Latest / Talk Commerce / AI-Powered Marketing Revolution: Insights from Udayan Bose

Transcript

AI-Powered Marketing Revolution: Insights from Udayan Bose · Talk Commerce

• Brent Peterson (00:01.533) Welcome to this episode of Talk Commerce. Today I have Udian Bose. He is from Netlixir. Udian has been on the show before and we're always excited to talk to him, especially around gen AI and Google and all the fun things. Udian, go ahead, do an introduction for yourself. Tell us your day -to -day role and give us a passion in your life. Udayan Bose (00:21.165) Awesome. Brent, it's so amazing to really sort of join back this cast with you. I I had a blast last time when we talked. So thank you so much for having me again on this show. My name is Udayan Bose and I'm the founder and CEO of NetElixir. We are an AI first digital agency. By AI first digital agency, I think we have really built our own machine learning platform. through which we are able to uncover hidden revenues from the Google Ads and Meta Ads and so on. It's a combination of machine learning and predictive optimization. And today, what I'm going to be discussing a lot and sharing with all of you is our latest and really excited about that is basically our first three Gen. AI tools sprint. So we really have taken the API from GPT and built a security layer on top of that. and really sort of started imposing the workflows for SEO and paid search, et cetera. So we have really built our own Gen. AI powered marketing operating system. And these are three tools which our team has been very successfully using as well, which we wanted to launch and share with all of your viewers as well. to check it out for free. So very passionate about AI, really, really a big believer in this concept that humans that use AI will always beat humans that do not use AI. And that's, think, what my topic of discussion is and really excited to be here. Brent Peterson (01:57.661) That's perfect, thank you. Before we get started though, you have volunteered to listen to a joke and give me a rating. Call it the free joke project still, but you'll give me a rating one through five. So here we go. I heard they made the world's strongest suction cup. I'm not quite sure how they pulled it off. Udayan Bose (02:06.917) Absolutely. You would love that. Udayan Bose (02:17.447) I would actually give you a 4 out of 5 Brent. He'll give you a 4 out of 5. Awesome. Brent Peterson (02:22.899) All right. Thank you. That's great. Thank you very much. great. yeah. So, you know, I, I do, everybody looks at GPT and they think of generative AI and they don't always think about the analytical portion and how it can help. So give us an idea of how you're going beyond just generative AI and helping people with more, more of the analytics portion and more of the predictive portions that people can use AI. Udayan Bose (02:53.575) Yeah, so I think, I mean, we started building our machine learning platform, Brent, way back in 2017. And we wanted to tackle an important problem, which we have experienced firsthand, working with over 200 retail and direct to consumer e -commerce companies. And the problem was a small percentage of the customers drive a big chunk of revenues. And we wanted to combine machine learning and predictive analytics to see if we can predict which of the customers are likely to be a higher value customer, right? I who's likely to spend anywhere between three times to 10 times more. than the average customer with your website. We were able to pull it off around just around COVID times. And then we started building some sub -algorithms as well. For example, now our machine learning can actually predict with almost a 92 % accuracy, which of your customers are likely to churn, which of the customers are likely to come back in the next two months and buy again, what is the estimated purchase value going to be in the next 12 months, et cetera. So that's, I think, the first part, entirely driven driven by machine learning algorithms and works amazing with first party data. So we have first party data integration with Shopify, BigCommerce and so on. So that's something the first part. Gen .ai really sort of got into, think swept the entire world off just post launch. I think it was roughly around end of 2022. ChatGPT 3 at that point in time was launched on 30th November, 22. And we initially... had to just sort of soak it in as to what does it even mean. We are extremely pro technology. So any new technology, are probably one of the first ones, if not the first ones to dive in and just do some testing and playing around. And we quickly figured out by the month of about Feb that this can potentially transform work. When I say transform work, we said that this has the opportunity of really, really sort of partnering with humans and taking on a lot of the work which Udayan Bose (04:59.643) can be done by machines a lot quicker, thereby releasing a lot of time, a lot of time which gets freed up or opened up for the humans to think about other creative projects which pertains to their business growth and so on and so forth. So we started a pretty audacious project, Brent, and the project was called Elixir GPT. We wanted to build our security layer on top of GPT, or chat GPT essentially, overall platform. So use their APIs, build our own security layer. Security was important because we knew that anything that you put into ChatGPT is available in open internet. And understandably, as an agency, we worked with a lot of very confidential information pertaining to advertising campaigns on Google, Meta, TikTok, and so on. And we didn't really want any of that. So we had to spend about seven months building our security layer on top of that just