A16z Growth Expert On ChatGPT Ads, Impact of AI On SEO & More

On today's show, we're going deep on marketing and AI with one of the biggest marketing experts in the world, Andrew Chin. He's a partner over at a 16Z. He famously led growth at Uber for a long time and he has a must read newsletter if you're in the marketing growth space. We're going to dive into everything marketing, AI and the future with him. Let's get to today's show.

Welcome to the show, Andrew. Andrew, who is investing at the intersection of gaming and tech over at A16Z. We are really excited to have you on talk about AI and the change that it is going to cause for marketers, for growth people. So thanks for coming on the show. Amazing.

Thank you for having me. Always a huge fan of the podcast and excited to talk about all the topics.

Hey, Andrew, one thing we could start with is when did you start writing your original newsletter? The reason I'm asking is because your newsletter, when I was really just starting out in marketing, was the one that I would never miss. I would never like miss the edition of it. I always make sure I read it. But when did you actually start that originally?

Yeah, thank you for saying that. I mean, you know, the funny thing was I started to write an anonymous blog when I first moved to the bay area in 2007. So I was 25 years old, moved to the Bay Area. I was like, okay, I'm this, this will be a great way for me to stay in touch with a bunch of my friends in Seattle. That's originally where, where, where I'm from. So I forcibly subscribed my bunch of my close friends, my sister, my mom, you know, like, that was, that was my original blog audience. And I would just write about whatever I was learning. And what I found though was that I was learning so much about kind of this new discipline of user growth and marketing, kind of in a world of Web 2.0 and mobile and all that stuff that it started to pick up from there. But if you go back and look at my early posts, they were literally about everything and anything eclectic that I would. It was almost like a journal. And it wasn't until I think two years in that I was like, oh, maybe this is actually pretty useful. And I started to meet people like Mark Andreessen and Josh Koppelman at first round and a bunch of the PayPal mafia via the blog. And I was like, wow, this is actually like a pretty amazing thing. And so, yeah, I felt lucky to be early enough to build and sort of quote unquote build in public back in 2007.

Because I always think if you were writing back then was similar to like Paul Graham. People kind of refer to those essays as timeless classics. And you managed to reach the pinnacle where you have these kind of timeless Andrew classics that always get brought up time and time again. And you had a. One of the reasons I'm kind of bringing that up is because you had a really cool model that everyone always refers back to about the saturation of marketing channels. Like it's a, it's a guarantee that these things will saturate at some time. There's a ceiling to all of these kind of different growth channels. And you had a really great article recently around AI and the evolution of marketing because of AI. And one of the things that most marketers and like businesses in general are really scared about is how quickly AI can saturate things because it can copy and paste people's tactics. Now we can get into whether it really does copy and paste great tactics versus average tactics. But just how, like how is your thinking evolved today with AI kind of really starting to be integrated into go to markets about how quickly these kind of channels and tactics and things get saturated.

The essay that you're referencing on my blog from years ago is called the Law of Shitty Clickthroughs.

Right? That's it.

That's a great title by the way. And it turns out great titles are part of. Of timeless classics. And you nailed that.

That's right. That's right. Yeah, you got to do a little, hopefully there's some substance, but you definitely need a catchy title. So the thing with A lot of Shitty Clickthroughs is I basically looked at and said, look, if you examine the first websites on the Internet and I was looking at Hotwired, which is sort of the predecessor to Wired.com and the first banner ads that were ever created for websites like that. The crazy thing is the click through rates were actually just enormous. You know, these banner ads had 70% plus click through rates and people were so interested in them just because of the novelty factor. And you fast forward 20 years later, the average click through rate on, you know, across the Facebook network is probably, you know, under 0.1%. And so you're just talking about an insane drop. And so we could, we could run that same analysis not just for banner ads, but for email marketing and for, you know, SMS marketing and for, you know, all sorts of things. And the argument there is, look, you know, marketers are sort of in a perpetual kind of forever war with each other. And as a marketer you kind of have like this choice that you can make in the early years, like am I going to go to a cutting edge new marketing channel? Okay, and that's great. And you know, today it might be kind of variations of what's going on in the creator economy. It might be variations of what's going on in, you know, all the interesting things in Web3 and incentivized marketing that people are doing. It could be, you know, many other things that we could talk about. But your, your addressable TAM is just smaller on that, on those channels for the definitionally because they're new. On the flip side, you could go for really, really big and broad channels, but then you're stuck in there with all the other marketers that are sitting there copying each other endlessly and you have to decide like, okay, well you know, the performance characteristics on those channels are, you know, law of shitty click throughs. They're getting worse and worse and worse. And you know, where do I want to play? And so, you know, so, so, so you often have that happening and then, you know, and then Karen, to your point on this, with, with AI, every emerging technology is intertwined with the marketing techniques that sort of, you know, sit on there. You obviously can't have email marketing if you haven't invented email. And by the way, it was funny enough I was doing a deep dive on like kind of archaic Victorian age technologies from 100 plus years ago. And literally as soon as they invented the telegraph, there was telegraph spam.

