Personalize EVERY Customer Experience for Higher RPV

Personalize EVERY Customer Experience for Higher RPV Speaker 1: Welcome back to another episode of Chew on This. Today's a special episode brought to you by Envive, and we're here with the founder and CEO of Envive, and he's gonna be talking about all things AI, what makes their platform incredible, some of the incredible brands using it, and why you should be using AI the way Envive has been productizing it. Speaker 2: Absolutely. I feel like AI is so daunting for some people. They just don't even know where to start beyond ChatGPT. So I think this is gonna be a really good episode for people. Speaker 1: Okay, let's start with I think the broader viewpoint because I think you have a great bird's eye view to where AI is going, but also obviously you're in the weeds and getting to see literally every interaction. But if you had to kind of maybe talk through somebody who doesn't understand this, if the question came up, which is what will be the biggest shift in how consumers interact with eCom sites in the next 12 to 24 months? What would be, how would you explain your prediction? Speaker 3: Yeah, so where we got to start is where is consumer behavior today? So a fun fact is that ChatGPT utilization for consumer use cases, weekly active users has Doubled in the last three months to 800 million weekly active users versus the prior 12 months. So what that means is that we're seeing exponential increase in weekly active utilization of ChatGPT for consumer use cases, including shopping. And then like last week, OpenAI just announced inline native shopping. So as people shop, they can transact directly through ChatGPT. And so what that means is that like, if you think about that plus exponential growth, it stands to reason that like this is going to be Something similar to what Google search was when it came to discovery of products in the early 2000s. Speaker 1: This episode is brought to you by Envive, the most brand safe AI for eCommerce. You know that moment when you're on a site ready to buy something specific but the search gives you nothing useful and there's no way to ask real questions about the product? That's frustrating and it's happening to your customers too. Envive fixes that. Their AI-powered agents makes your site smarter, doing dead-end searches and clunky chatbots into real conversations that convert. It's search, sales, SEO, and insights all in one seamless brand safe experience. We'll dive deeper later in the show, but if you want a preview, head to envive.ai slash Chew on This to learn more. Now, let's get back to the episode. You know, the inline shopping piece I think is like, you know, you touched a little bit on this and it's more so around like a lot of people start to react and like, oh my god, is this going to change everything? And I think you have an interesting viewpoint, which is like, you know, things are going to evolve. Right. And I think what you do to stay ahead of the times and stay with the times is equally important. When it comes to users and let's say customers or the average consumer, right? Do you also think like something like ChatGPT and AI is also not going to be maybe generally used as wide as let's say what like Google got to, right, with search? Do you ever feel like maybe there is a piece where like, AI and even like a ChatGPT won't get to as widespread of these just given like maybe there's a little bit more knowledge required, maybe there's a little bit more, you know, just the user experience is a little bit different from something like a Google or like website shopping. Do you think there's ever going to be that like breakpoint? Speaker 3: Yeah, so is there a breakpoint where ChatGPT may not completely replace brands? Is that the right way to frame that? For sure. I definitely believe, despite what my friends in the venture capital community might think, I don't believe that there's going to be some crazy disintermediation risk. And the reason for that is just that someone still needs to be thoughtful about consumers, understand what their needs are, design great products, buy inventory against them, manage and warehouse that. It's not going to be dropshipping through ChatGPTs the future of commerce. We all agree with that. Because we've been through this before with eCommerce, with mobile. It's all going to be about how do you adapt to this future and still create the best end shopping experience for your end users. And what we're building at Envive are the tools to help with that, both in terms of becoming more relevant in generative engines, but then also creating the best on-site experience possible so that the qualified traffic that's coming to your site can be meaningfully engaged and is more likely to transact and then come back to transact more from your property where you have a higher margin and opportunity. Speaker 1: That's solid. Speaker 2: All right, let's move to Envive. So before we even get into that, I was at a summit earlier and a lot of this stuff is happening with these chatbots on websites. Explain to me what Envive does