Latest / Growth After Headcount / Can One CMO with AI Agents Do the Work of 10 Marketers? (FRAME) #014

Transcript

Can One CMO with AI Agents Do the Work of 10 Marketers? (FRAME) #014 · Growth After Headcount

• Haris: Banu, welcome to Growth After HeadCon. today we're gonna explore are your journey into marketing, but then also how that journey now changed through AI and how you're working nowadays with AI and marketing. but before we go into that, You actually started blogging as a teenager in Delhi and what draw you into that? Bhanu Chawla: Yeah, yeah. so thanks for having me, Harris. I really appreciate that. And yeah, it's it's nice talking to you. But yeah, I mean essentially I'm from New Delhi, India. and that's where I grew up. And I came to the UK like sixteen, seventeen years ago. And basically when I was a teenager I got into ethical hacking and penetration testing. I don't how. I mean just maybe just accessing different kind of information on the internet when I grew up. It's like, you could hack into things and stuff like that. So I used to use like This Kali Linux Backtrack OS. it's an operating system, a Linux operating system that, ethical hackers or penetration testers or not so ethical hackers use as well. And Haris: Mm-hmm. Bhanu Chawla: basically, yeah, I mean I started blogging about, you know, how to do stuff, and how to use that operating system and you know, access vulnerabilities, you know, and fix them and that kind of stuff. And Blockspot used to be the the choice of blogging platform back in the day. And basically started blogging about that. had a good subscriber base, I think more than 50,000 subscribers on that blog. Haris: wow, it's big. Bhanu Chawla: Yeah. And then essentially I realized you can monetize that with Google Adsens. so that was quite new back then as well. And then I started doing ads on my blog, and then I realized you could earn money with this. And when my first check came, my my parents thought it was a scam. Like, who's sending you a check named by Google from the US? Like, no, it's not, you can cash it out and that kind of stuff. So basically, I think the inflection point in in my life, getting into marketing was just the idea of growing that personal blog. and SEO was something that I learned in the beginning. and it was quite new back then, you know, when Moz was just coming around, even after that, it was around Fishkin and that kind of stuff. So yeah, basically just wanted to grow my own blog and the rest is history and you know, I can talk more details about, you know, like what I did afterwards. But that was like the first thing that I started doing. When I seventeen or something. Haris: Okay, that's very very cool. Also at this age and to reaching that level and also considering that was sixteen, seventeen years ago, I guess, fifty thousand followers was probably equal to half a million nowadays, right? So like the internet population grew by quite Bhanu Chawla: Probably. Probably. Yeah. Haris: a lot. And okay. And then afterwards what you continued your marketing journey? Bhanu Chawla: Yeah, so basically in India I did my bachelor's degree in computer science. So engineering kind of that's my background. and essentially after straight after that I came to the UK, Kingston University, did my master's degree in information systems and management studies and that kind stuff. and did my dissertation on SEO again because I was getting good at it and I showed a real use case of my like growing my blog 3x in organic traffic in like s less than six months and got a distinction in that. So that was pretty well. and straight after that, I got into SaaS businesses. So B2B SaaS and started managing teams for web developers, digital marketing, and then slowly took into demand generation, then senior leadership roles, and then today, yeah, I mean last sixteen or or seventeen years grown into a role that's fully marketing. today I'm a CMO for a company called Frame, which is a financial infrastructure interoperability company. And yeah, that's that's my kind of like trajectory on a helicopter view. Haris: All right, great. And now there you're the CMO. How does your day to day look nowadays? Bhanu Chawla: everything we do. And it's it's not just me, it's the full I mean the whole leadership team at frame. we are AI pilled, as they say, these days. and yeah, it's essentially it's a part of everything we do. It helps us accelerate processes faster. get shit done quicker as well, much, much quicker. I'd say more than 2025 ex productivity gains from it. so yeah, I mean in terms of our models, we use multiple, like cloud is our major one. And you know, most of most of us use it in terminal or in in the app itself, but cloud code is is is our choice. And then we also experiment with other things like, you know, codex and open rate models and that kind of stuff. But essentially our marketing and sales and product engineering operations, it's all boosted with AI efforts. from a marketing perspective, everything we do, I mean, to give you a a a perspective of like my team size, it's just me right now. in my last company it was like fifty five people in marketing. So you can imagine like how