Right. We're not doing anything new here.

Exactly what you'd expect. And as soon as you invent newspapers, which you had to invent these steam printing press newspapers in the early to mid-1800s, one of the big innovations that put what eventually became the New York Times catapulted it into success was they used an ad driven business model and you could buy these newspapers for one penny and then off you would go. Instead of, at the time, the original model was something that cost ten times as much. And so this was sort of like a freemium ish kind of thing. And so I just think that I can't help as somebody that has this lens around growth and marketing to look at everything that's happening in AI and not immediately consider all the consequences that are going to occur. And I think, Kieran, I think you're absolutely right. I think what it means is that there's quite a few things like SEO, I think that's probably the most obvious one that are going to be completely reinvented, upended, maybe made irrelevant, possibly. If Google decides to. One box, every single one of the answers that you have, what does that mean for, you know, the 20 search results under it? Yeah, you know, like, it may be completely different. And then on the flip side, it may be very exciting for other things like video, where it takes, you know, we were just talking, you know, just, just before we started about how. How much of a pain in the ass it is to start a video production. Maybe now it's just as easy as, like, you know, writing a little blurb and then, you know, full on, like, amazing interactive, you know, video gets generated. And so maybe we'll have an explosion in video or maybe we'll get an explosion in interactive video. I mean, all that stuff is very, very exciting and, and cutting edge.

Yeah. I think what strikes me differently about this, right, in the past waves of technology evolution, let's say, from web one to Web two, where you had the rise of mobile, you had the rise of the social web, it's not that the Web 1.0 channels, the banner ads, the print ads out of home, all those things, it's not that they stopped working. There were just these new, more effective channels where with AI, it's facilitating saturation across channels, new and old. Like you used SEO as the example. I can tell you from personal experience that the most established companies who are great at SEO are getting disrupted the most. Right. It's normally like if you have this big moat and you have this expertise, oh, you can kind of grind out and hold off that saturation longer than everybody else. And with AI, the opposite is happening where, like, the people who are most reliant on a channel are actually getting disrupted first, which is like a very different dynamic than what has, I think, happened over the last, like, 20 to 30 years.

Yeah, I think that's right. I think generally these channels have life cycles to them. And what that means is, you know, is direct mail, you know, does it still work or not? I mean, you know, I assume it's working for somebody based on, you know, how much, how much, how much direct mail I'm still getting. At the same time, I think that, you know, for, for all the marketers that are looking for that edge, you know, we're going to newer and better things. Right. And then, and then, you know, simultaneously, I would say, you know, if you are a very large branded incumbent and you are able to take a very long payback cycle, you're maybe willing to, you know, stick with channels that may or may not, you know, make sense for other people simply because for you it's almost like brand spend, you know, it's just like you've had it for years and you might as well keep going. And so, so I think, I think within, you know, the way that, that I might disrupt things, I think it has a lot to do with, you know, where the consumer engagement and attention is going to go. If it turns out that, you know, the hypothesis that people will start mostly interacting with LLMs and the search engine result page will matter less and less, you know, that's a big shift now. You know, the world is made up of a lot of, you know, late adopters, right. And, you know, and it may take 10 years or 20 years for people to stop typing keywords in a certain way and instead talk to their phone and just ask them questions. Or it might happen quickly, who knows? And I think if you're established, you probably keep it all in your portfolio and you probably experiment with some of the new channels that are coming out. But I think if you are a startup and you are building something from scratch, you have to go where all the new innovation is. It's really hard for you to compete in these established channels as a result.

Yeah. One way that I've always thought about marketing is as the, as the, as the channel becomes more mature, the complexity to do great marketing that stands out increases. Right. You can actually, you know, graph those two things out, which is basically the law of shitty click throughs.

Right.

Like I remember sharing a stage with Gary Vee where he was talking about his early days in marketing and then his initial open rates in his first newsletter were something like 90%. Right? Because the effort to get a 90% open rate back then, where it wasn't a commoditized channel, it was quite an easy thing to do. And as those things mature, people get much, much better. It gets much harder to do marketing that stands out. And then to Kev's point, what's fascinating about AI is with these other changes, these other technological shifts has opened up new ways to market to your audience, has opened up new channels to market to your audience, where AI actually doesn't do that yet, like maybe it will, it actually makes the channels you have access to today, it actually accelerates the diminishing returns from them. And to your point, in startups, I think that's a good point to maybe just riff on a little bit because I get asked this a lot from the companies that I invest in is should I start to invest in the thing, should I start to do marketing in the channels that have always worked, or should I start to pivot towards these new channels that should work within the AI world? And the challenge that all startups have, for the most part built a scalable and forecastable model around, have been paid, search, virality, and all three of those are getting harder and have diminishing returns. And I don't know what the. Like, if that's the Today slide with those three things, I actually don't know what the tomorrow is that is actually forecastable, measurable and scalable in the same ways. Like, I haven't seen glimpses of that because where I've seen AI add a ton of leverage is when you've acquired the demand. Now you actually have a ton of interesting ways to extract value from that demand and close them to customers. But it's not clear to me what are the replacements for those three things.

