You know, you climbed the mountain, you climbed one of the highest mountains tackling ad tech, and then you're like, Oh, that mountain's boring. Everest is next.
Speaker 2:I'm gonna go do a consumer thing too. Consumer scares me because distribution is very, very Yeah. And distribution meaning is the consumer side. But what's also hard is, and we'll get to that, but what's sometimes even harder is getting the merchant side. How have merchants reacted to this and give us this idea of the scale.
Speaker 2:Are you just in fashion with 10 merchants or what give us a scale for the supply side of this equation? And then we'll go on
Speaker 3:the Great. Consumer So on the supply side, so firstly, there are a few things that are principal to us as Glance, right? So we're not a merchant of record. We only want to be the agentic layer, the layer that provides the intelligence, enables consumers to make better choices, which means that the brands like us because we are not taking the consumers away from the brand. We want to make sure that the brands actually build their direct relationship with the consumers or their agents.
Speaker 3:And so, therefore, just that principle really helps the brands to say, Okay, fine. We will be open to work with you. Now, the second piece of this is we have roughly 400 brands already that work with us and many more in the pipeline that are just like connecting very, very quickly with us. And the idea is that each one of these brands, they want to make sure the agent is able to discover the brand because, you know, I could look up a product if a brand is not connected to us, like just by scraping the internet and I could still surface that and send the user away there. But if the brand agentically allows us to talk to each other, talk to them through the agent protocol, then it becomes much better because the agent is doing a better job of like understanding the brands.
Speaker 3:The brands are starting to be, hey, what do we need to do to be agent ready? And therefore, they're open to the agent being a new channel. And I think that openness is great. But I think coming on the consumer side of this is where we you talked about a two sided network. We actually built both the sides together.
Speaker 3:And then the consumer side, I think we did something which I think is very different and unique. Instead of doing individual user acquisition so that consumers can call my agent whenever they want. We believe that the agent is an element that should stay that should be out there and wherever consumers are. And therefore, what we decided to do was to build build our age or have our agent be part of a network, meaning like, of course, you can you can have the agent on your phone. But as part of the network, our agent actually sits in on the phones.
Speaker 3:Our agent sits on laptops, our agent is already sitting on TV sets, our agent goes and sits in the brands and their websites and their apps and retailers' websites. So, what we did is we basically did this massive exercise in the last two years to open architect our agent in a way that a phone manufacturer could actually pick our agent and integrate that agent as part of the firmware of the phone. Or they could integrate our agent as part of the firmware of the television or of a laptop or tablet. Similarly, we architected our agent in a way that a brand can actually take our agent and place the agent on its website. So therefore the idea for us was that, hey, instead of doing this one by one user acquisition, what if I essentially create a network?
Speaker 3:And therefore when a consumer logs into Glance Agent, they actually in turn log into a network. So, wherever that user goes, so he moves, you know, if he is on his phone, you know, there is a widget on the phone. There could be something on the lock screen that's active on the camera. They could use the same agent on their television. The idea is that if you're truly an agent, the agent needs to be supporting you in any context.
Speaker 3:Unfortunately, the context of shopping that has always been thought through is to say, I will come to the app, I will have an intent and I will basically type it in. But for example, if I'm watching a TV show and I see something that is pretty inspirational on the show, I can just call my agent. The agent would look at the scene and say, hey, do you really want to shop the scene? And then if if I say yes, it'll actually take me to the journey and figure out how I can shop what's on the scene. Same is true if I open my camera, I can actually go into, you know, to a filter on my camera and say, hey, what I see here is what I want to shop.
Speaker 3:So if I see something in the open world, that's a different context. So therefore, what I'm building in this user environment are two things. One, my user scale is pretty rapid. And secondly, more powerfully, I starting shopping journeys from different contexts. And that is my context graph that is very, very supreme.
Speaker 3:So therefore, when these brands and merchants see this, they're very excited by this, both in terms of the scale and the context with which we bring these users. So today, you were to ask me, my strength of my network is roughly about 100,000,000 network points in The US. That has led to on a monthly basis, we now have roughly 15,000,000 users who are using My Agent. And the brands love it because that's a pretty large scale user base available through an agentic pipe for the brands to essentially utilize. And the interesting part is about nine months ago, about 3% of those users were transacting.
Speaker 3:Today that number is roughly 15% on a monthly basis. So We are very excited about the prospect because I think we're just getting started.
Speaker 2:Again, I consume a lot of your news because I see it everywhere. You're very good at PR. I admire you for that.
Speaker 3:I will pass that on to my team.
Speaker 2:The, you know, so you're on some subset of Samsung smart TVs. You know, that's where I always bump against Glance when I'm going to that user interface. And then an increase in the data shows and, you know, I read a lot of the other just big picture data. People are using their TVs for shopping all the time. It's like, you know, they, I remember when it first came out, everyone was like, no one will ever do that.
Speaker 2:And I see it. You probably, again, you probably, this is where having your ad network gave you a really nice little bit of vision. You can kind of see where the puck is going to use that analogy because in the ad networks have really lit up with over the top, I guess they call it OTT. That was, that kind of gave you the indication that that was a good, smart place to go. So...