to ensure that none of this information gets really leaked out. shared out and so on and so forth. And then on top of that, we started to really map out all of our workflows. When I say mapping out workflows, we were talking about, let us say you do SEO optimization for a Shopify customer. So there are certain steps. It's a sequence of steps that you normally take. And we said that what are the different tasks involved through this entire sequence? So let us say there are about 15 tasks which are there in this entire sequence. to the point that you really get to the first round of optimization of SEO. And we said that which of these tasks can be also done by generative AI, much quicker and much better as well. And we identified, think, across the processes, whether it is paid search management, SEO, paid social, and so on, we identified initially about 25 % of all the tasks which would be done by Gen .AI. So we started building those tools, or very specifically, we use those as problem cases or use cases, and started building essentially very specific Gen .ai tools or widgets effectively, which sat on top of our, what we call LXR GPT, our secure layer GPT. And that was remarkably successful. That actually, by the end of last year, we had been able to complete the project. The project was completed around the end of October. Udayan Bose (07:19.271) And then just during the holiday season, we were able to test it out at high volume because we worked with a lot of holiday customers and we saw almost a 20 % lift in the overall productivity. the team had 20 % of their time freed, which they could really engage in other things. we said that why, I mean, at beginning of this year, we said that why don't we start to release some of these tools to people for free, let them test it out. We knew that these tools worked because we have been using them for now almost a year now, eight, nine months now. And we said that why not really give it out? So effectively, we launched essentially last Friday. And you are the first time we are just, this is publicly been announced that we are going ahead and sharing our first lot of. first three tools essentially, and the platform is called LXR smart .ai. I can probably sort of type it essentially as well. It is LXR smart .ai brand. So that's the part which effectively I would really want to sort of offer. And it has just three very simple tools. One of the tools in LXR smart essentially helps you just write meta titles for your SEO rankings and just cleans up your description. typically writing a meta title and specifically with bulk often takes some time, probably about 30 minutes to 45 minutes. And in this case, you could really get all of these things done within seconds. The second tool that we have is more of an image alt text recommended, again, a very geeky sounding, a very tech SEO sounding one. And this is where we found some tremendous success, because even if you had used chat GPT, you probably can upload one image at a time, but we really allow you bulk upload of about 10 images. And for all of them, you are automatically getting this Udayan Bose (09:16.051) image all tests recommendations sort of there. And through one click, you can really append them as well. And the last one has been incredibly useful as well because we've managed literally a few hundred of millions of dollars essentially in terms of Google and Microsoft ad spend annually. And ad copy and testing ad copy is a big thing. And this really enabled us to really use GenAI to go ahead and write those in a very fine -tuned customized manner and really in a format which suits DSA, RSAs and so on and so forth. So have been extremely helpful and extremely useful and we are seeing collectively, typically, let us say for a company which probably has randomly, maybe about 1000 keywords that it is advertising on SEO and paid search, typically we believe that the time saving per session of use can be almost about three hours. So we figured that if folks are using this tool in a week, can, let us say you do it even once in a month, probably you have about 12 to 15 hours of saving, which is not that which can really be utilized in a much better other areas as well. So that's the idea. We are extremely passionate about this brand. How can we really truly utilize AI to really transform as to how advertising is done? That's the Alexa Insights tool that we have. And this one is our newest, essentially very fresh launch just end of last week, which is Alexa Smart. How can we really help you do more with less? Brent Peterson (10:53.085) Yeah, I think it's a, it's a really good way to look at the productivity boost. And I, I think a lot of the tools don't recognize or they don't advertise that you will save all this time around what you're trying to do. And maybe some bosses may think, you know, I, I, I'm introducing this thing that lets my people be more lazy rather than more productive. why do you think that? Udayan Bose (11:07.366) Mm Brent Peterson (11:21.865) a lot of e -commerce platforms haven't already done this. This just seems like low hanging fruit. And I specifically think about Adobe commerce and how they have Adobe sensei that they released way back in 2018 that they still haven't utilized or markets marketized that they're still just doing Firefly, which is nothing to do with, with the, with the generative, or not without, with the e -commerce portion of it. Why, why do you think they're just so lagging behind? Udayan Bose (11:35.686) Yeah. late. Udayan Bose (11:52.337) So that's a great comment actually, In fact, there was a Goldman Sachs report which came out in the last month. I can probably share with you the link of the report as well, which said