How Smart Founders Use AI to Win w/ Hiten Shah

They're calling it AI Sloth. So it's like, don't post something that looks like AI wrote it. And I see things that people post that AI has written clearly and it gets a lot of likes, totally cool. But just don't build that brand because that brand shouldn't be what you make your whole thing about, which is posting content that's obviously written by AI. Well, yeah, I mean, everyone's using these tools, fair, but are you lazy about it or are you actually going to talk like a human to other humans? Because remember, AI is not reading this stuff. The majority of this stuff is not read by AI yet. The majority of the stuff is still read by humans. We could have AI talk to AI when we're there, but we're not there yet. So to me it's like be human. And I see people not being as human, even though the opportunity is being human. Telling your stories and AI helps you tell them, it really does. You just have to know how to use it.

Do you see right now ownable opportunity in the AI and marketing space that's far enough from the pack to stake a claim, but not so far out that people, people can't connect it to the current problems they're working on.

Welcome to Best Story Wins. I'm your host Jason and I'm joined by my co host Josh. And today we are with the one and only Heaton Shah of Dropbox. Heaton, it's great to see you again.

Yeah, likewise. We've known each other, all of us, for quite a long time. So it's always fun to chat either, you know, live like this for other people or privately. So just happy to hop on and talk shop.

So Heaton, what are you working on right now? A lot of people are surprised, like, oh, Heaton's at Dropbox. What are you doing over there?

Yeah, I mean, I started a company doing enterprise search now almost five years ago and then we pivoted into doing sort of IT and security related, kind of solving IT and security related problems that we at the time and even today we didn't really think anyone was doing super well.

Part of that is just, you know, a little bit of a specialty we have on sort of with the folks I work with, which is like pretty good at marrying user experience and engineering challenges to do something that's great for people.

So I think overall.

To kind of take a step back, I've never had a job in my life and so going and working for someone is a really weird thing for me. Probably weirder than Most people, because I'm 43 and I'm essentially getting my first job. I was about 42 when I got it. Be 43 or 44 this year, I guess, losing track of time, as we all do as we age. Right. I want to be 21 again. If I could say that, I'd be very happy right now. I think all of us probably would. Not that I'm regretting anything. So, anyway, long story long sold to Dropbox because the things we were doing five years ago and the things we pivoted into have a real good fit with the company and where they're going and what they want to be building. And got really lucky.

Early on in the process as I joined the company that they were interviewing someone I had known for 15 years, and he's my boss now. His name is Morgan Brown. And, you know, it's one of those situations where I'm sure there's other people that could manage me and do a good job. But I would say that out of everyone in the world that I could think of, besides my father managing me, Morgan is probably the best person that could manage me and deal with, you know, a person new to corporate life, so to speak. So, yeah. And, you know, to be honest, like, the main reason that I am excited to be at Dropbox is I've known the company since the beginning. I've known Drew since the beginning, but I didn't go to him and say, buy my company. I actually went to other folks in the company and sort of wanted to partner with the company for a number of reasons. And it turned out that, like, they were very interested in what we were doing and vice versa. And we were looking for a way to scale that product we've created to more and more people sooner, more companies. And so it made sense for a bunch of reasons like that. Plus.

The.

When it comes to startups and these businesses, there's always what. What I would say is a peak of value where, like, you know, it's kind of like a stock chart or crypto chart where things start going up and to the right. And. And this isn't necessarily like revenue or any of that. It's more like market sentiment for what you're doing. And. And as that sort of tailwind develops, you kind of ride that. Maybe you write it for revenue, maybe you write it for attention, whatever it is, and then it's going to drop. So, you know, if you think about last year, it was like. Like we had started talking to the company a little bit in very late 2023 and the acquisition actually got announced in October, but the acquisition was completed in May. And we, we integrated all of our technology into the new product they were building in that from May to October. And so we didn't announce it until October with the new release so that we could say, hey, we're integrated in. And that just felt right to everyone.

And so as all that happened, if you look back today, there's something that is so true that if you understood what we were doing, you would see how it all kind of matches up. So let me explain. What we were doing is managing permissions on documents. So we made it so that IT and security teams understood who has access to what and they could change those permissions in bulk. So for example, it's very likely that if you're a 200 plus person company, or even a 50 plus person company, you have thousands if not millions of documents that have public links, which means anyone who has access to the link can access the document regardless of if they're in your company or not. Sometimes there's really good reasons to make links like that that employees do. Sometimes there's no good reason and they just did it out of convenience to make it easier to share with other people. That also applies to company links. What this means is there's documents out there that are accessible to anyone in the company if they have the link. Now imagine if things like salary information are in there that most companies wouldn't be okay with everyone knowing, but somehow HR person might have accidentally made it company wide as they were sharing it. Then imagine that information being ingested by AI for like a co pilot or a chatgpt or any of those things. Then me as an employee can type in what's my boss's salary? And get an answer. So what we were doing for three years, four years, probably 3ish, was really deep on solving that problem. Today. Even today, Nira, which is the name of the company that Dropbox Bot that I started, we do that better than any other company on the planet because we let you change millions of document permissions at once. We don't know any other tool out there that does that. The ones we've seen that came out after Us were like 100,000 items at once, things like that. But teams have this problem even when they're pretty small because documents get just created every day, right? Tons of them. So now if you think about all these AI tools connecting to Google Docs and Microsoft and all this stuff, they're essentially creating this risk for every business that they work with and our technology solves that risk. That's kind of the reasoning and that wasn't clear a year ago when we were talking to Dropbox and other folks about kind of what we were doing. Now it's very clear that this is probably something that everybody needs. Everyone that's trying to connect to your company documents probably needs some kind of tooling like this or every company that uses this tooling needs tooling like ours to ensure that nothing gets leaked. So it's kind of a new risk vector if you want to talk security language and people get really anxious about things like that.