that's shrunk into just, you know, a lot of different functions. and from brand and design, content, not like AI slob, but like really humanizing content and taking our time to review stuff and not just putting anything that AI publishes out there. That's done by AI, performance and paid marketing done, you know, in with the help of AI, ABM campaigns, personalization, ABX, that kind of stuff, social content, video content, motion graphics, editing, marketing operations, analytics. It's you can just get me started and then I can dig deeper into every single function of it and talk about how we use it, but we use it extensively. Haris: Yeah, it sounds very extensively. And you also mentioned your all of the other co founders and the whole team are also very heavy on it. How do you collaborate so that everything is aligned in the same database or knowledge that everyone that doesn't go in different directions? Bhanu Chawla: Yeah, so well, we go with the concept of second brain, as many are going these days. So having a knowledge graph of different nodes essentially structured in a way that that helps us organize it in just one repository, you can say. and then we can connect you know, model agnostically to to to that particular knowledge graph and say like, Hey, I wanna talk about this project today and it would just grab everything that you know we've been working on collaboratively, in our own terminal instance or cloud instances and just work on the same things. Haris: Okay. Do you have preferred tech setup Bhanu Chawla: Yeah, so so our tech stack looks like we use Cloud for AI. we have Notion, that's you know, our kind of like a mini second brain you can say, but more from a project management perspective, but it's all connected with MCPs. then we have Slack, and we are working on Google Workspace, so everything's connected from that. From a CRM, we're using HubSpot today. and as I said, everything's connected with MCP. From a GTM perspective, we started using this platform called Swan, getSwan.com. That's a shout out to them. from an outbound perspective, that works really well for us. and what else? I mean, yeah, there's many more, but that's like the basic of like, you know, what we use day in and day Haris: No, I think gets one Amos. he was also I think the first episode of the podcast. Yeah, yeah. That's Bhanu Chawla: Yeah. really? That's amazing. Yeah. I so basically how I discovered them is like I was connected to ammo's and probably some other some some other person from the from that team and I realized from the content, organic content, maybe fund content. So this says that fund content works. I I I discovered them and started to try them and I'm an early adopter, so I would just like go in, sign up, try it out and see if it works. And it actually works really, really well. I mean like we created like lots of pipeline from that. So the ROI is massive from us. So really, really good platform. And their next iteration from a product perspective as well is amazing. So yeah, that's that's what we use. Plus, of course, from a con from a tech stack perspective, like if you if you get into models, some other models outside of just you know text and code for image and video generation. So we use Higgs field essentially to Haris: Yeah. Bhanu Chawla: have access to everything in terms of like image generation models like you know Dana Burana Pro or G B D Image two point or from a video generation model with these days we're using quite a lot of C dance two point five because that's super realistic, really high quality and they just announce four K outputs for enterprises. So that's getting really heavy users as well. Haris: No, yeah, yeah, the video space is just getting insane as well. whatever I'm just playing Bhanu Chawla: Super insane. Haris: a bit sometimes, and I'm like, wow, this is wild. So before we zoom in more into detail, the marketing, I have a category which is called human or agent. So basically I'm gonna give five ideas, topics in marketing, and you can then say, okay, a human should do it or an agent should do it, and then we move to the list and then we can discuss it. All Bhanu Chawla: Okay, cool. Haris: right. choosing the central idea for a new marketing campaign. Bhanu Chawla: Mm-hmm. A human should do it. Haris: Okay. Interview Bhanu Chawla: We should e we should I D it from AI but not not just leave the decision on AI. Sorry, I'm it's a long ratzer. Yeah. Yeah. Haris: Yeah, the decision making. No, no, it makes sense. And interviewing a customer for a case study. Bhanu Chawla: of course human. Haris: Yeah. Writing a public apology after a brand mistake. Bhanu Chawla: Absolutely human. Haris: Okay, adopt a campaign for a different language and culture. Bhanu Chawla: super led by AI with human in loop. Haris: All right. pausing an advertising advertisement campaign that is underperforming. Bhanu Chawla: That is underperforming, I'd say these days I'm actually relying quite a lot on AI for that as well. and then human and loop at the end. Haris: Okay, good. All right. So we had three times human and two times AI, right? So the last one even. Yeah yeah yeah yeah. And you bet Bhanu Chawla: Yeah, you could