that the adoption of GenAI tools has apparently dropped dramatically because there has not been any measurable impact in terms of, or probably even applicable areas have not been identified and the impact has not been measured. It's an interesting paper which I do not agree with, primarily because that's where it's more of a de -validation for our side. So some of the points though mentioned in the paper as to why there is a bit of a slow adoption in some of these tools. I think they talk about three things broadly, Brent, which prevents the adoption of any such cutting edge technology or new technology. I would say transformative technology in this case. The first one is just a level of discomfort that humans have initially adopting anything which potentially may have an impact on how they have worked all the time, right? So I have been used to working in such a way and suddenly there comes a tool, so I really have to change how I really sort of work and apply this, right? So where does the tool, what role does the tool play in my day -to -day work life is a very important component here, right? That's think one of the things that has been called out. The second thing which has been called out is many cases, I think the productivity is a collection of multiple tasks and incremental improvements per task. And if it is not done on a continued basis, it's difficult for what is called the aggregation of marginal gains or incremental gains to add up. So it really requires it's just grows and builds over the time. Just a simple example through the Alexa smart example that I gave this free tools. Typically it's about a. Maybe once a week you will do these tasks. You save about three hours per week. So in a month, you are still probably saving about 12 to 15 hours. But when you really do it every month, it really adds up. It gets to 150 hours suddenly in a month. So that's where the only thing is the continuity is the second part. The continuity sometimes does not get prompted because the immediate session -wise gains may be smaller. Udayan Bose (14:14.585) So it's just maybe marginal gains or incremental gains. And most people are not willing to wait or don't have the patience to wait all through to really see the aggregation happening. So it's almost a bit of an atomic habit playing out in a different way. And the third part I would think is just. I would, for a lack of better words, say a little bit of a fear as well. A little bit of a fear in the sense that will this task going away really dent the belief that I am really useful for the company or the organization? Will it take away a certain part which really legitimizes my role in the company? Now, most people, I think, are still struggling to shift from this adversarial relationship with AI to more of a collaborative relationship. And it will take time because it's at the end of it, it's a big, big, big step forward. And at some point in time, we really have to acknowledge the fact that the future is going to look like this where humans and AI really collaborate. I mean, that's going to be the case if not today, then tomorrow. So the faster you sort of get onto the bandwagon, the more successful you likely are to be. And those are, think, the three reasons Brent that I would call out, which has typically really seen almost like this curve, which is called gradually things happen. And then it suddenly happens. the more, let us say, future forward companies get into the gradual phase and the others essentially follow in the sudden phase when the entire leap is happening and that not much of a competitive advantage can be gained then. Brent Peterson (16:00.047) What do you think right now the models are increasing in their scope and their usability so quickly that it changes often week over week and now there's this new zero one, there's a new GBT out already that's available for enterprise clients and it's making sometimes, well sometimes the return results. Udayan Bose (16:12.473) Right. Udayan Bose (16:17.104) Yeah. Udayan Bose (16:21.543) Yeah. Brent Peterson (16:26.545) are vastly different week over week, even on the exact same prompt. Do you think there's gonna be a level off at some point that allows people like you that are using the APIs to be more determinant of what you get? And the second part of that question is how are you ensuring that you are getting consistent data over time right now? Udayan Bose (16:28.123) Mm -hmm. Mm -hmm. Udayan Bose (16:41.56) Yeah. Udayan Bose (16:48.443) Yeah, absolutely. So I really look at it from the user perspective. Right. So for me, I mean, 4 .0 is good enough, right? So do I really need anything more for the technology and the purpose that we have? So I really start from the goals that I'm trying to accomplish essentially, or what my, what the purpose of my using JNAI is. Now, will O1 sort of give me an advantage through reasoning? I'm probably, but I think I am good for now. So that's where it is a question of starting with the goal and really seeing as to. understanding that the the gen AI is a tool for you to really reach that goal right and that's I think the first part which I really wanted to emphasize. In terms of the model the recent one the O1 essentially apparently has some reasoning capabilities as well so this is the first time it is becoming oddly and scarily human like right so reasoning capabilities and all that stuff. The only question is I am at this point in time not very comfortable that the reasoning be done by AI, primarily because then since it is accessing open data. So that's where I think everyone, so the point that I'm trying to make is everyone will have their own cutoffs. And I think the other part which cannot be ignored is the