It's a really interesting problem that you all sought to solve and I can see why Dropbox had an interest in you guys. So people that are newer to B2B marketing, they might know you from Nira. Folks that have been around for a while, they know you from kissmetrics Crazy Egg. You've built a number of successful companies and I'm curious to hear you talk a little bit about the role that contents played in building brand and driving revenue for you and your companies over the years. And is there sort of a common thread?

I wish there was a common thread. I would say my brother in law who I started our first set of companies with probably has a more common thread. He's a marketer. If you hear his name, you know he's a marketer. That's kind of what he wanted for himself in the world. I think for me.

I get excited about probably more things than him on a daily basis about the world. And he gets excited about marketing very singularly.

Probably by choice more than anything else. Part partially by accident too. Same here. So I found myself in the Bay Area, I found myself around startups and you know, things like that a lot more. And so I'm probably known for a number of different sort of things. But I just love building products people love. I think that's probably the. And helping other people do that is probably the line that I would say about myself for a lot of, a lot of years. But it's not as focused as my brother in law or a lot of other folks like even a Gary Vee or a number of different folks out there that have really pushed their personal brands. Fun story. I wrote the forward on a book called Founder Brand because I taught the author how to do it and those are the kinds of things that are of interest to me. It's more about humans and humans that are creating things from nothing including myself and folks that I work with. That's Probably a weird explanation to your question, but that's what I got today.

No, that's funny. I. I hadn't read that book. I don't know why, but I actually downloaded it right after we had scheduled you to be on this show. And right in the beginning of the book, it's, you know, Dave's talking about Heat and Shaw and I'm like, oh, wow, you know, like synchronicity. So that was cool to hear. And, you know, obviously people know you for things other than just the companies you created and thought leadership and so forth. What's as sort of a follow up to that question? I hadn't seen a lot of your content maybe over the last couple of months, and then just I'd say in the last month or so I've seen.

A lot of it.

Curious to understand what's behind that more recent surge.

Yeah. And you know, to go back to your original question, I think like, this is a good add on to that we had when 10,000 signups a month on B2B was a lot. We had those at Kissmetrics because we were doing about a million paid views a month on our blog for a very Nichy topic, analytics, stuff like that. Right now those numbers don't sound impressive, but back then they were because that was like 2010, you know, 2009, something like that, maybe 2011. And so I think content has always been something that I have been attracted to and something where I actually cared about the quality of it and the words we use, whether that's when we speak or.

When we write and other people read it. They say a lot of things out there like write for yourself, but then they say the opposite, which is write for your audience, know who your audience is. So this stuff's confusing and trying to get into that and being like, I want to get really great at it. I think the one thing I learned to answer your question about what I've been doing lately is that you can't get to quality without quantity. That doesn't mean that quantity needs to be posted publicly, but if you want real world feedback, you got to put the stuff out there. So what I realized is we're about to enter and we have been in a world where quantity wins.

As long as you realize that you can learn from what you put out there, and the more you put out there, the more you learn and you can improve it. And what I've seen work really well is when you push out a lot of stuff and use the information you get back to learn what works and what doesn't work and then start refining it. Maybe you do less, maybe you do more, I don't know. Right. But I think one thing I would say is I have a lot of fun being goalless and aimless and the stuff that I can do on Twitter or LinkedIn or whatever. @ the moment, I'm goalless and aimless. And I just want to figure out, what are we dealing with in this new world? Because back. Back in the day, we would be able to create blog posts and Google would gift us traffic. That's no longer the case. Even today, everyone's scrambling to figure out, you know, new channels OR hey, is ChatGPT referring me traffic and all that stuff? And I'm sure a bunch of people have figured a bunch of stuff out there and we all will kind of hop on that bandwagon because it just makes sense and it's trackable and all that good stuff, but it's not really trackable. I don't know what convers led to it. And so it's another black box, like Google, if you really start thinking about it. Plus, like, imagine the day that ChatGPT decides to dial that up and start linking out, right? Like, there's gotta be a situation where they decide they might want to do that. That could be a debate they've already had. I'd be having those debates if I were there. I'd be trying to figure out the intent of queries and questions and chats and things like that. Even when I look at a lot of the cool stuff out there, like Bolt or Lovable, where you can like build an entire app and stuff, and you start really looking at the conversations. Because a lot of these are public, because they have public galleries. If you haven't yet, you should do that because that's going to tell you a lot about the state of AI today. Because that's probably the largest public repository of chats for a specific intent, which is create me a new thing. And I'm not going to like, what is it? Share the insights from doing that, but everyone should consider just browsing some of those public things just to understand how AI works today. And the short of it is, we have been obsessed with prompting instead of being obsessed with conversing with AI. And I think that's probably the big epiphany I've had as I've been sort of trying to get my thoughts out in a much more structured way, whether it's online or using AI to help me think things like that. But I definitely use AI quite a bit to help me think, and that's what has been really helpful. But I don't necessarily use it to generate content. And I think that's a little bit of a confusing thing for people because they're like, wait, it can generate content? I'm like, yeah, it can, but it's more of it's way better as a thought partner to help you think than it is to help you generate content. So if your end goal is I want to generate content with it, I think you're going to hit a lot of hiccups and problems if your end goal is it's going to help me think about things that are on my mind and maybe some of that stuff makes sense to share. I think that's the attitude I've been taking with it. That's been a lot more.