and why is it better than everything else? Speaker 3: Oh yeah. So I'm going to explain that. I want to just say at this point to start, right? I have absolutely nothing to sell anyone. I'm building Envive because I want to build the best tools for brands and retailers to adapt to an agentic future. And there's things that we are world-class at, and there's things that would not, which I'm happy to talk about, where we're not a great fit for everyone. And so I just want to say that as a disclaimer. And then in terms of what Envive does and how we differentiate is that there's a lot of tools out there which are basically chatbots on your website. And our goal is, when we were a seed company, We started with a sales agent, which is sort of like a chatbot on your site. What was differentiated and unique about that, and then I'll get into what we do as a whole and the vision, is we realized early on that the best brands, and we work with amazing brands like Spain, Supergroup, Coterie, brands that are doing amazing amounts of GMV and have an amazing brand that they built with their users. In order for them to embrace AI, We have to build for three things. Performance, because you guys run a great company and I imagine that you all care about conversion and GMV, right? So performance matters, but then also controls matter. You want to control the AI and be able to... So it's not a black box. You know what's going on underneath it, right? And then safety, right? Like you have cultivated an audience and you have a relationship with that audience and what we're doing is crazy, right? We're putting our AI In between the trusted relationship that you built with your end user, you've worked hard to acquire, you paid money to acquire potentially from paid channels, and so you need to trust that the AI interactions will not have any downside risk for your brand. So our sales, our core architecture is those three things. And then we built agents on top of this, this custom large, like we trade a custom LLM, that's a large language model. And on top of that, we have agents like a sales agent, which is a chatbot. And then we built a search agent too, which can replace onsite search. And then we have another agent which does Content creation for better ranking SEO and generative engines. So our whole vision is, in the future, having a system of interconnected agents that helps drive an AI-driven growth engine for your brand that you completely control and own. So that's the capital division we're doing here at Envive. Some part of that is a chat, like a sales agent, which can help users who are further down the funnel better interact and engage with your brand, but the bigger vision that we see is in order to prepare for an agentic future, your brand has to itself become an agent, and that's what Envive does. In fact, Envive stands for Into Life. We bring your brand to life with a custom LLM that's purpose-built towards your datasets that you sit on, like your customer transactions, your brand voice and tone, your product catalog, and we unlock that into a custom trained model and then have agents on top of that which go to work for your business. Speaker 2: That's really cool. How long do these models usually take to train after you get the product catalog and all the data? Speaker 3: That's a great question. So these models can, depending on the size of your brand, they can take between two to four weeks to not just like train, but then also do the post-training process. So if we do multiple rounds of validation work to make sure that in different scenarios, the brand voice is fully dialed in or that the model is working just how you want so that it's fully aligned to, like I said, like the biggest thing is brand safety, right? Like in different scenarios, is it responding the way you want it to? So it takes typically two to four weeks to go live. Okay. Speaker 2: That's very quick. Speaker 1: You touched a little bit about the training piece of this, which I think is a great prompt around just understanding when it comes to whether you're using Envive or some sort of solution here, whatever route you're going, when it comes to even just the training piece, when it comes to agents, is there a right and wrong way to do that training piece too? Or sometimes I look at even ChatGPT and stuff like that and I feel like If I just hoard all the information I have and I dump it, ChatGPT will figure it out. And I kind of just walk away. I'm like, all right, I'm going to come into this magical world of it all solved. How does training really work when it comes to agents? Speaker 3: Oh man, what a great question and I really wish that what you said worked because that would make my company so much simpler. Speaker 2: It never works. Speaker 3: That would be awesome if that worked. I like to say, I like things that are simple and unfortunately we have had to build quite a complex product and the reason for that is what you said, which is a really great point, that why can't I just take all of my data and information and dump it into this model and be like, hey, figure this