say that. Yeah. th there there will be some AI used even in the human processes, but then I I was like going like majority of that is human versus AI or the other way around. Yeah. Haris: Exactly. Ex sometimes it's also not a black and white answer, right? So it's a little bit newer. Bhanu Chawla: Yeah. These days it's it's it's very gray. Yeah. Haris: No. I think the first one choosing the central idea or theme for a new campaign, you instantly said human. why was that? Bhanu Chawla: because I think central ideas, like that's we call it umbrella campaign or whatever that is. Like that's that's crucial for the company, right? you just don't want AI is getting really good at ideation, especially when it has Haris: Mm-hmm. Bhanu Chawla: access to all of the context and memory from your previous work and that kind of stuff. So let's say if I'm going into twenty twenty seven and I'm I want a new campaign idea, that's like the central idea of the company for that year. I can ideate with AI, but I don't wanna just ch let AI choose and go full blown, do full blown work without me. Right. So I I would wanna I like I would wanna give it ideas from like, hey, this is what I'm thinking. So I would have a lot of back and forth, you know, have a sparring session with my AI and some like, you know, correct me if I'm wrong and I'll correct you. And we'll we'll we'll go into discussions like like I would do with another human. and then I'll actually ha involve other humans for the subjective opinions and objective opinions. And then there's gonna be a lot of different ideations from that, and then I would run scenarios in terms of like if I were to choose this idea, this is how the execution would look like, this is how a a good, you know, campaign looks like at the end of the day. And I would do that for actually more than just one idea. And then I would just choose personally in terms of like, you know, d with the discussion with the leadership team, like what makes sense from a company level if you're going to d you know, certain direction from from from a campaign perspective. Yeah. Haris: and ha does it have to do with the quality output of the AI or is it more like okay, Bhanu Chawla: Mm. Haris: this is a too big decision, we don't Bhanu Chawla: I think a mix of both, because AI can do quality stuff, especially like cloud models. And when it comes to content and campaigns, it's better than anything out there. I've tried, you know, Astra and everything else as well, but still I think clothes like ops five point five would be my choice. but then from a complete campaign perspective, as I said, maybe a discussion and ideation from a very bare min minimum perspective is fine. but it's it's it's both, you know, too big and I can't just rely on quality on AI, for that particular thing. Haris: Do you think AI has taste? Bhanu Chawla: it's getting there, not yet. I mean, everyone's trying to replicate taste in AI, especially, you know, Silicon Valley and our friends over there. so I hear I hear about all sorts of different things of like, you know, how even biologically, like if you if you if you were to think about it, like, you know, it comes to my mind as well. Like everything is hormonal and emo even including your emotions and your DNA and like how your brain reacts from brain chemicals and neurons and how they, you know, kinda like work. at the end of the day, when you see something great or create something great, right? So if you try to replicate that from a perspective hey, take these thousand iterations of different ideas and these humans and how their brain and hormones operate, you know, operated, then they had that, you know, f you're like big movement of ideation or creation, and try to replicate that with AI, sure, maybe there's some merit to that and that can get to a stage where it would start to have a lot a lot of taste. Maybe it already has. Maybe I don't have access to to those things right now. But I I think I think humans still do taste really, really well. like you and I are talking right now. You are having different sort of thoughts in your mind in terms of like, hey, what's what's my next question's gonna be, what's my next comment going to be? he said something about this that that actually resonates with me because of my ex like your experience in the past. And that would actually drive the question which is gonna come next. I don't think AI is at that level yet that it'll actually have that kind of taste. And it's not just that. not just with podcast, but even with design and it's very sub like subjective things, where it's hard to give a definitive goal and what really like good looks like to an AI, that's where it struggles. Haris: I see, yeah, yeah, yeah. No, it's a typical experience when it just asks sometimes so generic questions as a follow-up and follow up. Let me ask one more question, one more question. But it's not how you would maybe have a normal conversation with somebody, which sometimes can be random, but the randomness at some point even starts to making sense in a way that it's becoming a coherent conversation. Bhanu Chawla: Yeah, s same with music, right? when people take hip hop artists for