Hey, everyone. You know, we're always getting requests on the show for more recommendations on how you grow with AI. Well, we're bringing you another one. We've got this free guide from HubSpot on how you can use AI to completely change the way you create content. This guide, it's got it all. Prompts to assist you, writing and editing steps to incorporate AI into your marketing plan, and guidance on how to roll out AI across your team. It is the guidebook on how to integrate AI into your content creation. I've dropped the link in the description below. You're going to want to check it out. Now back to today's show.

Yeah, yeah. I think what is benefiting a lot of AI startups right now is just, you know, the tremendous amount of consumer engagement. You know, we have seen more AI startups go from zero to a million of ARR in a month, you know, like of launch than we've seen anything, you know, growing that fast. Right. And I think the reason for that is because when all of these new technologies emerge kind of in the early years, all that matters is does it have the it works feature or not? It doesn't matter how pretty it is. It doesn't matter, you know, what the brand is or the design. I mean, we're all using a product called ChatGPT. I mean, think about how from a branding standpoint, how crazy that is, right? It's like a bunch of just letters just strung together. Like no marketing person would have come up with that. But, but the fact that it works is enough that people want to want to try something new. Now, of course, we're going to, of course we're going to, we're going to evolve into a different phase of the market. You know, we're going to get to a phase where it is about, you know, functionality and you know, what is it that this LLM can do over this other LLM and is it integrated into, you know, my workflows the right way and you know, all that good stuff. And then eventually, you know, and this might be 10 plus years from now, these products often get into a world where it's more about, you know, brand and design and ux, you know, that's sort of like where, where we are in kind of non AI mobile, you know, apps. And so I think, you know, because product categories themselves kind of follow this, this life cycle. The funny thing I would say from, from a growth perspective is that a lot of these products don't have to do anything clever with, with marketing. They just have to do something amazing that previously was, you know, couldn't have been done. And there's so much buzz around, you know, around it in social media, et cetera, that it just works. Now of course, like, you know, I think we all, we all are cognizant of the fact that because so much of social media is visual in nature these days, right? It's video, it's photos, et cetera. That if your product is in particular a creative tool, you know, that allows some kind of very interesting transformation, whether it's in video or it's you know, upscaling photos or it's doing something interesting with audio, the, that, that certainly spreads further and faster, right? That's a little bit of this like kind of product channel fit kind of idea. And so I do think that, you know, that that certainly exists. And then Kieran, I think the other thing that is really cool about kind of, you know, where we sit now with having so many different, you know, marketing channels is that if you're a new product, you don't actually have to go for the biggest channel right off the bat. You know, like you can, you can go build your 10,000 person discord channel, you know, and you don't, you're not competing with, you know, a Coca Cola discord channel that's you know, 100 million users. Like that doesn't exist. It doesn't matter. Like, you know, that's its own thing. And by the way, like, you know, I've still seen a lot of products launch on college campuses, you know, in this really classic thing that Facebook and Tinder and many other folks have done and you know, well, why does that work? It's like, you know, well, college campuses are just. It's very. It's very lindy. It's something that's always been around and no huge company's gonna go after it because it's just. It's also a little bit too small, but it's a great way to pick up your first couple thousand users.

Yeah, I definitely think for the AI startups, the product demos are really their best source of virality at the moment. I had a company that was talking to two great founders called Hyperbanded AI, and they have a AI tool that allows you to create a representation of your customer, to allow new reps to sell into that person, to train the reps to sell into that customer. And you can basically create. They created a bunch of bots for people and that people, that those people could actually, you know, film this. I did this film myself trying to be a salesperson. And then I share that with everyone because I'm like, oh, this is really cool. Like, look at me trying to be a salesperson. And again, it's because it's such a new use case, like, and so the virality is built into the fact that the, you know, the use cases is so. Is so new. I think one, one of the things that you did in your article, which you kind of riffed on, like, what are the opportunities for marketers? I think a lot of the AI narrative is around what are we going to lose? But I think it would be kind of cool to kind of talk about what were some of the opportunities you spoke to in that article.

I think that there's, you know, I sort of made like a couple core arguments.

Right.