out. You're an intelligent system and figure this out. And so in certain situations that can work, but it doesn't work all the time because like prompting, as we know, it's a brittle approach and it doesn't create the consistency that you want. And there's some really fun examples of this on YouTube or TikTok that we've seen where it's like, hey, you want the model to behave a certain way and it does once or twice and then you make one change and then it breaks everything. What we do is a whole training process using fine-tuning, where we're updating the actual model parameters, which is like the neural network architecture of the model, to actually make it smarter. And the way that we like to explain that is if you watch The Matrix, shout out as an early 2000s kid, there's a scene where Neo learns Kung Fu. And then you like basically they inject something into his brain, and it goes to the whole simulation process and he learns all these martial arts. That's what we're doing with our models that we're simulating thousands to hundreds of thousands of synthetic interactions with between your users based off of data you've already have to Understand potential conversion paths to train the model to be really smart in your particular brand and customer base. And that's how we achieve breakthrough results. And then that's part one. Part two is then using reinforcement learning based off of people coming to your site and other properties to engage and convert so that empirically over time the model improves so you can always trust that you have the best AI for your brand. Speaker 1: A lot of people talk about like, AI is only getting smarter, it's only getting smarter, the more you feed it, the better it gets, etc. Is there also an element of like, it can also just get more confused? Like, like, if when I when I think about all of the different places, let's say we're driving traffic to from App-V, okay. And let's just say like, whatever we have from from an AI solution, is, you know, being trained to learn like, okay, we drive traffic from Meta, Google, maybe some direct and organic, okay. And all of a sudden, I kickstart TikTok shops, all right. And like now I'm getting a bunch of leak leakage from there and like traffic coming spilling over from there and, and seeing a whole new channel get created, right. In a scenario like that, where an agent for let's say, let's say for the last year has been built around trained around Understanding our ecosystem and now you kind of come in and let's say you kind of change that up or you create a whole new entry point of data. How do typical agents that have been trained well interact with new data inputs like that? Speaker 3: Yeah, that's a great question. So just to play that back to you, because I want to make sure that I'm answering it well. It's like, hey, you've trained this agent, and then you have this like new sales channel, like TikTok shop, right? And then you start getting traffic from that, like, how does the agent adapt to that new data set? And so our perspective is like, two part, like, one is that the way it depends on on I think that a lot of folks who use AI tools today are using ones that wrap around ChatGPT, in which case they're feeding more data. It's just another data source that's going into the model context. That's a bit, like we talked about, that's a bit of a more brittle approach because then you're just like cramming a bunch of stuff into context and kind of hoping it works. What we're doing The way that it would work is like, hey, this is a new sales channel. People will have a different intent. People will be interacting in different ways. The model will learn from that and then adapt based off that. And then if we feel like we have an evaluation mechanism, so like if we see that people from TikTok have a different set of queries, like for example, you might have TikTok specific content and so they're Intent is different. Then we can retrain the model to be, and you can add a new skill to the model to be better or more proficient at TikTok based users and their intent. Because what I imagine is that, you know, people coming from TikTok just like Reels are in that doom scrolling mindset. And what you want to do is maintain that on your onsite experience and maybe fetch more videos and fetch more stuff that maintains that dopamine treadmill we all have when we're going through social channels. Speaker 2: Absolutely. Just to kind of go off that, let's talk about guided shopping for a little bit. I don't know if the listeners know the difference between an AI or just a chatbot and what guided shopping is. If you want to explain that. Speaker 3: Yeah, that's a great question. So, we gotta bring this back to, like, let's just take AI out of the equation, because, like, AI is, like, as we know, is artificial intelligence, but most people have, like, inside sales teams, which are, like, BI, biological intelligence, you know, which is, like, which are, that's actually our first, one of our first customer curated, you know, they were doing really well. They had thousands of experts that