example, they would go into studio, and they would just jam and like, you know, create some random sounds in in these machines and or talk about some random things or that particular sample that worked really great from a previous perspective and try to recreate that, but create something by mistake and that actually belongs to the main track that actually gets released because of of randomness, I think that's where it's kinda where humans kinda shine today. We're just discovering new things all the time from a perspective that's purely human taste. Haris: No and I think that's where our core strength is. And you mentioned earlier the output is almost twenty to twenty five X more. And do you also see the return of it? Bhanu Chawla: Absolutely. I mean we could talk about our friends at Swan, Samos. I mean what he's done, he's he's kinda like, you know, creating that second brain of GTM at the end of the day. going that from a pure outbound perspective and sales perspective. Yeah, I mean it's creating pipelines for us. This is one example. but then it's the same from a marketing perspective. I've actually created big piece big rock content with with with the help of AI, but then it what what it did is really it took me a long, long, long, long, long time to create a piece of content with AI. which because I think it should take you a long time. And I think Amos or someone else said that on LinkedIn that they have an internal AI process of like, hey, if it's taking you more time if it's taking to taking you more less time to review than it took you to create, then it's probably slop. so we are taking our own time, of course, in terms of like what goes out in the into the public. And I think that actually adds Haris: Mm-hmm. Bhanu Chawla: to c as to the quality of it because reviewing and optimizing and tweaking anything that's AI generated is is the way to go because that multiple iterations of human in the loop is what what we have been doing. And yeah, we're seeing pipeline, Haris: Yeah. Bhanu Chawla: we're seeing demand, we're seeing the right ICP from a marketing perspective, right ICP and persona coming to our website, raising their hands, downloading our assets, you know, showing us intent. helping we're helping them do their research faster. and they are as a consumer as well, people are getting their research done faster as well because of AI, rather than going into 20 pages of Google. they're just having a question and a follow up in an AI model and just getting that done. So how you from a market perspective, for example, how I show up in AEO and all of these things, it's kind of driven with AI as well. Like, you having glossary pages and FAQs and JSON schemas and that kind of stuff. Haris: No And was there maybe something that you tried to automate or a put towards the AI, but it's still not there yet? Bhanu Chawla: Let's say Haris: It's already a sign that you have to think a lot, meaning you already put a lot to AI and it worked. Bhanu Chawla: Yeah, that's true. Because if you think about the modalities today, what is it? So text, code, image, video, and audio. Right, these are the five kind of like major mod modalities of generative AI. Text, you know, Chat D P D cloud, that was like, you the GPD three movement when people had starting just doing that. First one. Second was like code, engineering. We've all seen that like everyone's a coder these days. That any anybody can create. then it's about image and video, I think image models are really good as well today. That's sorted. video as well with C dance 2.5 you know the new cling models and everything. That's also getting there. but I think, yeah, generating super realistic apart from just like let's just say it's a small trailer, you can create that with AI with a lot of lot of credits being used on Hicksfield. But then if you were to really create like a full blown, super realistic looking campaign that's like let's say a 10 minute podcast interview, if you want to create that. You can't Haris: Hm. Bhanu Chawla: do that. You can't generate that with AI today. but I that doesn't mean I wanna create that, but there's many more use cases that people are exploring over there as well. Audio, I mean, I don't use that much apart from just like, you know, audio generation models like, you know, eleven labs or speechmatics, but mostly eleven labs. So that's their and I'm not a musician, so I I don't use eleven music or Suno, that's another platform. But I think I think, yeah, I mean, it's it's getting there. I mean, at the end of the day, all of the knowledge work, believe it me or not, is fine it's starting to get replaced quite a lot. you can you can tell by the size of my team right now. You know, like if if it was th three, four years ago, before Opus F open five came out, I would have had a larger team, you know. But today, don't really need that. might be controversial, but that's the truth. Haris: It is the truth, right? It is called growth after headcount and it's a reality that we are living in and it's also not that people will not do it because of whatever reason, right? If it's it's always with technological progr progress, that it goes towards that direction. And