I think, I think the first thing is that there's a bunch of boring answers to what, what's going to happen? AI plus marketing. I think we all know that, like, everything's going to get cheaper and people are going to create more video and, you know, there's going to be infinite variations of landing pages and ad creatives and, you know, all that stuff. I think that's obvious and, like, we're gonna all reap the benefits of that. I think the other view is just thinking through, well, what is it that infinite kind of access to marketing labor would give us? Like, if we all imagine that we had kind of the world's best marketing agencies and growth teams and product teams kind of in our pocket and we could do anything, like, how would that change the way that we think about this? And I think kind of underneath that if you kind of pursue that thought experiment, I think one of the first things you'd say is, well, first, in a world where you have infinite, you're able to create infinite amounts of content, why would you ever create content that's targeted at a million people when you could make content that's targeted one on one to people?

Right.

And in that way, does that start to almost erase the boundary between marketing and sales? Because if it turns out that, you know, every piece of creative is custom tailored to Kip and knows everything that Kip's ever bought and is interested in and everything that's ever happened. And by the way, and it may not be an ad, it may actually be interactive maybe, you know, you want to be able to click into something or interact with something and then have a whole conversation at some point. Haven't you just built kind of a virtual sales agent and kind of where does that go? And so I'm very excited about kind of the idea of the companion or the interactive agent as almost the form factor for the way that we talked about desktop software that's in the form of kind of executables and applications that you install. For the Internet we had the web, right? For mobile we had apps. And I think for the AI era, companions and agents sort of seem like that's kind of the UI mode that we're all going to end up interacting with. So I think there's going to be a lot of innovation there. A related thing to that is thinking about internationalization. I mean, one of the amazing new things that I've been doing with my elderly Taiwanese parents has been, you know, I can, I can actually, you know, my Mandarin's okay. You know, we all moved when I, when I was five and so I just don't have the vocabulary to describe to them like what I do for work or what I'm thinking. And so it's amazing to be able to, you know, ask, ask ChatGPT, hey, can you reply in Mandarin? Can you tell them what venture capital is? You know, can you tell them, you know, what a limited partner is? You know, can you tell them what the law of shitty click throughs is and just have them explain? I mean, it's just amazing. And so the idea though that, you know, that that can really be built into anything and we're going to get to a world where we're one global audience as opposed to kind of this very, you know, Balkanized kind of, you know, certain countries speak certain things, certain cultures have different things. You can kind of instantly you know, internationalize. And then maybe, maybe the third one I'll kind of leave us all with is we sort of been marching from, from a content and media perspective towards more and more vividness. Okay, so what do I mean by that? The Internet, if you go back 25 years ago, was all text. There's really no. In fact, actually before Marc Andreessen built the original Mosaic browser, the Internet, literally the web did not have an image tag. Right. It was literally all text and just colors. And so then we added images that was important. Then we added eventually video.

Right.

In 2005, kind of around YouTube and everything that was happening. Obviously Web 2.0 was a lot about interactivity. And then now what we see is all people want to consume is video. We were just talking about the idea that you can't build a podcast as an audio only format anymore. You have to go video. And so the broader question is, well, what's after a video? I think that's a really interesting one. It's one of the reasons why at Andreessen Horowitz, I've been these days focused very much on the gaming side of things. And the reason is because there's nothing more vivid than a 3D environment where you're with a combination of people as well as agents, you're exploring a new environment, you're talking, you're solving problems together. That to me is kind of like where this all goes as kind of media becomes more and more vivid. And like today, if you were to try to build an amazing 3D open world environment, it's going to cost you $100 million to build something like that. But imagine again, in a world of infinite labor with all this AI running around, are we going to be able to, as marketers, spin up almost brand new, really interesting, interactive, high fidelity experiences, almost the way that we spin up meme images or individual creative ads. Is it the case that when something big happens in the world that we immediately are able to build a whole experience, you know, kind of, kind of around it? And so, you know, these are the kinds of things that I think can get really, really interesting because it sort of allows us as marketers and folks working on these growth teams, sort of keep pursuing this like consumer engagement and vividness trend that's kind of. That's ongoing.

Yeah, I think that's something Kieran and I have been obsessed with, is like the cost of web app development eventually approaching zero. And you can have highly personalized, highly disposable, very niche web experiences. And you're kind Of I think you painted a really good picture for everybody watching. It's like if you started with the Internet with just text, then we added images, then we added video, it's obvious to me that the next stage is going to be very personalized video and images right through the use of an interactive. And then we're basically saying even beyond that there's going to be deep interactive web experiences that could potentially be one to one because you can have AI agents building out a content, media, gaming, whatever it may be experience just for Andrew, just for Kip, just for Kieran. And that in the world where that is marketing, life gets very interesting, right? Because then you have personalized channels that you're then doing personalized call to actions, monetization, events, all of those things, which is very, very different than the world we live in today, which is like, let's guess what this segment of thousands or millions of people want. And we'll keep guessing until our guess is good enough that we're like at the high end of the shitty click through range.