cost of GenAI as well. So there was this brilliant piece which came out from the Wharton Professor, Ethan Malik. I think it sort of just came out last Monday. And he talks about that maybe at some point in time, we should think about GPT -5 being the level of factor. Udayan Bose (18:32.807) Do we need anything more or anything more sophisticated? Or let the application areas really catch up to five. I think that's what I'm really thinking of as well. For us at this point in time, for the purposes for, let us say things like digital marketing purposes, we are pretty well set with 4 .0. And understandably, I think that's where we are trying to sort of get to as well. Brent Peterson (18:59.817) We talked a little bit about the zero -click searches for Google and how Google is changing the whole search landscape. Talk a little bit about that. Udayan Bose (19:07.697) Mm So this has been an interesting one. think, again, I get to the point, Brent, such game -changing technologies like Gen. AI, things happen gradually and then suddenly. So we are very clearly sort of getting into a situation where we see things happening suddenly. When I talk about suddenly, we are talking about essentially a situation where, I mean, let us say, AI overviews once it was launched, we saw that essentially, I think, the number of clicks which happened on the Google SEO links dropped. Now, that was an assumption. There was some amount of talk about this concept called zero -click SERPs or zero -click search engine result pages. Now, what really happened with that print was We wanted to really, I mean, as a company, we are constantly for the survey and the research part. So we wanted to test it out ourselves. So we commissioned an interesting study, which handpicked about 1 ,000 keywords from about five retail e -commerce categories. And we wanted to compare what is the number of clicks that the SEO listing is getting pre and post AI overviews. We have all of this data pre AI overviews launch, and we started collecting the data post AI overviews launch as well. I was quite amazed to see that after the AI overviews launched, I think at every stage, the navigational search queries, the transactional search queries, informational search queries, there was a drop across the board. Udayan Bose (20:46.235) Very specifically on the informational search queries, the click through rates actually diminished quite a bit. Now, if you really sort of think of it, informational search queries mostly are questions, right? I'm asking something and I'm sort of asking a question, trying to find something out. And most likely, mean, the AI overviews is answering that question quite adequately, right? It may not give you a huge amount of range, but then when you compare the benefits of time saving, time saving from you're not having to click into different links and different sites versus the AI overviews giving this. I think I personally really prefer time saving a lot more compared to because I'm getting information which is good enough for me. So I think that good enough component I think is an important component. So, what we saw was very specifically in the navigational search queries and informational search queries there was a pretty big dent. Surprisingly transactional search queries, I had not really expected the clicks dropping as much, but in the transactional search queries also there was a fair amount of drop in clicks as well. Again, I can sort of share, we have created a one page PDF as well. So I can really sort of attach it for your customers or your viewers as well. So that was the interesting part. So the question, mean, two questions are ramifications of this. One, this is still the initial phase of AI overviews, right? The gen AI sort of thing and so on. So my first question I think has been discussed about for a long time. And the question is, is the future of search a chatbot search? or a chatbot led search? So that's, I think, the first question. So far, based on the results, we cannot completely ignore the fact that the search as a function may actually be replaced by chatting as a function in this case, A chatbot led sort of a functionality. That's a massive one. The second one, which has a more immediate implication, and specifically, of your viewers may be sort of Shopify e -commerce companies or big commerce e -commerce companies and so on. Udayan Bose (22:48.423) Google accounts for still over 50 % of all the traffic driven to your site. Let us say unique number, unique visitors to your site. If that number suddenly starts falling, the folks coming to your website, then you have a problem, right? Because half of your traffic comes from Google. If that number, that traffic keeps falling or continues to fall, during holiday season, especially, what options do you have? How will you sort of balance out the fact that this decline also directly is proportional to maybe potential loss in the number of conversions as well or orders as well, and as a result of loss of order revenue also, right? So those are, think, the two things. One is a slightly long -term strategic ramification, which is more of is search, is the future of search driven by chatbots or search bots, let us say. part effectively is much more immediate right I mean 50 % of the traffic plus essentially comes from Google or driven by Google the unique visitors I mean what if that drops then what happens right so effectively how do you really make up for the direct to consumer that the website traffic and the conversions there Brent Peterson (23:57.149) Yeah, I mean, that's a problem for Google as well, right? Because Google makes their money. And if you're suddenly dropping, you're going to start looking for other