talking about people, when would you start hiring somebody in the marketing team? Bhanu Chawla: from a scaling perspective it depends what I'm trying to maxing out myself. So basically I when I interviewed for this role, Alexander, our CEO and I had a really honest discussion about users of AI. It was like I I was like I use quite a lot of AI. It's like, yeah, we need someone like that actually. It's like cool, screwed. And then you know, it was my promise to him as well, like, hey, I I don't wanna hire and I wanna really wanna have a lean team at the moment because that's where I kinda like, you know, shine and also it can help the company more. and then I wouldn't want to hire until a Mac start with AI use it as well, or something that I can't hire, can't do work for. So for example, events, right? You can't really outsource all of that to AI. PR, sure. There can be outreach campaigns that you can do in a fully automated way with you know crockbot, which we use as well these days, or you know, chat dbd like dot. Sorry, I was just gonna name my it dot assistant, but yeah, so dot from ChatGPT or Instinct, whatever you can do automated always on a system that can just go out and do that. But then in the physical world, they can't really have manage an event that a human would. That's one area I would want to hire first. second would be maybe someone who can do that human outreach perspective from a PR PR point of view too. We have a PR agency, but at the end of the day I think that's somewhere I would want to hire as well. And the second and the third would be from a scaling perspective. I can only do so much at till a certain point. There's a need here and there that I can hire freelancers for. But then when it comes to a point where everything's becoming too much, which is a good problem to have as a company in terms of getting more Haris: No. Bhanu Chawla: customers and all that stuff and more marketing campaigns needed and expansion processes in different regions. I think that's where I'd I'd hire hire more as well. but then what's also changing is like probably your follow-up question, but answering that first in terms of like, you know, the the way I hire would be different in terms of like I'm I'm gonna hire someone who's like as AI pilled as I am or probably more, and the team so that they can actually create their own agents or take over agents that we have already created, like you know, hundreds of those and then just start managing them as well. Haris: Mm mm mm yeah. I think Amos is talking also of something similar, creating like the superhumans who can have twelve twenty, fifty, one hundred X productivity. And this brings me to my next question. Let's imagine such a superhuman or even you are leaving. Doesn't that also create a massive gap then for the company, or are the AI systems strong enough that they somehow can operate themselves? Bhanu Chawla: Yeah, so this is where you need policies and stu that stuff in terms of documentation and that cause the second brain kinda like at the end of the day it's all about data. So how you govern that data matters, right? So if I'm working on something day in and day out and that data is governed by the company and our CTO is is on the same page as well and CEO two and then they know that what I'm working on how I'm working on it and I'm quite honest and transparent about it. And, everything's been documented in a way that they can have access to it. but at the end of the day, yeah, I mean there's gonna be some gap left for sure. I'm I'm not saying Haris: Hmm. Bhanu Chawla: that, you know, can't just fully take over everything and expect similar output. Maybe it's better, maybe not, that that truly depends. But what the best you can do is just go on the data as a company. so, you know, have a central repository of like what everything's work everyone's working on. And when someone leaves, the other person can just easily just take take that over and plug in their agent and just start running. Haris: I think the that is the future. And do you still logging into tools or you do everything from a terminal or MCP? Bhanu Chawla: yeah, I mean both, but less platforms and dashboards now. Everything is terminal MCP I'd say these days. when you said tools, you mean like dashboards and UI, right? No, no, no, it's that's Haris: Exactly. do you log into HubSpot and stuff? Bhanu Chawla: probably one percent of the time now. and I was telling that to our chief strategy officers and salespeople the other day, is like I I don't use the AI UI anymore. with HubSpot, it's all connected to MCP. So it's got full permissions, like a private ad within HubSpot that I connected to. So it's got like full permissions on everything. Like even from a re reporting perspective. I don't like the HubSpot dash button reporting from a point of view that it's so, you know, you know you have to it's clunky, you have to customize quite a lot and you drag and drop all of that. What I've done is like literally taken all of that data and created local reporting that actually makes sense for me. And I could iterate that with one prompt. That's what I want to do. Same with updating deals, working on analysis