were interacting with consumers to sell them skis and snowboards and all these, like, considered purchases. Yeah, so this was a guided selling experience, a guided shopping experience because, hey, if I'm new to skiing, I don't know what to get. Like, I need some help. I need some guidance to the shopping journey. So what we've realized is like AI needs to do two things. Like it needs to build rapport in a way that's appropriate for AI. You don't want an AI model asking you, hey, what's up, dude? Yeah, yeah. You're not trying to be friends with an AI model, maybe you are, but generally what it needs to do is ask the right questions and then also have the specific skill of understanding and clarifying that higher consideration journey to guide users and then also retrieve reviews, retrieve social media and other proof points that make you feel more confident that, hey, this is going to work for me. So a great example, this is my fiancé, shout out Aaron, She's petite, she's 5'1", and so she was on actually using what we did for Spanx and she was trying to find the right shaper for her and our guided shopping experience with Spanx. was pulling forward all the site reviews and all the metadata for Shapewear that was for more petite women. And when Spanx has like hundreds of Shapewear garments on their site, it's really tough to know what's going to work for me. And so that guided shopping experience kept her on the platform instead of going to Google or some other platform where they might lose the transaction. And she ended up staying and transacting with Spanx. Speaker 2: So interesting. So is that I guess that AI, is it recommending upsells, bundles, things that you might be interested in? Is it doing everything that a sales agent would do? Speaker 3: Yeah, that's exactly right. So think of it like, hey, I'm buying cookware, and I added, like, I come in off, you know, non-branded search on Google, like, searching with best, like, non-toxic frying pan, and I go to, like, one of our cookware, like, cookware co. on Greenpan, and I added, I add, like, their non-toxic cookware set to my cart, and then I get a notification, which is like, hey, do you have a non-toxic way of storing leftovers? Speaker 2: Wow. Speaker 3: And then I can get the storage container set. And that's a great user experience. That's a good point. If I'm going to buy this, I'm going to have leftovers. And that's a great upsell opportunity, but in a way that's a better experience for the brand. Speaker 2: That's so cool. And because you fed it your brand voice and kind of the way you talk, that AI is talking to that consumer like that, I'm assuming. Speaker 3: Yeah, that's exactly right. Everyone wants the brand voice. No one wants ChatGPT voice. Speaker 2: Yeah, you want to feel like a human experience now. It's really cool. We were talking about it. I was talking to Ash about it the other day. So we have when you come on our homepage, you have two options. It's like shop all and then take a quiz. And we try to push people towards this quiz funnel. But I'm wondering, is this the better move where you chat with a person or whatever the CTA is, and then now you have this guided experience versus, you know, a generic quiz that's going to bring you to the same page every time? Speaker 3: Yeah, that's so interesting that you ask that because like what we see all the time is like the quizzes were our attempt before AI to do like basically help people navigate that journey. It was like you said, it was guided shopping. It was like, I want to collect data from you and then I want to transact. But the challenge that we saw with quizzes is that those typically drop off after three or four questions. You're like, listen, I'm here to buy stuff. And so what a sales agent experience does is by giving them some value with each interaction and then asking the right follow-up question, it's more fun to complete that quiz because now you're building a relationship versus going through a checklist. Because imagine that the best sales agent or sales associate at a store is not going to come up to you and basically ask you a bunch of monotonous questions to figure out what your preferences are. There's going to be some rapport building. There's going to be some showing you this. There's going to be some social proofing to show, oh, this has worked for other people. And that's how humans are wired. Speaker 2: It's a conversation. Speaker 3: Exactly. It's a conversation. And it's relationship building. And that's how humans are wired to actually want to transact and build trust, which is overall the point of a brand is that you're building trust that you have a unique solution to someone's problem. Speaker 2: Makes total sense. Speaker 1: Let's get back to Envive. Because this tech is honestly changing the game for eCommerce. Their AI powered search understands how people actually shop and Typos, vague queries, natural language, and their sales agent act like the best rep, guiding shoppers and answering questions in real time. It's all designed to protect your brand while increasing conversions. And