solutions and, you know, picking up other, there's a myriad of social platforms now that are, that are taking that traffic without even having to send it to your site. So you may search on TikTok and you may buy directly on TikTok. And that, that's a, that's a direct competition with Google that they don't even see. What else do you see in, Udayan Bose (24:11.644) Yeah. Udayan Bose (24:15.055) Exactly. Exactly. Brent Peterson (24:26.087) that's coming up then in terms of how people are going to use search and how Google, well, I think one interesting thing is the flip -flop Google has done on content. Originally, they said they're going to penalize AI content. Then they came out with Gemini, or they really pushed Gemini, and now they're kind of flip -flopping, right? Where do you think that's going to go? Udayan Bose (24:34.031) Mm -hmm. Yeah. Yeah. Right. Right, right. Udayan Bose (24:50.439) It's an interesting one, Brent, because again, this is purely my candid personal opinion. So I am sort of making a pretty big comment out here. I think Google is yet to fully recover from that shock that they got when ChatGPT was launched end of 2022. And they have not still fully recovered. They essentially are, I would believe, still playing the catch -up game in terms of JNAI, which I think Chad GPT very clearly is leading this game by a wide margin. And on top of that, you correctly mentioned that there are multiple other forces which are in play as well. In the last one and a half, two years, and we recently did our holiday forecast. The marketplace searches have really spiraled up. They have really gone up. So a lot of people are really searching on the marketplaces. Google already had lost about, I think, 65%, 70 % of the searches, product searches happening to, let us say, Amazon and something. So I think that is continuing to go up. For example, the other example is like the TikTok and Gen Z very clearly. So within the walled garden, the search and shop behavior is going up. Now, if you think of it, let us say you are running Google. you are having to deal with multiple such mega forces. One is the Gen .AI force, which really forces you to get into AI overviews, which you knew from the beginning would actually sort of, to an extent, cannibalize the advertising opportunities that you have, which is probably going to happen at this point in time. The second part is you are seeing a fair amount of traffic really move to the walled gardens, just because of the simplicity and the walled gardens like the TikToks and the Facebooks of the world, et cetera, and Amazon, and sort of getting a back share. And then on top of this comes the third big blow, which is all these antitrust rulings, which the Google is sort of having to deal with. So I don't really envy them at this point in time, honestly. So when they had started off, and this is purely my sort of opinion on Udayan Bose (26:54.313) this thing. Initially, when they started off, had the right thing in their mind that we will not really use generative AI -led content. I think they underestimated the power of GenAI and the usage and the adoption of GenAI. mean, recently, CatGPD came out with a number that they have 200 million active users now. Right? So if Suddenly, the force has become so big that even a giant like Google essentially has to flip -flop and sort of go back on what it really sort of came up with a plan. So that's the part, I think, Brent, which we are going to are seeing very clearly from Google. And now this ad tech sort of antitrust rulings which are just going on, this case will, this trial will, I think, come to an end roughly around the middle of October. 9 September it started, around the middle of October it'll end. And we'll get to know as to what happens. The previous one, which happened on Google search, unfortunately did not go in their favor, which has an implication as well. So it's a difficult situation. It's almost like you have this elephant, and there are multiple areas where they are getting poked by different competitors. And these competitors are not just companies. These are mega forces like generative AI. And that's what I think it remains to be seen as to how exactly they deal with it. Three questions that I would sort of put, I think the first part, and maybe this is, I wouldn't say as much of a question, I'll just probably make a comment initially, Brent. And this is just to probably again, maybe strengthen the point that I mentioned initially. I said that generative AI is such a force at this point in time, you may dislike, you may hate gen AI. But you don't have an option but to really collaborate with GenAI in the future. And Google's flip -flopping on the content is just exactly the validation, which I think my statement probably gets directly there as such. Because when a company of the size of Google and the clout of Google really sort of flip -flops on this entire their content policy, then I think any other businesses Udayan Bose (29:05.297) too small and too trivial to really sort of go ahead and just try to fight and sort of really stop and say that, no, we will not really write the JNI ban. So that's, think, a message which I wanted to pass. But there are three things which I think Google now is sort getting into the question. The first question is, how do they monetize all of this lost opportunity? We are here, right? I didn't want to mention it, sort of, that he sort of brought it up directly. But yeah, mean, people are not getting clicks, but you are losing money. So how exactly will Google really make up for this lost revenue? What is the monetization? There is currently no monetization paths. The search engine, mean the generative AI chat engine, perplexity, tried