Yes, that's right, that's right. And I think we're, you know, I think with all of the AI kind of conversation via alums that are, that are happening right now, we're sort of in a golden age where like I feel so grateful every time I open up, you know, Cloud or ChatGPT or you know, one of these, one of these apps and I could type things in and I just get the non advertising, you know, like I feel like I'm using, you know, Google in the earliest version where before, before any ads. But like it seems obvious to me that, that the entire AI business will eventually, you know, become ad supported. And so what that'll mean is, you know, we're gonna have to figure all this stuff out. But it seems pretty obvious that in the same way that there are, there's SEO, that there's going to have to be this entire thing of. Okay, well when LLM answers your question about the best places to travel in Iceland, how does it decide kind of given the inherently commercial question that the customer's asking, how to actually prioritize what to show and is there a business model behind that? And I think it's inevitable that consumers are going to want free as opposed to paying a monthly subscription in order for that to work. And this is even doubly true for international customers. And so I think that presents a huge opportunity for marketers over the next decade to see whether that conversation is to LLMs or it's to agents. Or it's to companions how advertising and marketing will sort of be part of that mediation that'll exist.

Yeah, I think soon, I think we can kind of pre announce it here. We're going to launch a free app at HubSpot to help basically help companies understand how they rank in LLMs and how to improve how they show up in LLMs. Because we have that same hypothesis, right, that as more searches go there, especially with the iPhone 16 in the fall, which I think is going to precipitate like another round of AI adoption with the OpenAI integration there, like you're. It's SEO is still going to be important, but there's going to be a new level of how your company shows up across AI agents and LLMs that you're going to have to basically monitor and keep track of. I think that's right. That thesis is totally right.

That's right.

Yeah. And I think that all these kind of like marketing workflows are just going to be very, very different and maybe more indirect.

Right.

I was recently had dinner with a couple sci fi writers and I was asking them about using LLMs as part of the writing process. You know, they did not love the idea, let me put it that way, of course. But one of the things they talked about, which I think is fantastic, you know, kind of a great encapsulation, is, you know, they view their job as they are supposed to write things, write about ideas that don't exist in the data set. Right, yeah, that, that, that, you know, novelty, which is something that human, human beings just, you know, we love kind of novel stories, novel approaches, novel brands, novel messaging, that if you are as a marketer able to do things that sit outside the data set, that there will be a huge advantage to being first versus a world where everybody's going to be able to create infinite creative and infinite new ways of messaging. I think that whole area gets commoditized pretty quickly. And kip to your point on using all of these tools to monitor where things are in AI, I think not only is that going to happen, but ultimately that will be part of almost an instantaneous, you know, monitoring workflow. Right. Because this is an interesting question too. In a world where you have pushed out an infinite number of variations of an email marketing campaign that you've done, how do you decide, you know, that, oh, this one's working and this one's not working. It's easy. In a world where, you know, you can a B test and you're like, oh, well, you know, I'm just gonna have like four groups and I'm gonna send it out to a couple hundred thousand users and I'm gonna just do the analysis.

And.

But if you're literally sending very, very personalized things, then even the process of monitoring the performance, iterating on the output, doing the next set of sequences, you know, even that probably has to be managed, you know, via AI. Like there's no, there's no, you know, there's no kind of way around it. And so I think, I think it's going to become very interesting because the world of being a marketer is going to become much, much more indirect and much more something where you're managing a bunch of workflows as opposed to sitting around doing the creative yourself, et cetera, et cetera. And then for the rare folks that are able to do something completely novel, I think they will end up doing very well.