here's what sets it apart. Every interaction powers deeper insights. You learn what customers are searching for, what content's missing, and when they're dropping off. It even helps you rank better in LLMs, so you're not just fixing on-site friction, you're preparing for how discovery works now. Envive is helping brands like Spanx and Coterie prepare for the next era of commerce, where discovery is conversational, AI native, and always on. See it for yourself at envive.ai slash Chew on This. Now, let's get back to the episode. So I'm curious when you look at the typical websites today, right? I'd say more sites obviously do not have agents and tools to help guide the customers through journey and most brands do typically rely on quizzes and whatnot. And I think part of it, another, I would say like maybe myth or hesitation that comes up is like, oh, I don't want this like, Chat bubble and like to take over the screen on mobile and like I think it'll take away people from the user experience that may be coming from just like people guiding themselves to shop and browsing and whatnot. It's similar to how like you know some people feel about like exit intent pop-ups right or or even like triggering the pop-up too many times within one experience. What do you say to the brands, I mean I would say as simply as to like go and test it first before just claiming some result, but what do you say to the people who are more on the line of like, oh I just don't want to disturb the people who are coming to our brand to experience the brand and browse themselves? Speaker 3: Yeah, I agree with their perspective. Like, yeah, you shouldn't disturb them. I think it's about, like, engaging them, and I don't know, it's like a, it sounds very, yeah, but like, the way to do that is going to be like, you don't, you know, from a UX perspective, you know, you want to, the way that we do it, and we're receiving some great results, is like, for example, hey, someone is looking at product PDP1, and they go to PDP2, and then what we see in session data is like, there's a lot of, like, bouncing back and forth, and you guys have probably seen this before as well, and it's like, well, A smart system in that situation will just start like to do an automatic comparison between the two and kind of like show like pros and cons between these two and it's just embedded on the page. There's no pop-up but it's like right there on the PDP page above the fold so that they're gonna see it on mobile, right? And so that's a smart way to solve for that pain point in a way that isn't super intrusive because you're right, like no one wants to like have like the spin wheel thing or the pop-up, you know, like win the prize. Speaker 2: I hate that thing. Speaker 3: Totally, yeah. Speaker 1: That makes a lot of sense because I think even from my buying experience, when I go to shop, typically I know exactly what I want to buy, but the times that I don't, I find myself looking for answers outside of the environment of a website. So I'll go on like it reviews or I'll like go on like, you know, Google and just search something and whatnot. And but I feel like if it if it was able to just know what I'm going to ask right there, I feel like I would love an agent that helps me through that. But I I've always looked at it as like, oh, like, I don't want to talk to this bot. And I think that's where the maximum shift has to happen, which is like, this isn't just chalking up to letting your customers talk to a bot. It is creating an experience that's going to elevate your customers' time there. Speaker 3: Yeah, that's totally right. And that's why the way we think about it and the way our placements work is like, we want to minimize friction and cognitive load on the user as much as possible. So in my mind, If the user is having to type in a question, I love that as that's a great way to understand someone's intent. We want that. We enable freeform input, but we want to lower the friction as much as possible, which is why we're going to have voice-based input, right? And make it really, really frictionless. Because what we see with open AI utilization and their data is that voice-based input, because it's so much easier to talk than it is to type everything out. Speaker 1: I do voice all day. Speaker 3: Me too, right? It's such a great input mechanism, which is why we're enabling that. Plus, you can then train your brand voice into the voice. Speaker 2: That's crazy. Speaker 3: And so that's why we're always thinking about like how do we just reduce the friction and kind of like our mental model is we're building cheat codes to your brand and to your website. Because people want to like, people come to the site, they have some sort of destination line or they have something they want to get to, want to make it as simple as possible to get to that, whatever that is. Whether it's like my fiancee's use case where she wants to find the best shapewear for someone who's like 5'1 and she wants to Not only see a bunch of products, but she wants to see the reviews and have that trust that this will work for me. And