The Best Marketers Are Becoming Part-Time Engineers

You know that feeling when the strategy is done, the brief is written, everyone's aligned, and you realize someone still has to sit down and actually create all the content? That someone is you, and it's due tomorrow.

Breeze Assistant can help. It works right inside HubSpot. Drafting campaign copy, blog posts, emails, all in your brand voice, all grounded in your actual customer data. So you don't just create content, you create content that converts.

Check out HubSpot.com, the agentic customer platform for growing businesses. Basically, Neil... everyone is now becoming a part-time marketer or engineer. So if you look at the little table over here, you can see that every single role, row here, right?

Every role slash row is, they have some form of engineering that's happening over here. And they also have some form of marketing that's happening. It doesn't matter if you're a customer experience, design, engineering, whatever, you're doing some form of marketing. Now, obviously, a marketing person is going to spend 36% of their time doing marketing.

An engineer is supposed to spend 53% of their time on engineering. But you can see these rising as the main things. And the thing is, Neil, you and I, we're marketers naturally, but I do find myself building more stuff now. And sometimes, yes, we're going to think about finance.

We're going to think about design as well. But I think this is an overall... This is one of the big findings in their study at Procter & Gamble. So this is just people at Procter & Gamble, I guess.

But I find that really interesting that we've talked about the Coinbase philosophy of the pot of one where one person, like one marketer can be part designer, can be part engineer and vice versa, right? So I think that's interesting and it's going to continue to rise. I think a lot of this stuff is just going to merge together.

Yeah, totally. And one caveat here, just for everyone to know, Procter & Gamble is the fifth largest advertiser globally as well. So advertising and marketing is also in their DNA. They spend a lot more than most companies.

But I do agree in general, people are going to start doing more things because of technology and their ability to code and stuff like that. Because if you also look at Procter & Gamble, people aren't inherently engineers, right? On the marketing side, yes, they're a big spender on marketing, but people in different departments generally aren't engineers, right?

I was having lunch with my CTO today, I remember walking, and he's like, I actually think that we're gonna need more people that have product builder type marketers that are deep in marketing but can build products versus a standard engineer from the past that will build up a UI and build up a typical SaaS product, right? And so he's and I'm like, I honestly think it's going to be a mix of the two. Like you might need some of that. You might need to kind of build out like a traditional UI.

And then you're also going to need like marketers such as yourself, Neil, or myself that has a taste and judgment. And we know how to at least build a V0 of something to get at 20, 30, 40, 50 percent there, which then brings me to my next point. Elon actually tweeted this yesterday. So.

He said this, source code is on the verge of becoming like assembly. So CTO sent this to me yesterday. I'm like, I don't even know what that means, right? So I asked Twitter or X to summarize it.

And Elon basically is saying that AI coding tools have advanced to the point where developers increasingly trust outputs without reviewing generated source code, mirroring the historical shift away from manual assembly programming after reliable compilers emerge. I have no idea what that means, right? So... He's like, the next evolution is AI system will produce optimized binary executables directly, eliminating traditional source code as an intermediary step in software development.

There's too many big words here, okay? So I'm just like, dude, what does this mean practically? So practically, it means software development stops being about writing and editing lines of code and becomes about describing what you want and verifying that the resulting program actually does it. Okay, so here's the key thing here.

So Neil, you and I, we don't review code, right? Now, people talk a lot about technical debt and things like that, right? They talk about the code base being very bloated. And NVIDIA is actually notorious for having a very bloated code base.

Whereas before, if you came from Silicon Graphics, it was very perfectionist, right? Elon's saying it's probably going in that direction where you're not even going to be talking about having too much technical debt because the stuff is going to be so good, eventually it's going to clean itself up and we're not even going to need to understand what's under the hood. That's all that he's saying, and I think it's already happening right now.

But the crazy thing, again, is when I talked to you on the phone yesterday, we have people that we know where they don't let their engineers, they don't trust AI, it's not reliable, right? I think a lot of this stuff is just gonna go in this direction, and humans are just gonna say, hey, I want this, and you verify, hey, is this actually what I wanted? If it is, great, no complaints.

Yeah. So let's dive into speaking of what marketers are spending their time on. Let's talk about how the best marketers are becoming part-time engineers. And I know you have this as a topic here.

I'm curious to what that person said on X, but I agree with this. And the reason I agree with this is marketers now can create product and free products quicker than ever. If you look at how much people spend on advertising, in many cases, it's actually cheaper to release a product that a lot of people pay for your competitors that costs money and release it for free and gobble up a lot of visitors and audience that way than it does for just running ads on Google or Facebook from a profitability standpoint. I did a lot of that in the early days with Ubersuggest.