I think it comes indirect in a couple of ways. I think it becomes indirect in terms of marketing is going to have to go back and sit at a place where even the work you do is measured in indirect ways. We've been very fortunate over the past decade that it's been really easy to measure the return on investment on marketing because it's again because channels like Search paid, they're just like, much easier to look at the return on investment as you start to push into these other channels. Even if LLMs become the way that people search and actually get information back, you're gonna, you're gonna have to use correlation models, conversion lift studies, all these kind of different things to actually figure out the, the, the impact of marketing. But just coming back to like how marketers use AI, like part of this, you've hit on all the points that we agree with, by the way. The personalization of content, personalization of code, go much more faster in international. The other one I want to maybe come back on is the AI concierge, which is like the companion that you have with you. The one I, one I want to just touch on, which I think is really interesting because you kind of made this point, which is I think AI will enable marketers in two ways and maybe all people in two ways, but it will enable you to copy and paste good much more rapidly. But that's good within the current data set, right? So you can figure out what good is. And one of the examples that we have done on a previous show is, and I was showing people like I was using Claude to figure out, write in style so I could figure out like it would do A pretty good job because you have enough data in the data set. If I give it a template and then replicate your write in style, it would actually replicate your write in to an okay degree, right? And your write in is great. But I could replicate your write in to a good level, right? So now I can actually replicate like good. But to your point, because that becomes so much, that's going to get really saturated quite fast, right? Like the amount of marketers who now get to good is going to be accelerated. What I'm always interested in is like, well, what about the great marketers? And the great marketers I think are great or the great growth people are great, or most people who are great are great because they are curious and can figure out new things. I think for that group of people, I think AI unlocks the ability to be much more creative because it takes away some of the barriers to do that. And so the example that I've been thinking about over the weekend and playing around with is actually there's a great example from Darren, our producer, stuck in my head, which is like Darren was talking about the fact that he is not a coder, right? But he had a bunch of people over for a football match. He wrote a. He wrote an app in Claude. Claude wrote him the app. He published it to Claude, and then he had the people in the room use the app during the match to look at data and stats around the players, right? That is wild to me, I don't think. Wild. That is right. And I, over the weekend I had a newsletter. I was telling Kip, like, I sent a newsletter basically teaching people how to do this write in style thing. And everyone was like, can you release that as a product? And I was like, huh? And so it took me an hour to write the app in Claude and I could have it published and have it back to my newsletter readers to actually use that app. And to me, that is where I get excited about AI. I think the replication of good is going to be cool, but the ability to go fast on and I don't think it's AI helps you. I guess what I'm saying. My point there is the idea is still the value for the person, but the execution of the idea should be much more rapid because AI enables you to do it much more quicker. So AI can't give you new ideas, but it can enable you to execute on your ideas much, much faster. And I'm just wondering, do you kind of agree with that summary or how do you think about that?

All the anxiety and concern that's happened is people are thinking defensively about, hey, I have a particular function, I have a certain thing that I'm good at and I would hate for an AI to be able to replicate my writing and etc. Etc. I think the funny thing though is I think that the opportunity really exists in the kind of thing that Kieran, you and Darren are experimenting with, which is how do you take a person that maybe they are master of a particular function that maybe does get easier to do over time, but then you give the ability for that person to color outside the lines into other functions if you're not technical, being able to build a web app is amazing. And then by the way, all the people that are building, you know, video and you know, can maybe now build a whole like video game experience and they haven't been able to do that. You know, I often think a little bit about how amazing it is that someone like J.K. rowling can, you know, she can have a thousand dollar Chromebook, go down to a cafe and write a book in, you know, 18 months or whatever and then millions of people will read it and how powerful that is for a single person to be able to do. Now what she can't do and what none of us can do is to make a Marvel movie by ourselves.

Right?

But that's going to happen. You know, Grand Theft Auto 6 is going to cost a billion dollars and 11 years to make. What if a single person could build something like that the way that an author or writer can go down to a cafe and just open their laptop and do the thing. So I do think that there's, it's exciting, especially when you're thinking about kind of coloring outside the lines. And there's also kind of an interesting question of, you know, within a function, you know, if we, if we take, you know, video as an example, do you end up with like the Gary Vee's of the world were already popular, able, you know, Mr. Beasts or whatever, they actually become even more productive. And so then they get bigger. Right? That's one really interesting dynamic that will happen. The other interesting dynamic will be the content creators that are at the, you know, that can't create any video all of a sudden can. There's more people who aren't doing something than are even doing it. And so maybe you get something that's very interesting starting at the bottom too. And so it seems like the folks that really kind of are pressured in all of this are kind of the folks that are maybe in the middle. And it's a little bit mediocre. And what happens there? Well, for them, they're probably going to have to then cross into other forms of content, other forms of media in order for this to work. And we're talking about content creation a little bit right now, but just to overlay the marketing and growth angle, what I would say is I think all of this applies to marketing channels, marketing formats, et cetera, et cetera. Right now. I mean, we are in a world where even as everything moves to video, the percentage of companies that would love to do a ton of amazing video related to their products, whether that's great demos or customer testimonials or kind of little movie trailers, you know, they have to go hire people right now to be able to do that in order to, you know, to go viral on, on, on, on, on X and you know, Discord and all these places. But I think we're going to, you know, very quickly approach a world where that will just be part of that standard repertoire and it will benefit the very best people in the ecosystem who can then be a thousand times more productive. And I think it will also benefit folks that, you know, have never touched, you know, Final Cut Pro in their life.

Right?

And all of a sudden they're going to be able to do a lot of new things. And so we have a bunch of companies working on this. You know, one of the companies that it has been really cool to get deep with has been a company in New York called Captions. And it's an ex Snapchat team that's been, you know, working all this. And you know, and they're, they're going hard on this idea that, you know, from a paradigm standpoint, you don't need to have a timeline, you don't need to have kind of all these channels and things that you're editing kind of in a traditional movie format that people are actually willing to create video using, using a more like TikTok like kind of format. And that's going to be the future. We also see this with one of our portfolio companies, Descript and Andrew Mason and team there are doing fantastically well. And that's another great example where it's like, yeah, you can do audio, you can do video, but maybe the right format is actually editing the text that's in a document and the text is tied to the kind of video frames as opposed to all that. So I think a lot of this stuff has the potential to be reinvented. And with that reinvention, I think the marketers kind of Follow along really tightly with all this product innovation because every time there's a new innovation, a new technology platform, it's inevitable that it will be used for advertising and marketing and growth. And so that's one of the really exciting things.