we want to serve all that up as quickly as possible to them. And the first challenge there is going to be making that input friction as low as possible so that people can get there. And the way to not do that is like that thing on the bottom right which expands and takes over everything. And suddenly you're looking at a group of products and now you're looking at this like blank text-based input. And they're like, well, now I have to think of a question and type that. We want to anticipate the question, put it in line with the other products. And so you can just click that and be like. Speaker 1: That's so cool. Speaker 3: This is like, we want to read your mind. And so then we can get there. But in like a cool way. Speaker 1: Yeah, cool. Yeah. You know, I'd love to know, I know this is like a tough question to maybe have one answer to, but if someone asked you like, what's the one way to future proof your storefront experience, right? Where would you tell a brand to get started? Speaker 3: What a great question. What is the one way to future-proof your storefront experience? So, the question is, we need to go back to where user behavior is shifting, right? The second part of your question, future-proof, like where is the future? So what we see is that user behavior is shifting towards these, like using generative engines to find products, right? So what that means is that, but that doesn't mean your storefront will go away, because I think people love to have really great branded experiences, right? Like people want to go, like you're investing in these assets, people want to see your reviews, your assets, and want to be part of the shopping experience. So in order to future-proof it, I think that there's two things that need to be true. Like you want to rank better in generative engines, but two, and I would say just as importantly, if not more importantly, like you want to make sure that the qualified traffic that's coming to your site has a really great experience on your storefront. And so I know you asked for one thing, so I give you kind of two things. So what I would say is that like, how do you, depending on one framework to think about this is like, Offensive versus defensive, right? Defensively, yeah, you want to rank better on generative engines because that's where consumer behavior is going. But offensively, how do you create even better experience on your site than it is to shop through generative engines for the qualified traffic that's coming to your site? Because today, probably for most people, 2% to 5% of traffic is coming through a generative engine today. And sure, that will grow and perhaps it'll grow exponentially over the next 18 months. For the traffic that's coming in the here and now, how do you make sure that that experience is really, really top-notch when it comes to finding the right products, anticipating people's needs, and so on? I think that's going to be the key area to focus on. Speaker 1: One other key question to that, and then we can jump into the rapid fire. You look at the different categories of brands today, right? I mean, beauty, you have THC and cannabis, you have apparel, you have health and wellness and food and beverage, etc. Are there, and you kind of said this in the beginning, which is like, maybe there is products or brands that you guys aren't a fit for, maybe agents aren't a fit for, but Are there certain categories that you think this still doesn't work for yet or it's not ready for? Or do you think there is some level of generally applied principle where this is going to more times than not help than potentially not? Speaker 3: Yeah, it's a great question. We see the best results, we see great results across the board, but we see the best results with higher consideration categories and brands that really value and understand what it means to be a brand. And let's be like, y'all operate a brand and I've worked at and partnered with a lot of folks who are operating brands and not everyone truly values brand equity and what that really means. And so, The way to test that is like, hey, do you want to make sure that your branded experience comes through? Spanx really cares about their brand. And they care about their audience and how they engage with the audience. And so if you care about that and you have a more considered buying journey, then you're going to see stellar results with what we do because your users today, I guarantee you, you have a lot of qualified traffic that's coming to your site that's churning off because they just can't engage with your brand that they want to. And what you want to do is help them engage on their terms. So that's where we like to focus is helping brands like that really reimagine the shopping experience for their audience. Speaker 1: It's awesome. Speaker 2: That's great. Just wanted to add to that. So some of these brands that have, let's talk THC or anything in the health and wellness space, there's obviously compliance issues there, right? As far as things you can and can't say, is that bot trained on all of the compliance? Speaker 3: Yeah, that's