That's why we bought ads to the public. And that model's worked out really well for me. And I know there's a lot of other companies that have released free products to create their flywheel. Yeah, so it is the table that I just showed you, Neil, but let's talk about what we're doing right now, okay?

So I'll give you some examples. So whenever we build product right now, we say to the engineers, we say to people, hey, everything that you build, we need to have a marketable moment. So whether you're using a hyperframes video where it's like a 20 to 30 second video, and by the way, when they make one of those that I put up on my Twitter, 40,000 views, right? Yeah.

Um, so, and there's a lot of these hyper frame videos that have like a million views plus, right? There's a lot of these launch videos. So we want to launch video on a specific feature or product, or by the way, for our skills dojo, which we've given away for free, if you want access to skills, evals, loops, that type of stuff. Well, we're just going to, we're promoting other people's skills.

We're promoting our own skills. That gets a lot of reach as well. And then we can market to those people. So that's for free to click full API to MCP.

That's for free as well. Um, the API gateway, we're going to have a free version of that too, uh, Carrot, we're trying to give a free version of that for free. So we're just trying to give a lot of these free things out.

And that way people can use it. And then we can figure out, hey, who in here is actually a product qualified lead that fits our ideal client profile? And then we can reach out to them from a sales angle side and then follow up with them, right? And then just continue to follow up with them with different angles over time.

And that's working for us right now. But the key thing that we made a change in is, hey, Engineers, you're all marketers now. We're going to teach you how to make these marketable moments.

And we literally have a channel on Slack where every single week we're just talking about marketable moments because then that informs the salespeople. That also informs the paid media people on what we want to amplify. That also amplifies what we might want to do from an organic social standpoint too. So that's what we've done.

I'm curious to know what you guys are doing right now. We focus on one simple thing when we try to release product from a marketing standpoint. In other words, the product helps just build a bigger audience.

What are we giving away for free that people charge for and they feel value in that you can't really find for free right now? And if you can't answer that question, then what you're building probably won't get you the results that you're looking for from a marketing perspective. You can have the best email sequences, the best social posts, the best marketable moments when it comes to creating a video. But if the offer itself isn't valuable enough, it doesn't work.

And then you just wasted all this time building whatever you're trying to build. In conclusion... In conclusion, make sure whatever you're building as a marketer is something that people are used to paying for and they wouldn't expect for free. And the moment you give it away for free, they're shocked, they're ecstatic, they share it.

That's what causes virality. That's what does really well when you're building and releasing products as marketers. Yeah, and keep in mind, by the way, Seth Godin used to say this.

Basically, today's wow is tomorrow's average. And that changes a lot faster now too, right? I think it's, not Neil, but Seth Godin had a book called Purple Cow. And Purple Cow is just a cow that stands out, right?

You want to be that Purple Cow. And that Purple Cow is changing every single day. The target's moving very quickly. So you just have to be ready to adapt.

By the way, Neil, I have a little topic here. So do you want to know how Palantir land and expands for cheap? No, but I'm curious. Well, this was early days Palantir, right?

Obviously, they popularized the forward-deployed engineering model. So I was watching a talk with Joe Lonsdale, who co-founded Palantir, and he was talking to 8VC. I think it was like a VC firm. And the main thing that they did in the beginning, they just made sure that they got their tentacles in there in terms of the forward-deployed engineers working at these companies, right?

And then doing some stuff for maybe, you know, not too expensive to start with, but what ends up happening is because they end up asking for more and more over time, the expansion revenue just goes up to, we're talking about seven figures. We're talking about multiple seven figures. We're talking about tens of millions of dollars a year.

And what happens is you build so much goodwill when you build with these people where they just don't want to get rid of you. And Neil and I were talking on the phone yesterday. The thing is this, and this is from Neil's share, and so maybe I'll let you share it, but when you do a damn good job and you're so integrated, it's hard to get rid of you, right? So Neil, you can add a little more there from our conversation.

Yeah. So then when you look at it from a perspective of land and expand, if you get them with a company and you help them out with one product or service item, and then you convince them to do two or three more.

So if you help them with eventually multiple products, and then you do it in multiple languages and territories, whether they like you or not, it's not easy for them to rip you out.

And this is what they call in, at least in SAS switching costs or in business switching costs. Um, A lot of times it's hard to rip out Microsoft or it's hard to rip out Salesforce or whoever it may be. So you just stick with the solutions that are cheaper to pay for than it is to rip them out and switch them to something else.