What seems to be clear is that if you're a marketer, if you're a creator, you're not going to be able to exist in one format, one medium, right? If you're a great writer, the tools are going to exist. That, that writing is going to also eventually be turned into video and audio and other formats. And like, we are going to a more multimedia, more immersive world, which is kind of your point, Andrew. And I think some of these early apps are helping to, just to demonstrate that where you're like, oh, cool, I have this 30 minute episode. And that 30 minute episode can now be short form video. It could be text, it could be an email newsletter, it can be an audio podcast. It could be so many things instead of like, hey, I just write, you know, 1500 words a week and that's what I do or whatever that may be. And that the world is a world of more, it seems like in the future, more so than less.

Yes. Yeah, I think that's right. I think there's going to be more, more avenues of competition, more complexity.

Which.

Is very, you know, very exciting. You know, I think for all of us. And I think, you know, even, even the way the growth and marketing discipline has evolved over the last couple decades, I think shows that the discipline is just getting more complex. It used to be that you had product management, you had brand marketing and comms and pr, and then you also had sort of direct marketing, kind of all as pretty separate disciplines. You know, these days, whenever I go into one of these companies and I talk to somebody that's in any of these disciplines, they're sort of, you know, fully expected to be conversant on all the things. You know, like they have to have, you know, aesthetic taste and you know, have a point of view on, you know, brand guidelines. But they should also be able to talk KPIs, you know, they should also be able to, you know, think, think about product strategy. And, and, and I, and I think, you know, to, to your point on this, I think what that means is that in many ways I think the discipline gets bigger. It's a bigger umbrella. And I think that's especially true where if it turns out that cloud and kind of the sort of artifacts idea, being able to use LLMs to build applications and products gets more and more powerful. I think then it might be that building the product becomes trivial. And it's the distribution that's the hard part. And if it's the distribution that's the hard part, then it's like even more complex in terms of how you actually, you know, get your product out there in this really competitive environment. And we've sort of already been running, you know, it used to be that in the 1990s you would actually go raise a series A to go build your product. And you know, and you know, these days it's like if, you know, by your seed round you're expected to have your app fully built and have your first test set of users. Now play that forward another 10 years. If everyone can build a product that's like an Instagram or like a Twitter just in the LLM and that can be your quick evening project. We're just going to go from having, I think there's like 4 million apps in the app store. Are we going to have 40 million? Are we going to have hundreds of million? I mean, is the concept of a kind of pre built app even going to make sense? Or is everything going to be kind of generated, almost bespoke? I mean, it's wild to think about where this all goes, but I think it sort of reinforces the importance of marketing and all this because the more trivial it is to build these products, the more it becomes a challenge of, okay, well how do you rise above all the noise and actually get to the customer in the way that you want?

That. That, that's the wild thing, I think about software and AI in general is like, typically it has not been economically prudent to try to build an app for like a thousand people who want that customized experience. But if the cost of code goes to zero and anyone can create that app, you suddenly have the opportunity for people to have apps that service like 50 people and pay $99 a month and just have a whole set of those kind of apps, right? It's like the tiny for, it's like the tiny version of tiny. And that to me is the wild part. Like, I actually wonder how you think about that with your VC hat on. Because you know, those companies don't need to raise, right? And so like, if everything goes into niche and it's not, it's not going to be true of everything, there's always going to be mass, mass software that everyone needs to use. Everyone needs an LLM, everyone needs a Google search. But have you, do you have a perspective on that? As an investor, even, like how that might shift the way that shift the kind of companies that you're going to start to see.

This can only be a good thing, I think, for, for the ecosystem. Right. And the reason why I say that is because the idea of making a living being, being a content creator, like having a Substack newsletter or having a YouTube, you know, that was just not a thing, you know, 20 years ago. Like, you just, you just couldn't do that. And then once the market actually got big enough and you had all these people interested in all these kind of niche things. I'm on the board of Substack and you know, the Substack guys will tell you, you know, there's tons of people making over a million dollars a year writing about the weirdest, you know, most niche topics out there. And how great is that?

Right?