a great question. A key thing that we do, which actually does differentiate us because we can guarantee against the downside risk of being non-compliant. So to put it differently, yes, we're 100% compliant and we're able to, the reason that we're able to achieve that compliance is our training process where we systematically simulate and train out the undesirable behaviors that might violate, for example, if you're a medical device company, FDA guidelines or FTC guidelines or your own internal brand safety guidelines so you can ensure consistency in performance. Speaker 1: Okay. I wanna get into a little bit of, I'd love the last round we had. Aki, you wanna kick off the first question? Speaker 2: What's a UX best practice that holds brands back? Speaker 3: UX best practice that holds brands back. So, I'm gonna say that we gotta reimagine, is that like a lot of our best practices are based off of how eCommerce has been pre-AI. Yeah. So when you reimagine AI as a whole, it creates a whole new set of UX paradigms. So I'm not answering your question, but what I'm saying is that, like, the question comes from that assumption that, like, hey, our current set of UX practices are the right ones. But they are, but for a world that was more pre-agentic. Speaker 2: Wow. Speaker 1: Damn. You just, like, threw everyone back in the matrix. Speaker 2: It's almost like rethinking everything. Speaker 1: That's awesome. If you could remove one element from the average eCommerce site forever, what would it be? Speaker 3: I could remove one of that spinning wheel. Yes, it is going to be the spinning wheel. Speaker 1: I love that. Speaker 2: All right. Has customer behavior ever contradicted your assumptions? Speaker 3: Yeah, man. So like I mentioned, our first product, Envive, was a co-pilot for inside sales folks. So that's where we really cut our teeth in knowing, that's how we know that our models are good, because we were using that as a co-pilot for people who were chatting with customers to transact them on high ticket items like skis and snowboards and so on. And what we found is that Not all the, like, only half the sales folks really used it, even though the best, the ones that did use it saw great results. And what we saw is that, like, hey, we need to, like, Rethink co-pilots to just full autonomous agents and that's what led to this current evolution of what we're doing today is fully autonomous where there isn't a human in the loop because we realized that like it's when you put AI and humans together in a workflow perspective it's tough for people to always embrace it as something that's going to help them be more effective. Speaker 2: That's interesting. Speaker 1: Very, very cool. Speaker 3: Hot take about the future of shopping The future of shopping is very bright, you know, like I I just I don't know about your guys LinkedIn's but mine is Mine is full of advice that is actually doing, you know Fear mongering about the future and I'm I just believe that like look what's happening today has happened before with with the internet with search with mobile and What you do see is that the brands that are successful are the ones that adopt. I worked at Walmart, which didn't believe in eCommerce until 2017, maybe. But Amazon did, and they took all the market share for a good 15-year run. Same thing with mobile. The companies that adopted, today 80% of our customer traffic is mobile web. And so the brands and retailers that adapted have seen tremendous success. And so this is just a new technology, but ultimately, consumers still want a fantastic shopping experience. And I think that this is going to be a really exciting time ahead to use this. The brands and the retailers that really experiment and dial this in will, of course, adapt and continue to grow and thrive in this new future that we're all going into. Speaker 2: Very cool. Speaker 1: Last question for you, Nikita, is you touched on, and this isn't a rapid-fire question, but I'm very curious, you touched on brands that have high consideration typically see maybe a little bit better results than brands that don't or it's just maybe a better fit. You look at the examples of, let's just look at like t-shirt companies, okay? You have something as simple as maybe as like a true classic tea which averages out to like nine bucks a tea and then you have other t-shirt companies like maybe a Built or Rory or some of these that are like $34 to $68 for a tea, right? Simply put, you know, my question really is when it comes to how to identify if your brand is a brand that requires a high amount of consideration, right? Is it typically driven by price point that uses that determination factor? For example, True Classic on average sheet comes out to $8 to $9, right? I think a brand like that still has sometimes high consideration just maybe because of the filtration of people go down to there could still be people who don't have the budget or maybe reaching, right? Versus just buying a $2 Hanes Tees. But then Vuori, right, also has a high level of consideration. It's $68 for a tee, right? But again, the people