Speaker 2:Yeah. And then getting to 10,000,000, I know... So the big ones that get talked about are FIA. There's like this set of vertical agents. You guys don't really fit squarely in a bucket, which is fine, but maybe a little bit of that bucket and they tend to focus on fashion as you know.
Speaker 2:I don't think any of them are even close to 10,000,000 MAUs. You're definitely...
Speaker 3:15,
Speaker 2:yeah. They're not even close to 10 and you're already at So like that's amazing. And then the other thing I try to help merchants understand is, okay, yes, it's 15,000,000 MAUs, but the Agentic shopper punches way above their weight. What kind of data do you have on that? And by that, you and I know they, their basket size is typically larger and they convert like, you know, a whole different, like we've never seen before.
Speaker 2:So you can't take the old Google conversion math of, you know, maybe like 2% or something and apply it to this world. What are some interesting things you
Speaker 3:should I'll give you a because I think the like to like comparison is important here. So we have one of the brands who basically, you know, see traffic from both the channels. They see the normal the traffic from normal channel and they see our agentic traffic. And the Founder like pinged me a few days, few weeks back and he's like, hey, man, I got to just tell you the following metrics. And his metrics were the the return rate had reduced by 15%.
Speaker 3:The conversion rate went up by 2.5x.
Speaker 2:The
Speaker 3:basket size was up 35% and there was one more metric I'm forgetting. But like it was just like different. And it's like, we've like, he's been in the business for like a decade or more and he'd not seen anything move these numbers up. And like, I was like, okay, this is just like day zero, man. This is not even like we have not even turned all the elements off it that you're going to see because effectively what is happening was the agent when embedded into their website or the D2C brand's website was effectively helping the consumer make a better decision.
Speaker 3:And they were basically saying, you know, the agent would basically say, well, you're buying this. Actually, why don't you pair it with this? And then here's how it look and therefore and there's the guy who's like, we struggled so hard to upsell. And the agent was naturally doing it not because it was trained to upsell. It was just trained to do the right thing for the consumer.
Speaker 3:The upselling was just happening. And that was the best part. So... Yeah. So we...
Speaker 3:Yeah. The the numbers
Speaker 2:are just, like, better. Yeah. And I've I've been in ecommerce for thirty years. And, you know, my simple block diagram is everyone has their ecommerce layer and then their back office stuff. And, you know, you know all the ecommerce layer companies.
Speaker 2:And then you have your little thin layer of conversion enhancers, and there's like a thousand companies in our world in that layer. And then you have your customer acquisition layer, your Google's and all the things. And then that little conversion enhancement layer, I've seen retailers do two year projects to implement some personalization or a better search engine or a million things, and they get an 8% lift on conversion and they have a crazy party. And here we now see the ability to get not only... And we're not saying 2.5 x of that 8% or whatever
Speaker 3:it is.
Speaker 2:We're saying it's like 2.5% of the whole inshallah. It's crazy. Yeah.
Speaker 3:Yeah, exactly. What
Speaker 2:of those 400 merchants are you you talked about, you know, trying on things. Fashion is kind of natural. You mentioned that one. Can you kind of share a split of and then you mentioned home some home items and a couple of other things. What are the big category areas that your
Speaker 3:products So are so the major categories that have been on for a while, and therefore, the split is far more skewed towards the categories of the brands that have been on for a longer period of time. They've been in Fashion and Accessories. Those are the first two categories that we went out big with. We are now basically coming out with other categories and therefore the numbers for those are lesser. So it is not representative of the data where the consumers truly shop in, but it is a reflection of the fact that we started off with the categories which are very personal.
Speaker 3:So personal categories, those two categories have a much higher share for us today.
Speaker 2:Yeah. What's your advice to merchants that are leaning into Agent to Commerce and how do they optimize for Glance?
Speaker 3:Well, I think two, three things. I think one is the fact that you're leaning in. That's a great point because I think the bigger issue we see is that a lot of guys are just thinking of this as, oh, we'll figure it out. But, know, if the agents are essentially gonna be the, you know, one of the main shoppers in the next two years, you gotta really start working to market yourself to them also. So the brands and the merchants have to market to consumers and now also to market to agents.
Speaker 3:That's one part. I think the second is we have something that we're going to launch at our event. It's our platform for, you know, creating brand agents. And we have that platform that we'll, you know, talk about. But the idea is pretty simple.
Speaker 3:Just like you have websites, you have
Speaker 2:your
Speaker 3:own storefronts, physical storefronts. You've got to have your own brand agent. And the brand agent
Speaker 2:This is something that sits on my website and answers questions? Or this is more, right? This is kind of an elevated type the type of
Speaker 3:Far more elevated. Right? It's effectively the brand agent, for example, you know, I could if the brand agent's beauty is that it can sit on your website, but it can also sit outside of your website, right? So, it can go sit into a TV environment because the agent protocol can call different agents into that, right? It's just like a website, right?