Helpful. Wow, there's so much there. Okay, well, first of all, I love your goalless and aimless thing. It's so like counterintuitive. But you tapped into something there, which I've been thinking of it as like my 4am test. Like, is it like the thing I'm working on? Like, if I suddenly find myself, like, I don't want to go to sleep, I want to just stay up and just keep doing this. It's like there's something there and even if I don't know what that is yet, it's like I have to pay attention to that. Right.

And, and then you know that like action and doing starts creating more information for you. Right. To then go, okay, what's next? What's next? And I love from our previous conversations, and I think there's a human and AI version of this story, but you've talked about these individual tasks and like, what's next, what's next, what's next? And thinking about a series of even like micro actions and in the context of AI, though, with that distinction between you know, even, even the conversation of like, will AI take our jobs? Right.

You mentioned that a job is really more like a series of tasks. Right. How do you see shift here fundamentally changing the way like agencies operate or individual marketers operate if you start enabling AI to do individual.

Tasks? Yeah, I think we've talked about this a bunch and I work on the AI products at Dropbox and a lot of the newer stuff. And so, you know, trying to think about how do you create AI powered products that people can actually find useful and not scary and also that are reliable is very challenging. And so when I thought about what are we really doing And I kept reading that people were like, hey, it's going to replace our jobs? I'm like, well, it's going to replace tasks before it replaced jobs. And so if we can, let's say, take an agency, let's say this is an agency that creates content for people. You have certain workflow. And I really believe everything old is new again. So if you really dig into a framework like jobs to be done, and you take it back to its origin, it has a lot to do with tasks. A job is made up of tasks. So someone already codified this idea, not from a I'm working at a job, but more like I'm hiring a product to do a job for me. And if you don't know the paradigm they talk about, what do you hire a milkshake for at McDonald's? And the truth is, what they noticed when they started studying that at McDonald's is people are hiring it as a morning meal because they figured out, why are people coming to McDonald's in the morning and getting a milkshake, of all things. And it's because they wanted something heavier to start the day, whatever. And that learning was one of the key things they talk about when they talk about jobs to be done. So to me, if I were running an agency, I would look at all the jobs which are human jobs, and then I would look at all the tasks that each human is doing. And then I would look at which of these should we augment with AI? I wouldn't say which of these should be replaced with AI, but which of these should be augment. And then you start augmenting them. For example, like.

I have a YouTube show, and.

We'Ve realized that the high production quality, all those things are super important. And so we spend a lot of time on that. I mean, last time I checked, we're at 50 to 100 hours for editing a video, which is ridiculous. We can't do that with AI because AI is not going to be able to do what a human can do on that. But one of the biggest things you learn when it comes to content like that, which is like a YouTube show, not an interview show, but an actual show where you're, you know, scripting it and making sure you know what you're saying is the script, everything's in the script. The hooks, the open loops, all that. All those things that, you know, if you're a copywriter, you understand deeply. Good copywriter, great copywriter. That's why they tell you, hey, you should be manually writing those sales letters. It's like one of the oldest tactics in the book, you know? And it works because it makes your head think in those types of prose and words and how to sort of think about writing that actually is attractive and gets an audience. So when I kind of think about all of these things and how writing and things like this work and even the stuff we're doing, like, we spend a lot of time on scripts. We use AI to do research for scripts, but we haven't found AI to be able to write the script in such a way where the script is gonna work. Can we get it to write a script? Yes. Can we get it to draft a decent script? Yeah. If we just ship that script, though, and try to use it, it's probably gonna not do as well as if we take that draft, use what AI is great at, which is putting a starting point together for us, and then turn it into something that's manually edited and all that. But if you think about that, we are saving time on research first. Then we're saving time on taking that research and drafting something as a starting point. And then we can do a pipeline of it. Now, if you think about all that and you think about what people are talking about, agents and all this stuff, and you expand it to that, you're like, well, agents are just an automation of a set of tasks. So you could say that agents just put tasks together so that they can be bundled into a single task that an agent can do. But wait, what are humans doing? Humans are doing the same thing when they manage other humans or when they manage their own work with. To do lists and things like that. So I think, overall, the way I think about all this stuff now is that we are able to generate more than ever. And because we can generate more than ever, content, words, whatever, we can essentially replace tasks that humans are doing today. But replacing an entire human means replacing creativity, replacing decision making, replacing context that an AI does not have, that you probably couldn't give it, and then finally replacing intuition. So I think the final boss for AI is can it replace our intuition? And I'm not really sure what intuition really is or how to describe it in a way to an AI where the AI understands it and can either mirror it back to me or actually do that intuition thing. This is why, for me, I like using AI to help me think because it informs my intuition. But is it going to be intuitive for me? Is it going to be driven by intent? Unless I tell it that intent? Probably not. Not.