that are filtering down are people who are more apt to potentially buying it, just given the brand and stuff. So how do you look at, when you said something like, you know, brands that have higher consideration, or brands who have customers that take a lot of consideration, how does a brand, you know, going, if I'm a listener or a viewer, And I'm trying to figure out like, hey, where do I fall in the scale? Where does my brand fall in the scale of high consideration, mid consideration, low consideration? How would you help define that? Speaker 3: Yeah, it's a great question. So my personal definition of consideration is like, what is the tolerance for error or what are the stakes? You know, like, what is our tolerance for getting the wrong decision here? So think about like buying a car. That's like super high consideration. You don't want to, you know, go take out a load on the wrong car, right? Buying a house, like those are... Speaker 1: That's a great answer. Speaker 3: Or really high consideration choices, who you marry. So those are high consideration decisions. And then the other end, a lot of stuff at Walmart is like, hey, groceries, it's very commoditized. People don't generally care what type of dish soap they buy. People just want to say dish soap, whatever, just give me the same stuff I normally buy. I don't really care. So the stakes are low, relatively speaking. And what's really interesting, what you said, is actually a great point that is correlated with price point, but not always. So it could be that Something that's like $9 is higher consideration depending on your budget, but also depending on like, you know, just your value system, right? And so that's, and the other part of consideration I mentioned like tolerance for error, but it's also the number of choices, right? So if you have like a paradox of choice situation, then that also creates more consideration. That's why streamlining your product line is sometimes helpful or having something that helps you through that guided shopping experience navigate and traverse that paradox of choice effectively. Speaker 1: It's always been an interesting thing that plays in my head. A brand like NeoCell, which is an $18 collagen, has a 4.2% conversion rate. A brand like Opvi, which is a $40 collagen, has a 3.8% conversion rate. At the end of the day, consideration in some ways should be far less on a NeoCell. It's just a collagen, it's $19, let me just get it. I don't really care, it's a dish soap. Then you look at Obvi and it's like, oh, I got to really think about this. This is expensive. This is pretty much paying over a dollar a day for collagen. I got to make sure I take it all the time. I got to make sure I'm consistent with it. I don't want to waste this. And at the end of the day, you can clearly almost tell there's the same level of consideration for both. Speaker 3: Yeah, it makes total sense, right? And I think that the way that you frame that is so important because like collagen, if you think about that, it speaks to how you speak with your audience and who your audience is and they value their health and the dollar, that's like, that's nothing for the benefits of collagen. So it really, I think it comes back to your audience and engaging them effectively and helping them. And sometimes like they don't know and that's about where they need to be like, Like, it's not about persuasion, but it's more about education, right? Helping them understand the importance of collagen for their health. And as soon as you educate people, suddenly their values can shift and their beliefs can shift. And I think that that's where AI can provide a lot of help because you can't, you know, have an army of people out there, you know, slinging collagen at those price points, right? Speaker 1: That's sick. Well said. I think this is, again, I'm super, super glad that you're able to share so much knowledge. You clearly know this at the back of your head. I think you're building something really, really powerful. I'm excited to use it for Ferabbi. To leave our audience with one last thing to chew on, one thing to take away, what's that one thing that everyone should go and do and work on their business today? Speaker 3: Yeah, so here's a Chew on This moment. The biggest thing is, look, we all know that consumer behavior is shifting. And the question is, like you said, like, what is your brand of promise that you're making to your audience? You know, who's your audience? How considered is your buying journey? What does it mean to meaningfully engage users as consumer expectations are shifting and becoming more agentic? In certain worlds, that might make sense to partner with Envive or explore these types of agentic solutions. In other solutions, they may not, but I think it's really about taking that from first principles to really figure out what does it mean for your audience as this new technology breakthrough is changing how people shop? Speaker 2: Chew on that. If you want more from us, follow us on Twitter, follow us on Instagram, follow us on TikTok and check out the website ChewOnThis.io.