Speaker 3:The website framework was I can use a browser and the browser can call any website to open anywhere. The agent protocol is I can basically an agent protocol can call any agent to come in anywhere, right? So, is an agent. The Glance agent can call a brand agent. The brand agent would then appear within the Glance environment.
Speaker 3:Similarly, the brand agent could open into some other agentic environment. But that means that you've got to essentially enable the brand agent to have all the insights about your product, about whatever your brand is about and has the ability to also complete the transaction then and there. And so therefore, the brand agent is a living entity And that brand agent carries all the intelligence about your brand and the ethos of your brand and ethos of all the products. And that's the way world will start to interact. So therefore, investing behind the brand agent is possibly the best way not just to work with Glance, but actually to participate in the agentic world.
Speaker 3:And therefore as we unveil some of these things, we are very kicked about this because think about not just the brand agent sitting in the wide world, but also think about the brand agent as a new unit for advertising. So, if you want to essentially advertise, effectively, brand agent is the new ad. And it's just going to sit there in the context of, you know, Scot's, wherever it sees Scot or sees Naveen. And the brand agent basically kind of appears there and completes the full transaction. So therefore, you know, it's the atomic unit of the new store.
Speaker 3:It's just agentic in nature. That's where we think that that'll be the most important piece for people to anybody to essentially create and will enable help enable that.
Speaker 2:I think that I can kind of see where you're going with this, right? Because you will offer a distribution mechanism for these brand agents. And you're you've got you're sitting on this giant network of, what'd you say, a 100,000,000, 110,000,000 devices? A very big number of devices. Yeah.
Speaker 2:And, and then all these MAUs and, you know, now you're going to give a very, almost an app store is interesting. I know it's not an app store, but it's a good metaphor for it for people to kind of connect the dots. So I think that's going to be great. I'm very intrigued by this intersection of ads and Agent to Commerce and where this goes and how you do it in a way that's not obtrusive to the experience, that adds to the inspiration part of it, not detracts. I think our old era of sponsored listings just splattered all over the Amazon pages It was not a very productive version of the ad system there.
Speaker 2:That's awesome. And I know you're super busy and I want to make sure we get out on time. You've spent a lot of time thinking about this. Any interesting views of where you think things go? And I know it's hard to look past six months, but, like, maybe the next year or sometimes people are anchoring on 2030.
Speaker 2:What what do you think about where this is all going?
Speaker 3:It's so hard to anchor in 2030. It's like, I think you're right. The the things are changing so rapidly that, you know, just a year out year out view is a good enough view in my view today. But I, you know, the... I fundamentally am looking at the infrastructure for commerce is rapidly evolving.
Speaker 3:It is so rapidly evolving because the consumer behaviour is adopting to this new agentic world much faster than what we imagined. And so, I think the idea and I said this earlier in your pod that, you know, 50% of human population would actually end up using an agent. And that's a much more rapid evolution or adoption of any technology than before. And so, you know, even if like x 53% to 5% of the GMV starts to flow through agents, that is just an insane amount of context learning that will accelerate the uptake. So, if you take a five year view, I think you know some like a third of the transactions in the world will be agent driven.
Speaker 3:Not just for the choices that the agent will make on you, but the amount of savings that the agent will drive for you. So therefore, like the saving, like doing things to not overspend is an important human way of thinking about commerce. Agents will just like enable all of that and make sure that you'd never have to do this. And so, yes, we are going to start to see a very different world out there, both in terms of the way shopping happened and also in the way advertising is going to evolve.
Speaker 2:So if people want to try Glance, glance.com, that's easy to Excellent URL by the way. And then the event coming up is pretty exciting. The pod will be out so people will have time to look at that. It's called Glance Forward. It's kind of your, you do two events a year.
Speaker 2:This is kind of the second, the
Speaker 3:fall one. Second one, yeah.
Speaker 2:And it's going to be in New York, November 17. And we'll put a link to the people want to apply We to have a lot of merchants here. I think you're going to have some really there's going be a lot of merchants that want to learn more about this new launch you're having and everything. I'll be there, so that'll be fun. We don't know exactly what we're gonna talk about, but I'll be there in some form.
Speaker 3:We'll have
Speaker 2:fun. I sent your team a bunch of ideas. Have a I thousand think just a conversation like this would be kind of fun, at least for you and I. So there you go.
Speaker 3:Absolutely. We'll have fun.
Speaker 2:And then if folks... Do you put a lot of your thought leadership out and where where should people find that?
Speaker 3:I I put most of it on LinkedIn. So I, you know, just find me on LinkedIn and we we we put a lot of our stuff out there.
Speaker 2:Awesome. Naveen, thanks for coming on. Congrats on everything you've built over. It's just amazing to have built such a large B2B ad tech business and then tackling consumer. I have more power to you for you're taking on way harder things than I ever could.
Speaker 3:Great speaking to you, Thank you for inviting me to this.
Speaker 2:Absolutely.
Speaker 1:Follow, rate, and share to stay up to date with Retailgentic, your guide to the future of AI shopping agents.