Today. Yeah, that's good. We've Been talking about it internally a lot in terms of, you know, ChatGPT and Claude are really great right now for creating the kind of content out of the box that would have worked great in the golden era or the golden age of search. But now things are a little bit different. And it's not that people can just tell that you're using AI to write, but it's, as you've said, that intuition, that human touch. People want that now.

They can sense if it's not there. And for me, the way I write for LinkedIn or for Twitter or for our blog is I have conversations with people like we're having right now, and I try to understand, okay, what are the common threads, what are the things that I'm feeling people are sharing? And then I try to think about, okay, what's a unique opportunity or angle to start chipping away at? And then I bring these tools into the mix. I have them help me with some of the heavy lifting of some of the writing. But I think that the, the real hard work and where the leverageable opportunity is to write something that will resonate is in that ideation part and then in the editing part at the bit or at the end. So it's like that 10% at the beginning and 10% at the end. If you are leaving that to the machines, I mean, you're going to end up with stuff that, you know, five years ago would have been really useful for telling the search engines that you're writing on something that's topical. But right now it's just going to feel like anything else that people are seeing in their LinkedIn feeds that is obviously created by ChatGPT. I don't know, does that take resonate with.

You? With.

Me? Yeah, for sure. I mean, I think, I think the way you're describing it, as it would have flown five years ago, is very accurate. Like, a lot of the stuff I see, even people post that I know is AI generated, like, would have flown just fine. And then there's all these things that. Now I think this leads me to an existential thing is probably the most interesting part of this.

At least something that I don't think we can answer yet, which is like, does the AI train us to read like AI or think like AI? Because, like, I've posted some things and posted quite a few things and only once did someone say, hey, using AI for this, because, like, some of the pros, you know, sounds like that. And I'm like, actually, let me point you to content I've written that was before AI has the same pros. Like, exact same. That's how I talk. That's how I write. Like, but people think they know the patterns of what AI spits out. But the funny thing is, AI just spits out what we put in. So, yeah, there's little tells, like the word delve. Right. Like, and stuff like that that people popularize or EM dashes or whatever. And yeah, like, if you want your content to not sound like AI, don't flag human brains. That it sounds like AI. Right. Like, that's it. If you're using AI to generate content, fine. We all are in one way or another, or should be. Make it sound like a human. Make it sound like you read it, edit it, change it. Like. Like, who says you should take what it produces and go spit it out and throw it to the world and think that's okay? Right? Like that. That's kind of my current take, which is like, we're missing the taste when we just do that and let it go and let it ride. That's why I don't like thinking of it as a generation tool, the way I think a lot of other people do. I really think of it as a thought.

Partner. Yeah. And at very least, just tell it to not use EM.

Dashes. Right. It's a very challenging thing to get it to do that. So you're usually better off just editing it.

Out. To be honest, I was really excited when I learned the keyboard shortcut for EM Dash years ago. I was always like, how do people make that long line? It looks really nice. Some people write on the way.

Out. Some people write with EM dashes. I never have. It never made any sense to me. But I'm not an English.

Major. It feels fancy, you.

Know? Exactly. I'm not an English major. It's not my thing. I don't really care how accurate my statements are from a grammar standpoint or any of that, as long as you know I'm saying what I.

Believe. So you're pretty close to the AI world right now. Like, you're working in it. You're thinking about how that's affecting our daily lives all day, every day. Do you see right now.

What you'd call an ownable opportunity in the AI and marketing space? That's far enough from the pack to stake a claim, but not so far out that people can't connect it to the current problems they're working.

On. So a buddy of mine pinged me today, and he's like, hey, have you seen the new synthesia? Videos, which is the company that's doing synthetic videos of like avatars and people and stuff. And I hadn't seen it, and he basically said three months ago it looked like crap. But the one he just saw today, he says doesn't look like crap anymore. So I think the way to answer your question and even think about the question is fairly simple, which is.

It feels like the stuff that seems to work in the market is stuff that has an outlandish promise.

And that will get you the signups and the attention. The problem we have today is our ability to deliver on that promise. So I think the challenge we all have for building this AI stuff is how do we promise the right amount of stuff and be able to deliver on that instead of promising the wrong amount of stuff, which is too much.

And this example from Synthesia is really telling because they've been promising this more or less, but now they can do it. But they've been at it for quite a long time. But the thing is, now they can do it. So if now they can do it, should they have ever said they couldn't do it? I don't know. It's a little existential there for me because I'm sure the fact that these companies say they can do certain things causes them to get attention and signups. The fact that they cannot do these things very well yet gets them churn. So would you rather have the signups and the churn, or would you rather not have the signups? I don't know. Right. I don't know the answer to that. I like building products that don't have churn like that. But we might be in a reality where we're in a weird phase and it might last for quite a long time. And I can explain why I think it might last for quite a long time where this is just the state of things. That would be sad though, because that means that people are going to get really.