And what that does is it just creates, you know, many, many, you know, new small businesses and, you know, sometimes big businesses, you know, when you look at this and, and, and, and sometimes what that means is one of those companies, you know, that sits on, on these platforms, one of these media companies. And there's been quite a few as, as you guys know, kind of within the YouTube world, for example, you can build a media company and that might in itself be an investable kind of company. Or we're involved in a business called Voldax, which is a game developer that sits on top of the Roblox ecosystem, as an example. And so those themselves might be investable. Or what happens is you just have to figure out how to invest in the platform.

Right.

And what is the thing that's going to help you build all these. So, Kieran, I think an amazing business idea I'm sure somebody will go figure out is, is if there are many, many little niche SaaS, vertical SaaS, companies that can target, whatever, all the little workflows for dentist offices and plumbing vendors and all those other things. Is there enough commonality within all of those products to build a platform that sits on top of everything and make it really easy for anybody to spin up that kind of a business? And if that's the case, like, that's probably a pretty interesting company. Even if the individual companies that come off of it are like microscale. Right. And, and then on the flip side, sometimes these companies are, you know, the little micro companies that sit on top of these platforms, you know, sometimes they end up growing a little bit bigger than what we all think. And I think, I think for, for our role in venture capital, we really view it as how do we take things that show a lot of promise and might work and just scale it to something as big as we can possibly imagine? And sometimes this stuff is really counterintuitive. I wrote a tweet recently just talking about the idea that if you look at the last 10 years and you were like, okay, some of the top companies were, one, an app where you get into a stranger's car, another app that lets you watch other people play video games, not you play video games, you watch other people play video games. And the third being something that helps you find people's random houses where you can sleep in their house. I mean, that's just wild that those ideas were like the billion dollar ideas. And so a lot of times I think the reason why I say that this can only be good for the ecosystem is because if it creates the opportunity for people to become app developers, that never in their life would they have considered that as a possibility. Maybe they were only comfortable being, you know, bloggers, or they were only comfortable being kind of video people, or they're only comfortable doing whatever, you know, building a Shopify and then all of a sudden they can become like a SaaS company. That's awesome. Like, that's great. So I hope that happens and then I think we'll all be surprised by some of the companies that come out of there. I think we're going to read a lot of really fun stories about, you know, two kids that start and build, you know, dentist software and then all of a sudden it becomes, you know, the next, the next square or the next one of these kind of mega platform companies.

Yeah, there's a lot of untapped ideas hopefully AI can unlock. Maybe we'll just leave with this. Andrew, if you were a marketeer growth person, trying to figure out how to be incredible at your job within the AI era, what are like the couple of skills you think that person should be learning now, you think are going to separate the great from the pretty average?

Yeah, I think, look, I think right now there's no question that the rate of change in the environment is just going to rapidly spike. And so I think one of the first things I would really consider, it depends on where people are in their careers, et cetera, et cetera. But how do you put yourself in a place where you can see as much of the change and really figure out what's cutting edge as much as possible? And so I always tell folks that they should spend three to five years in San Francisco and kind of be right in the middle of it, and especially if you're earlier in your career, there's a huge bonus and kind of advantage of that. And I think that's kind of number one. Number two is finding a role where you're able to experiment and experience a lot of what's happening in the marketing world. There are folks that, you know, if you're. If you're working on something that's a little bit more traditional, you know, how do you kind of pivot what you're working on, you know, towards a lot of what's. What's changing? And then I think the third thing is, you know, trying to. Whenever there's kind of a lot of new emerging technology happening, I think you just have a huge advantage to actually going to the frontier and really seeing for yourself what's working and not working, rather than reading what's working for other people and what's listening to podcasts and doing all these things. The reality is that things are moving so fast, nobody knows. Everyone's speculating, everybody's kind of trying new things, et cetera, et cetera. But how do you put yourself in a situation where you're able to go and see it for yourself? And so that might mean for the entrepreneurial folks out there to be able to start their own companies and really experience that. That loop for other folks, it will be being part of teams that are very AI forward and finding a role where you're able to learn by doing. And that's the exciting part. If you were to join a marketing team that was all about Google and Facebook ads, the reality is there's not a lot of places to grow in that right now, because that's the part of pretty well understood. But if you go somewhere that's all about using the latest and greatest AI tools, or you're marketing an AI tool yourself, then you're starting to talk about being able to really grow your skillset quickly and really learning. Because right now nobody knows anything. And that's just fantastic.

Yeah, Yeah. I think 100%. If you want to really accelerate your skills, putting yourself in the right spots and then being a practitioner and actually learning for yourself is hugely important. Andrew, thanks so much for coming on the show. I know our listeners are going to get a ton of value from this, and we appreciate your time.

Awesome. Thanks for having me.

Thanks, Andrew.

Thanks for being here.

This data is wrong every freaking time. Have you heard of HubSpot?

HubSpot is a CRM platform where everything is fully integrated.

Whoa.

I can see the client's whole history.

Calls, support tickets, emails, and here's a.

Task from three days ago. I totally missed HubSpot.

Grow better.