Now, in that calculation, a lot of businesses also run the scenario of if we don't switch them out, what are we losing in growth or potential revenue or whatever it may be? But the moment you're just like, oh, the other software does everything that the current one we pay for, it's just half the cost on an annual basis. But to rip it out, it costs $2 million.

And annually, you only save, you know, 300 grand. You're probably not going to rip it out and take the multimillion dollar hit. you know, what Neil's talking about here, when we think about a moat from an agency, there's a book called Seven Powers. And the only moat, in my opinion, that a traditional agency has, at least for today, is switching costs, right?

Because if you're talking about you have, it takes a lot to get through procurement, you're a global organization, you're working with them in different areas, you have multiple services in there, it's hard to rip you out. Now, if you're talking about multiple integrations you have and things that you've built, it's even harder to rip you out, right? So,

That's what you want to think about switching costs. But then if you want to think about some of the other moats for our business, because seven powers, right? There's moats.

Someone, Gary Tan from Y Combin, argued that speed is actually a moat and that that's arguable. But when you look at Coca-Cola, for example, what do they have? They have brand power, okay? When you look at Toyota, what do they have?

They have process power. How do they create cars at such a scale? You can argue that Tesla even has that, right? Or you can argue that maybe they have a cornered resource, right?

So if you have all the copper in the world, then that's a cornered resource. Now, ideally, if you can have three or four of these, then you have a really defensible business. But in my mind right now, I think when you look at an agency, what can you do? Network effects would be another moat that you can have, right?

Like a meta, for example, has network effects, and that's not easy to defeat. And so the more moats you have, you think about a castle. In fact, Neil is actually in a castle right now. He's in a moat, right?

I am really in a castle. Yeah. Yeah, Neil's literally in a castle doing this. He just ran away to take care of his kids. And by the way, look, you can see there's a female knight over there in the background, right?

In fact, this looks like it might be World of Warcraft looking at the lion over there. Anyway, I'm going to continue to carry this episode while Neil's gone. So one thing I'll say is this. Oh, look, Neil brought his backpack over here.

Neil, by the way, are you in a World of Warcraft situation? If you're building an e-commerce brand, you should check out DTC Pod, hosted by Ramon Berrios and Blaine Bolas on the HubSpot Podcast Network. They speak with founders, marketers, creators, agencies, and platform experts about what it actually takes to grow a direct-to-consumer business from paid ads and influencer marketing to conversion, email, brand building, and consumer trends. I particularly enjoyed their conversations around scaling a brand without losing what made customers care in the first place.

Listen to DTC Pod wherever you get your podcasts. Real quick, if you want to acquire customers faster and more efficiently this year with the latest strategies and tactics, then check out singlegrain.com. That is my ad agency. Again, www.singlegrain.com.

Check it out. And if it seems like a fit, we'll get in touch and help you with a free marketing plan. No, I'm in the Lego castle hotel and whatever the theme is, but I'm assuming it says Spidey. Oh, okay.

He's in a Lego castle hotel. That's amazing. Okay, great. So I was just talking about the modes, but let me talk about one thing new, Neil. Have you used ChatGPT voice yet?

Have I sold that to you yet? Yeah, you sold it to me. I still don't use it, but you sold it to me. Dude. Oh, well, you can't do it with your kids. Okay, guys, let me explain how good this is, okay?

So ChatGPT voice, when I went to go see my dad over the weekend, I was like, you know what? I'm not going to listen to a podcast because that is mental masturbation, right? I want to get things done. So I opened my ChatGPT voice, which is, by the way, it's open on this computer over here so it can control all the apps that I have.

And it controls my other agents. I was like, hey, You know what? I want you to know the top three highest leverage things that I'm working on right now and I want you to tell me the next steps, okay?

I want you to tell me the next steps and then I want you to execute on them. It started executing on all those things and it kept moving the ball forward and that's the main thing. It kept moving the ball forward, right? And then I had a 40-minute commute to my dad and then 40 minutes back

And I just kept doing it. I kept working through it. My phone got really warm. And then what happened afterwards is I, on my way to see my dad the first time, I was actually worried.

I was like, oh crap, I'm gonna be late, right? And I just said to the day, I was like, hey, can you text my dad? Tell him I'm gonna be about five minutes late. And then it sent a text message and then it confirmed with me that he responded and he received it.

And I'm like, that's good. Let's continue to get back to work. Now, this is really big because you now have like a Jarvis from Iron Man. You can tell it what to do and it knows your context.

And not only that, I can control my other harnesses. So I had my Herbie's desktop app open too. I was like, hey, I want you to continue work trees on that one too. What else can you continue there?