Fatigued from trying new things that don't work for them. And I think that's the world we kind of have gotten ourselves into. And one of the reasons I think that, you know, this is a never ending problem, or not never ending, but it won't end for a while, is the models keep getting better. And our ability to integrate what the models do ironically is probably dictated by the lies we tell of what it can do. Right? Because like it's a chicken or egg, it's a true chicken or egg. What do you do? Do you lie and say it can do it, but then not do it and not be able to deliver on the other side until months later or years later. I don't know. It's a game of chicken. I mean, even if you look at things like Cursor or now Windsurf and these coding tools, a lot of the population of engineers believes they're useful, but most of the population doesn't believe they're that useful. They'll use it because it helps them do something faster. Tab completions, a bunch of stuff like that. But when you really dig in, you're like, hey, can you build something production ready? Where AI wrote a vast majority of the code, you see very few instances where it feels like that's true and reliable and successful. And then you start talking to engineers and they're like, yeah, I, I use it, but I check its work or it has this limitation or that limitation. I mean, the whole battle right now between Cursor and Windsurf is related to one simple thing. It feels like one product is better with large code bases. And Windsurf has come out and explained why because they're not Windsurf and they launched today, they had a whole business doing what large context window related stuff, right? Like, and so they are able to say we could do it better than Cursor. And all those little bubblings of why people like Windsurf better than Cursor. The answer is if you have a large code base today, the consensus is Windsurf does a better job. I'm only mentioning that because Cursor is one of the fastest growing companies I think on the planet today in terms of revenue. And they already have a competitor and that competitor is able to use exactly the same technology but a little bit better and as probably already siphoning off a decent amount of users. I did a poll on this on Twitter just to understand kind of what was going on. And kind of the consensus is not enough people know about Windsurf, but the people that do seem to like it. It's kind of the just rough, broad strokes consensus without getting into the data. And I think like that, that those kind of things are what I study and look at to understand the answer to your question. And so when I think about AI and marketing and content, I think that you have to provide a 10x better experience than ChatGPT for the same use cases. And that's the world we live in because that's the popular AI tool. That's how people experience AI today. I think in the hundreds of millions, if not now a billion users. I don't know what their exact scale is. But it's enough that the familiarity of using AI and the association with ChatGPT, I mean, just wait till Thanksgiving this year. I'm sure Thanksgiving last year was interesting on that level. Just wait till this year. People are going to talk about ChatGPT more than probably any other.

AI.

So you brought up an interesting point on that, which, you know, I think we've seen it in any SaaS segment for years where, you know, maybe you leap out in front with a new feature and you know, maybe a few years ago it took a direct competitor, you know, a few months to copy that feature. Now it can potentially happen much faster. And as people, I mean, you see it even at the foundation model level, right? Like they get a jump on each other for three days before, you know, then the next, the next shipment comes out from the competitor. And as I've been thinking about this, I've thought, wow, it's such an interesting state of things where like maybe it's been engineering led, product feature led, you know, that gets the headlines and so forth. But you see it with OpenAI, you know, starting to do a lot more brand marketing and storytelling and kind of these simple emotional stories, you know, do you see that marketing, I mean, the Windsurf example is a great one, right? Like they have maybe a marketing problem, right? If their product is potentially satisfying customers more, but fewer people have heard of it, that's a marketing problem.

Right? Yeah, I think, I think to some extent it is.

And at the same time, like it's funny you say it's a marketing problem because the Cursor founder, founder came out the other day and said we spent zero on marketing. You know, the typical first time founder, probably getting really excited about not spending on marketing. And I think I followed up on that and I'm like, yeah, you should probably start spending on marketing now. That's probably a good idea.

Someone else is going to outdo you and there is someone else sitting there that they spent money on marketing. They can grow faster than you. So the problem with this whole startup kind of ecosystem is that there's not enough marketers that start companies.

And I mean tech and venture funded, I think self funded. There's actually way more marketers that start companies or agencies or things like that. And as a result you have all this FUD about marketing. You know, but imagine how big Cursor would be. Imagine if they actually did marketing. Imagine how much harder it would be for Windsor to even come in and do anything if Cursor did Marketing. But they came out and said they haven't done any marketing and they've got to 100 million ARR or whatever. Then it's just like not another one of those, like let's please do marketing. Let's not say that not doing marketing is a good thing, you know, badge of honor. Yeah. It's a way to accelerate a great product and get it into more people's hands faster. Yes. It's not a way to go market something that you know or spread something that shouldn't be spread because it's not any good. You'll find out really quickly. Right. You'll have churn things like that. So yeah, I just find that topic just comes up every now and then when there's a blow off, you know, word of mouth product that just grows. But I always am just like, imagine if they did marketing. Just imagine. Would it be 200 million ARR.

Maybe. Do you think we'll see a shift there? I mean, I think a lot of people are waking up to the reality that it's really easy to copy your competitors now, easier than it's ever been. And that also brand is ultimately where you can create and maintain a sustainable competitive advantage. Like, do you think that VCs and other folks are going to start to say we want to back marketing.

Founders? No. No. If it happens, it'll be like a blip.

Yeah. Why is.

That?