of you know, like what's working, what's not, different kind of you know, fields I want to do analysis on. Everything is why MCP these days. And not just Top Spot, even Hicksfield is MCP. Swan for me is MCP. I log into the platform sometimes, but I think it's more MCP for me than the team. and Notion, MCP, I only open Notion for viewing stuff, not writing back to it. Well. reviewing, of course. yes, even my emails and everything, now it's all managed by assistants. So they proactively send me notification. I if it's a I have a two minute rule. If it's if it's done in two minutes, I'll do that right away. If not, it creates a to-do list for me. I I have my voice mode available almost twenty four seven on my computer. not when I'm talking to people, but so that I can just have a back and forth in in a in a way that makes more sense for me. And they were like, hey, an email just came out. Like, cool. Let's just open that up, reply to it in this way, send them this asset, upload to Google Sheets, or take data from HubSpot and do that. And it's all connected. At the end of the day, if you have a really connected ecosystem with MCP, you just do it without the user WIs. Haris: No. half a year ago I was I was researching should I get the private assistant? Now I almost have to laugh about that. And especially now you also mentioned Grogball and often with driving I'm like just talking to Grogbord and I Bhanu Chawla: Same. Haris: get so much work done in the car. It's like it's absurd just thinking of it. Like right? Bhanu Chawla: Super absurd. previously, but so I live just outside of London in Hampshire, right? and I'm in London also every week. So used to like, I mean still take the train because you know, driving sucks in terms of like getting into the city. but when I'm driving, I'm like literally opening my grogboard and just going into voice mode and like, okay, what's on my list this week? And what's on my list today? Okay, cool. Let's let's take that first action item. And it's like, okay, cool, let me pull that up. And it just pulls it up. It's like, okay, Alexander asked for this. And what do you want me to do about it? Or actually, more most of the time would suggest an action like I've I've already pulled in this HubSpot data for him. Do you want me to just send him a Slack message with this list of attendees from that event, with a new column added of all the actions that we took from an outburst perspective via small and hubspot, HubSpot campaigns? We're like, Yeah, just do that and then it's done. So imagine that. Getting into the training, opening the laptop, doing all of this manually, would have taken me one and a half hour. Or now in one and a half minutes, I'm getting that done with voice, proactively someone just doing that for me. It's like that minority report world that we're living in right now is crazy, man. And I I I personally love it. Haris: Yeah. In parallel also the ten other agents that just run in the background and doing their own thing, right? And I guess that's Bhanu Chawla: Yeah, yeah. Haris: how quickly we can get to this twenty, twenty five X per s percentage or a factor X increase. And I think for still so many people, like in our we are like I would say in the super tech bubble in the front of it, but for most people they don't realize that yet. Bhanu Chawla: Yeah, they don't because you when you when you so I got my wife on to GrockBot and she was blown away as well. And so I used to be that personal assistant for my wife quite a lot of the time. Tell us to hey, can you do this for me? Can you d get that sorted for me? The technical stuff. Now for what I've done is like all the repetitive processes. I've created like different bots for her in her app. Like, hey, if you want to generate that invoice, this is where you're going. If you want to generate that automation, we have replaced many chat into like Instagram automation platform. With a custom map that Grog built for her and then actually that's running in automation and chat and now she doesn't have to come to me with all of this stuff. I was like, well, it's not just saving me time, but my wife, but then I when she talks to other people about it in her group and everything, nobody knows about it yet. So it's it's it's I think Haris: Yeah, yeah, yeah. Bhanu Chawla: it's less than one percent of people who are using it at that level. Yeah. Haris: Yeah, yeah, yeah. I'd probably say even less. In the App Store it has one million downloads, that's nothing. And I think they have also now grog teams, so you basically can have a bot for you and your wife together and you both can access it, which is kinda also cool. maybe slowly coming to the end a few more questions. Is there like a marketing campaign with recently with the help of AI that you're like really proud of that you put this off like? Bhanu Chawla: yeah, there's there's been a few. the one we have at Frame, well, company I work for, essentially is that settlement blueprint campaign that you know I launched with the help of AI. Basically, I didn't want to spend too much time on having my own research being done with an analyst when I joined. So basically it said like, Hell, let's go out