to come up with some options. I think it's just more of a pre, I would say pre -alpha or beta stage, whatever it is. I don't think it has been proven. I personally don't think it really has too much weight. So what is the monetization model? Because without monetization model, understandably, the machine cannot really sort of keep going. The second is a fundamental question, which I hate to bring up, but I have to bring up, which is essentially, again, it has flip -flopped on its entire third -party cookie, first -party cookie policy as well, if you remember, So when we had met earlier in the year, the entire thing was the first -party cookies had the future and... Chrome is going to deprecate the third party cookies by the end of this year and all this stuff, right? So they went back on that decision a couple of months back, right? But at the same time, mean, Microsoft Edge is going to deprecate the cookies by the end of this year. Safari and Firefox already have sort of just gone to the first party sort of route. And I think the first party future is here and now. So Google can keep on relying on the fact that Chrome has a big clout. But honestly speaking, because if all of the other really almost like gang up essentially on Google, which very clearly is happening now, then I think it has to really be ready for a first party future. Now, if the first party future is there, we have seen Google's struggles. I'm pretty sure Brent, you would have seen Google's with whether it is introducing an AI module like Performance Max or let us say struggling with the GA4 and Udayan Bose (31:23.163) how sort of GA4 adoption has been much, much worse compared to what Google had expected initially and they have really extended that. So, there have been multiple forces there as well. So, that becomes another problem. They just have tried to postpone the problem or the issue, but it is coming. I think it's there for all practical purposes. How does Google really reinvent itself for a first party world? I think is the next question. And the third question, these are all questions for Google. I hope someone from Google listens to this. And the third fundamental question is, how does Google utilize its various assets, it still has the digital assets, to build a stronger business? And I think there is an opportunity. I think there is a massive opportunity in YouTube, which is significantly under -tapped still, primarily because of multiple reasons, whatever it is. I think there is a massive opportunity and many of their other assets as well. So what does that new Google, I don't know, 2 .0, 3 .0, 5 .0, whatever it is, strategy looks like, which helps it to bounce back, factoring in that the first two things are happening, right? Whether Google likes it or not, there is a monetization impact. Whether Google likes it or not, the first party world is here, right? Knowing that, if you were the CEO of Google, How would you really rethink your entire strategy? So pretty difficult and hard hitting questions directly for Google as such. Brent Peterson (32:54.097) Yeah, such interesting concepts and there's so much more we could talk about. So we have a few minutes left. It is lxrsmart .ai is your new tool and we'll make sure we get that in the show notes and I really appreciate you talking about that. And we'll have to schedule a call in another six months to see where you, what's changed and where your predictions are at. Udayan Bose (32:59.815) Good night. Udayan Bose (33:07.443) Yep. Awesome. Thank you. Udayan Bose (33:18.375) Absolutely, how much productivity improvement? Sure, 100%. I would love to do that. Brent Peterson (33:24.697) Odeanna, as we, as we close out, I give everybody a chance to do a shameless plug about anything you want. What would you like to promote or plug today? Udayan Bose (33:33.285) Well, I think more than promotion, I speak probably on the behalf of an AI first agency brand, which has been remarkably successful embracing AI and rethinking as to how marketing work in this new AI plus world really sort of looks like. And I call it AI plus world, primarily because like humans plus AI, think have a significant advantage over humans without AI. So my shameless plug would be for the viewers of this show to test drive these free tools in LXR smart .ai. And we would really appreciate if you can send me the feedback directly. mean, my email address, you already know I can sort of put my email address there as well. I am very open to your feedback. And there is a big reason why I'm sort of passing out my email address directly, Primarily because we are at a very, very initial stage in this humans plus AI revolution, which I think is happening. And I really want to activate that entire ecosystem because this is a much bigger force than one or two companies or individual resolve. But I think collectively, if we all join forces as marketers and we really exchange ideas freely and openly, and we really go into this with an open mindset and embrace AI. I think all of us will come out tremendous winners at the end of it. So that would it would be as much of a shameless plug for embracing Gen .ai tools and technologies as obviously I would hope that you use our tools Alexa, smart .ai and share directly your feedback with me as such. But thanks for the opportunity. Brent Peterson (35:15.037) Perfect, I'll make sure we get all that into the show notes. it's lxrsmart .ai is the URL. And I would encourage everybody to go there and try it out. Udian, it's been such a great pleasure to speak to you today. Udian Bose is the CEO of NetElixir. Thank you. Udayan Bose (35:30.413) Absolutely a pleasure. Udayan Bose (35:37.169) Thank you very much, friend. Such an honor.