Because they go back to backing engineers. Straight up back engineers. Why do you think Y Combinator exists? Engineers who started it, an engineer. Right. So we have this ethos where tech equals engineering. Tech does not equal marketing. Tech does not even equal product. It just equals engineering. Right. Part of the reason is incentives. If you build an amazing engineering team, you are building enough value where someone might acquire you just because you put a bunch of people together that can work well together and are a bunch of really smart engineers, for example. So I don't really see that shift happening to marketing or marketing becoming more prominent in the world of venture capital anytime soon in terms of them investing in marketing founders. Part of the reason is is a marketer, a builder. And again we're getting into your point now. So the shift would have to happen that it is actually objectively much easier to do the tech side. And I don't think we're quite there yet. I don't know what that'll take, but that's one shift. But still, it's not like developers disappear and then, and then the bar gets higher. Right. Like the investors today have been funding a lot of the model companies. Is a marketer going to start a model company today? No, it's researchers, AI folks, engineers, people like that. So I think we'd have to see a shift in a few areas, including the type of companies that are started, for venture capitalists to be in on that. I think there was a point in time where many, many years ago where this was a little bit of a thing, but we always swing back to investing in the builders. So if marketers turn into builders, somehow I think that can shift. But.

The bar gets raised then on what you should be building and then you're back to square one. And I've seen that happen over and over.

Again. You have this trademark like direct but not mean feedback style that I've always loved. And when we were talking a few weeks ago and kind of exploring some of these topics and I went and did some, what I thought was decent, deep research and organizing this and started screen sharing it, you just said, this is garbage.

And I loved that. And then you, I showed you what I'd prompted and you're like, there's the problem. You had a trash prompt to start. So you've talked about this kind of importance of input quality when using generative AI. Can you walk through the process and like you mentioned earlier, you know, using it to help you think, where do you start and how do you get to where you're not putting garbage in and getting garbage.

Out? Yeah, I was at a dinner last night and we were talking about this. It was a bunch of product people. And first I was shocked at how little they used AI, but I was also impressed at how much some of them used AI. So it was like a little bit of a. Just interesting thing. And all of them, it was an AI product in it. So everyone built. Was building AI products, every single person in the room. And we were talking about this similar stuff that you and I were talking about and have been talking about. And I think the conclusion I came to last night is pretty obvious. And it's been in my head, but I didn't articulate it until then, which is this. I think AI can be very. What I would call recursive or meta. And that's what you and I went after when I was showing you how I would have done it. And so the way I would generally do.

AI if I really had a very clear output or goal, research is a very clear output or goal is have AI help me make the prompt. So what we did when we talked about it is we described what research we were looking for. And then we had AI make a very detailed prompt.

And then we went and ran that prompt. I think it was like seven minutes, 10 minutes or something. And I had you compare the two, right? And I was like, hey, look, this one's better because the prompt was better. But I didn't make the prompt while in yours you made the prompt. I just used a similar prompt to what you did and told it. Make me a prompt about this for deep research so I get these outputs, right? And if I wanted to go further, I, I would be like, if there's anything unclear about what I'm asking for or there specifics that, you know, you as the AI would want to debate before you make the prompt, let me know what those are and help me help you make the decisions. I would have went that much further if I was taking it more seriously. And we had a little more time that day. And then that research prompt would have been even better. So the one hack trick will work for everyone is if you, if you're not getting the outputs you want, tell AI to tell you how to get the outputs based on the prompt and have it help you make the prompt. And you can even get more advanced and be like, are there decisions I need to make for you or with you, or are there things that could go one way or another help me. And even if you use deep research by ChatGPT, you'll type it in and it'll already come back to you and say, hey, I came up with some ideas. Do you want this or that? Right, It'll already do that. But if you preemptively do some of that in your prompt building, it'll just make the things it asks you that much deeper, that much better, that much more interesting, and then you'll get more of the output you want. Right? So that was kind of, that's probably the biggest trick that clicked for me that people might not realize this. Well, to me, you know, I spend three plus hours a day using these chat tools. It's just almost second nature that if I'm stuck or if I'm doing a deep, I will not do a deep research prompt without getting AI to make the prompt for me. Now, let's put it that.

Way. Yeah, it was, it was a game changer for me. I just, yeah, if you want to get into a more nuanced and specific answer, right. And it's amazing how kind of going back to your task analogy or description of this, it's like so often the problem is at step zero, right? Just like how Are you thinking about this work to be.

Done? Yeah. I mean, I was even talking to a friend of mine the other day, and I think that, you know, he asked me a really interesting question, which was like, hey.

What are the new role? What's the new role with AI? Because he's a marketer and his thesis is like, every paradigm shift comes with roles, like, new roles. Like, we have mobile engineers now. We have iOS engineers, we have Android engineers. That only happened because of mobile. There was no such thing as those engineers prior at all.

And so my hat in the ring on this one is I think there will be a title, I think it will be across departments and in departments called AI Manager. It's the person who manages the AI. But I think everyone should have that role. I am a product marketer plus AI manager. I am a product manager plus AI Manager. I'm a marketer plus AI manager. And I think that's the new role we all get to have. Do I think it's a role someone should be dedicated to? I don't know yet. Maybe not. But I'd say AI Manager is what's in my head. Because when I think about what I'm doing, I'm managing AI, I'm managing deep research runs, I'm managing. If you want to call it agents, call it agents, I don't care, call it whatever you want, but I'm managing these processes. Usually those processes required humans and me to manage humans to do them. Now I can do a lot of these with AI.

Right? Yeah. So you've been in B2B marketing for a long time. You've seen a lot of things change.