in the web Do a deep resort of you using fire crawl, and that was quicker for us and much deeper for us as well by by MCP. to do deep reset into X topic about, you know, settlement and you know the interoperability world that we are trying to solve, that we are solving already with frame. so we wanted to create like this blueprint top of the funnel campaign, and that you can access on frame.xyz slash blueprint. yeah, I mean, did a lot of iterations on that. And also, reviewed my the research personally, what it meant, what it didn't mean, which pieces of research we should take into account, which not, what's relevant to frame, what's not, what's relevant to our ICPs and persona, what's not, taking your time to review and create with AI, that was a good one for us. And that's actually generating a lot of eyeballs for us in terms of people accessing that via LinkedIn campaigns and organically. and then join that with our glossary on our website that's purely for AEO, like machine reading and Haris: No. Bhanu Chawla: that Top of the funnel AI sorry SEO perspective as well. That was AI driven as well. That was more so handoff AI driven to glossary writers, because I was like, this is very generic information that I just want to have on the website purely for machines. It's optimized for machines, that helps me get some topic authority on Google and AEOs, like sorry, answer engines. that was like purely A AI driven. I think these are the two kind of things that I'm like really proud of in terms of like, you know, that. that worked for us and were fast enough to do that myself as compared to, hiring a content agency or hiring a research agency, an analyst, which I'll do with time, for something much, much bigger. But, you know, this this was a campaign that actually had a, you know, good, good amount of success for us. Haris: I think also it relates into afterwards that the bar for agencies when you get them in is also much higher, right? Because all of the low hanging fruit or even medium complexes, you can do it yourself. Bhanu Chawla: Yeah, hundred percent. I totally agree. Haris: And we talked about the ninety nine percent who are maybe not there yet, and especially in marketing folks or marketing leaders, what advice would you give them to jump on the AI train? Where should they start first? Bhanu Chawla: experiment and try everything, man. Honestly. everything works. in terms of like, you know, different channels for marketing, traditionally speaking. But at the end of the day, you have to find what works for you. always have human in the loop. Don't just don't just generate something with AI and put it out there, no matter how good the models get. you still need that one percent taste that comes with you because that taste is gonna define the future success of that particular campaign. And also you wanna put your name on that campaign at the end of the day, right? So you wanna do it right. so Experiment, experiment, experiment, try new things. don't be shy about trying that new tool or new things. of course, I mean it's quite hard for like big enterprises where they have l you know, very slow processes in terms of like getting a Haris: And co pilot. Bhanu Chawla: I've speak to so many different organizations, they're like, man, I'm creating this presentation today with co pilot. It's like, man, it's twenty twenty six October. Like, come on. It's it's it's not we we that that's two years ago. You need to really go out there and try something new. But I don't blame them. It's it's also about the lack of knowledge of it's so like educate yourself. I think go out there, watch shit tons of YouTube videos or ask you guys this these questions, whatever you're interested in. If you can't get your company to get that tool, get that for your personal use. You know, really try it for yourself. If you wanna get grogbot for yourself, do that. Connect your Gmail and all of that. Don't just, you know, say like, I mean Of course, companies gonna have a long process of getting another AI tool that's owned by Elon Musk and people don't have really, hard trust issues with that still. but try it for personal. my wife can try that, you can try that. Don't connect your work email, get the hang of it. If it's good, I'm sure there's something an alternative that exists for your enterprise today as well. Even with chat GPT, dot, which is like the copy of, you know, Grogbot with just like single bot. It's only in the US and Canada and all that stuff. But I was in the US back then, when it came out, then I've been using it on VPN, you know, like it's probably not in the European Union or like UK today, but it's pretty good as well. Like, you know, really at that same level and it's coming coming to us as well. but yeah, that's my advice. Be open to experimenting anything new because if you don't do that, the cliche saying of AI will replace you will actually happen. Haris: Yeah. Those are rosy words to finish off. But no, also sometimes we people need a reality check, indeed. But thank you very much for joining. This was a great talk. And yeah, thank you very much. Bhanu Chawla: No, thank you so much for having me, I really enjoyed the conversation.