The last couple years. We've seen a lot of disruption. We've talked about search a little bit. What are some things you're seeing.

Right.

Now.

That you're just really excited about? And conversely, what are some things you're seeing right now? Maybe some trends that you think are maybe a waste of time or you're hoping will soon fall by the.

Wayside? You know, I think there's a lot of opportunity out there, and.

I think people aren't taking advantage of how much easier it is to tell your story. And that's not just the content and AI and all that. I'm not even talking about that. I'm saying that, like, the platforms exist for folks, even with small followings, to be able to make a lot of money or create a lot of value with a small following. Obviously the following should and can grow, but the ability to get attention, I think, is way Better than ever before. And I think that exists for B2B. So we should use that attention, or ability to get attention very wisely. I know this podcast. You have stories in the name of.

It'S stories. Right. Like, or story one or the other. Doesn't matter. Same.

Work. Yeah. Best story.

Wins. Yeah, right? Best story wins. There's a reason you folks chose that. I'm not saying that that's the name of the podcast and that's why I'm about to say this, but to me, I do believe in that. The best story wins. And so my advice for marketing on the positive side is the best story wins. So maybe you want to tell your story out there because the attention is available to you if you tell your story. Everyone has stories to.

Tell. Thank you for being the unofficial.

Ambassador. I mean it, though. You know me. You said it earlier, very direct. Well, I would tell you if I thought it was crap. It's not crap. I think it's good. And then on the negative side, I think we talked about this earlier, which is, you know, they're calling it AI Slop. So it's like, don't post something that looks like AI wrote it. And I see things that people post that AI has written, clearly, and it gets a lot of likes. Totally cool. But just don't build that brand, because that brand shouldn't be what you make your whole thing about, which is, you know, posting content that's obviously written by AI. Well, yeah, I mean, everyone's using these tools. Fair. But are you lazy about it, or are you actually going to talk like a human to other humans? Because remember, the AI is not reading this stuff. The majority of this stuff is not read by AI yet. The majority of the stuff is still read by humans. We could have AI talk to AI when we're there, but we're not there yet. So to me, it's like, be human. And I see people not being as human, even though the opportunity is being human. Telling your stories and AI helps you tell them. It really does. You just have to know how to use.

It. Yeah. What would you say to someone that's just starting out? I mean, I think if someone goes on LinkedIn today and they're thinking about using the platform to grow their business, maybe they see all these other people that have these huge followings and they get all this engagement, and maybe it can be a little daunting to think about writing and not getting a lot of attention or engagement right off the bat. What would you say to.

Them? I think the number One thing when it comes to engagement is to be engaging. And there's one way which is obvious, which is what we've been talking about, which is post content. I think the other way to be engaging reminds me of what I saw when the tech blogs were coming out, which I think you both have heard from me and we've talked about over the years, maybe an embarrassingly long time ago, but it's this idea that the way that Pete Cashmore grew, Mashable, again, people listening to this that don't know these things, TechCrunch and Mashable and a few others, Gigaom was another one, were big blocks, tech blocks. But the way Mashable grew was very different than Gigaom and TechCrunch. TechCrunch was run by Michael Arrington, a lawyer turned writer. I would say blogger. Lawyer turned blogger is really what. How I would describe him. And now venture capitalist. He started TechCrunch. It's still the number one startup blog. Tech blog, whatever. Tech blog, let's say. And then you had O. Malik, who, lifelong reporter, journalist, who started a blog as well. But then you had Pete Casmore over in England, I think, and he was number three in terms of attention. Why? He was the first one to comment on every single TechCrunch article with something of value. And he was engaging as a result. And then when you clicked his name, you went to mashable.com. that is the number one thing he did to grow his brand from scratch back then. So if someone's new to LinkedIn, I'd be like, find the engaging people, find the engaging post and try to be early on commenting on them. And that's how you'll build your brand faster than almost anything else you can do. But here's the thing, if you have no posts and you do that, that's like Mashable having no blog posts but a website. And doing that doesn't mean anything because you have no way to get them into your world after they kind of take a peek with a comment. You could do the same on Twitter. It works brilliantly, especially if you're able to reply and say something of value that adds value to the conversation. Other people read it, they like it, they follow you, all that good stuff. So I think that's the thing that I would do more of. I might even get better at that before I got better at posting my own content. But you need a drip of your own content too, at the same time, even if it's not popular, so that people start seeing it and know what you're all about, like. And I said I'm goalless and aimless today. That's not my intention. And so that shouldn't be your intention when you do what I just said. For me, it's the best place for me to be with a full time corporate job, a bunch of other businesses that I have that other people work on, that I use as a hobby. I guess right now in terms of my time, I can be goalless and aimless, but I wouldn't suggest that to anyone else. I'm going to stop being goalless and aimless at some point soon, but right now it's been working for.

Me. You're in your goal discovery.

Phase. There we go. I think that's, that's actually a fun way to think about.

It. If you were launching a brand new startup today, how would you approach the growth efforts differently than you did 10 years.

Ago? So I think 10 years ago and today I don't think I'd approach it super differently. The way I think about it is.

You want to find out where that audience hangs out, then you want to engage them there. And that's pretty much it. And so a lot of the stuff we talked about today is all the things I'd be doing. That's what we did in the past too though. What we would do is we would blog a lot, we would build a presence on Twitter. Back then, the way we built a presence on Twitter was for one of my companies, kissmetrics is we started, I guess hashtag jacking is one way to think about it, which is like there was a hashtag called measure and that was when hashtags were a thing on Twitter. So we would post an interesting link, not our own, someone else's. That's when links worked on Twitter too.

And put the hashtag measure and that community that's around that hashtag were a bunch of web analytics folks and marketers and ones that cared about measuring stuff. That's why it was the hashtag. And we use that to draft on that community and make sure that they knew who we were. And so I think there's incredible opportunities like that everywhere. Subreddits, LinkedIn, bunch of groups on LinkedIn are actually still interesting Facebook groups. So depending on what you're doing, find your audience, go engage with them, you know, and make sure that they know who you are. Anything more than that around conversion or signups, like it's not that important early on. What's important is people know you exist. So I would be Fundamentally trying to solve that problem.

First. Yeah, that's good. So one of the things that we're excited about currently is just seeing people kind of return to taking brand seriously. And you got a lot of people on LinkedIn talking about the importance of brand. I've seen a few posts from you that touch on this as well. Care to make any predictions about what companies are going to need to do in the future in addition to focusing on brand in an effort to stay relevant and stay front of.

Mind? Yeah, I think brand is a funny thing.

And it's really the trick with brand is like, how do you carve out a literal piece of someone's brain and say, hey, I'm going to be inserted into that piece. So today you think of AI. Most people think of ChatGPT first and it's not because it was first, but it was because it provided such a compelling experience that I didn't know how I could get that experience any other way. So now if I try that experience somewhere else and it's not quite like ChatGPT, I might just go back to ChatGPT. I've seen a lot of people do that kind of stuff and they're not probably the most fantastic example because it was new product, new technology, all at the same time. But it's still a decent way to think about brand, which is like they've carved out AI in your, in most people's heads. You go talk to like, you know, one of my older uncles or you know, grandpa, you know, not my grandpa, but like my kid's grandpa. So my father in law, my father, like you ask them about AI, they're like, oh, chatgpt. And then, you know, the silliest thing, which is a simple sidetrack, is like at one last year at Thanksgiving or around Thanksgiving, one of them, one of my and my uncles was asking me like, hey, I don't understand this. And my kids tell me it's complicated. And then I was like, wait, no it's not. And I kind of explained to them how it's predicting and how it works and how it, you know, all of that. And even the next day I saw one of the uncles again, it's like, thank you for that explanation. And so I think like I inadvertently probably carved out, oh, he's the AI expert in their heads. Not that I was trying to, but I wanted to give that example because like, I think one of the, and this is just a very tactical trick is you dispel someone's myths. Usually those myths are created by what other people have told them. And if you have an opportunity to do that, you can carve out a piece of someone's brain real fast. And I think a lot of the, that's a lot of the reasons why things like before, after graphics work so well for us, right? Like you're going to go buy a diet pill, you see the before after, you're like, I want it, I want that, right? I look like the before, I want the after. So it's like, I think there's strategies like that and very tactical things you can do as long as you're doing things of substance and you have something interesting to.

Say. So what, what are you working on that you're most excited.

About? So in, in all the products that I, I touched today, we, we are incorporating AI, as a lot of people are. So I think what's exciting to me is how to do it correctly. How do you incorporate AI into products in a way that's reliable, compelling? Right? How do you use AI in a reliable way? So I think that I'm not saying I'm excited about AI, I'm saying I'm excited about new technology that everyone has access to. And then since I build with this technology and try to incorporate it in the products I touch, it kind of extends into that. And to me, if you're working on anything remotely close to AI and you're not like trying to find a way to spend dedicated hours with the AI chat tools, you're probably doing yourself a disservice because then you're not going to understand what, what they're capable of. So I pay for the $200 chat, GPT, Pro, Pro, whatever, plus whatever they call it. And if they had a plan that was $500, I'd probably pay for it. If they had a plan that was a thousand dollars, I'd probably pay for it. But the reason I'm paying for it is I want to understand what is the cutting edge of what any consumer can do today. Because yes, I can use more advanced things. Yes, I can look at all these agent platforms, but frankly, I care more about what everyone else can do too, and how to push the edge of that. Because that's already good enough. It's like more than good enough. I don't need to get agentic yet. It's not reliable, frankly. And most of the tools are not really agents. They're just zapier with a twist. So anyway, that's my mini rant on that.

One.

So, Heaton, this has been awesome. And what's the best place for people to find you. And also your amazing YouTube.

Show. Yeah, the YouTube show is called the Heaton Show. It's on YouTube. You could probably look that up. And then.

I'm on Twitter, I'm on LinkedIn. Usual, usual stuff, you know, Follow me if you think it's interesting.

Reply if you want to chat and DM if you got a hard question. Get a lot of.

Those. You can answer in.

Private. Yeah, there you.

Go. Hey, thanks so much.

Heaton. Of course. Thank you.

Guys. Yeah.

Thanks.

He.

Column 5's best story wins is for marketing and branding professionals looking to unlock their growth potential. Each episode features a conversation with industry leaders about how they win the hardest hearts and minds of their customers and build world class brands. You'll learn about their success stories and their failures as well as ideas for how to take your own marketing efforts to the next level. Welcome to Best Story Wins.