And so... Because a lot of this stuff is about, sure, people talk about setting up loops and graph engineering where you don't have to prompt it. But in many cases, you still have to kind of check on the thing and approve it over time and make sure it's not just YOLOing and doing whatever it wants. Maybe you can do it with code, but when it comes to marketing, when you're sending like sales emails or when you're sending like recruiting emails and things like that, you have to have some form of human in the loop there.

And so that has been amazing and it's only going to get better. But you do know for the, at least the texting example with your dad, you can just talk into your phone and it just does automatically anyways. Right. I'm too late.

But by the way, I'm still working with the thing. And so like, I can just, as I'm working, I'm like, by the way, send a text message. Right. But by the way, I sure I can say, Hey Siri, send, send a text message, whatever.

But it, can I confirm that he received that? He sent the reply to me. I can't do that. Right. But with this thing, it's like, you probably can. Have you seen the new Siri?

No, I haven't. Tell me about it. I haven't used it either, but they say it's really good. You can't promote something where you haven't even seen it yet.

No, but I read about it yesterday. You probably can, right? But they're so far behind. But I think this is going to become the default way, Neil, where actually now Hermes came out with something where you can replace your Amazon Alexa with its voice because Alexa is really stupid right now.

It's going to get smarter over time, but it's really stupid right now. It screws up so many things, right? So imagine you wake up in the morning, Neil, instead of asking, how's the weather? How about how's the weather and knock out my top three things right now?

That's how the weather is. And by the way, with technology being so advanced, what I really hate about all the apps, whether it's AI or Siri or Alexa, if you ask how the weather is, it can be wildly inaccurate. Like there's been so many cases where it rains throughout that day and the weather apps will show it's not raining at that exact moment. And I'm like, man, you guys can't get this basic thing right now.

Yeah, and it's 2026, and I'm sure they're going to get it right. I think it's just, but I just look at ChatTBD Voice, I think that this is going to change how we work, and we're just literally going to be dictating. Because here's the thing, on that commute back and forth, had I spent 80 minutes to 100 minutes listening to a podcast, well, that 100 minutes now, I actually just got work done. And so that is a new expectation.

You're not going to be listening to podcasts as much because people are just going to be working more, talking more to these things. Yeah. I've never done the podcast in the car.

I've tried that out. What do you do? You probably call people, right? Uh, so I've done kids podcasts where like it's trivia for kids. I forgot the lady that does it, but it's, it's, it's, she's all about trivia for kids.

Um, I can't do the voice Siri thing cause my, or voice anything cause my kids will chime in the moment I'm doing any voice thing in the car. Like they'll ask, uh, you know, Tesla, you know how you can ask questions. My kids love it. They'll ask animal facts or random stuff.

Um, so I either do, phone calls or I talk to my children. Because with phone calls, even if they speak, the person on the phone knows it's me when I'm talking. So I can still get a conversation going and I can be productive and get stuff done through that way.

It is hard when it's not a phone call because most voice assistants, if I'm talking, will pick up their noise and they'll just get a kick just trying to throw me off and put in other stuff that they're not supposed to put in.

Well, yeah. I mean, you have to be engaged in a different way because you have customers in the car or kids in the car, I should say. Yes. You have to entertain them.

By the way, Neil, so I mean, on this note, okay, ChatGP Voice is great, right? But then I started, I installed Jack Dorsey's Buzz, okay? So I want to show you how I'm working with Buzz right now. And I think this is going to change how work is.

And to be frank, I've already been working like this inside of our Slack. I've been talking about having agents inside of Slack. But this is pretty cool. Like anybody can install Buzz.

I'm sharing the chat that we have over here. This is like Slack for agents, okay? So I'm making this world of business game in my free time, okay? Where you can basically, Neil, because we have this thing now and this is open source, I can actually finish off that game now where if Neil completes a quest, I give him $20 in stable coins, right?

Or if he does, if he implements a workflow and one other person uses it for the next four weeks, maybe I give him $500, right? And then you have a leaderboard for it. So I'm working on that right now with this thing and... I made this plan and then I have all these four agents in here, right?

I was like, hey guys, I want you to review these plans and give me a different perspective. And they just start working in here. Here's how it is right now. Here's my perspective.

And then they start talking to each other, Fable 5, Cloud Code here. Here's what I think over here. Boom, boom, boom, boom, boom. And it just gets work done.

So imagine I can use a chat GPT voice. Plus I can use this and say, hey guys, just do all these things over here, right? Bam, bam, bam, bam, bam. And then now we can do way more work.

So to clarify for anyone listening right now, if you watch the video version on YouTube, you'll see what Eric's talking about because he's screen sharing and showing you. So make sure you check out the video version to get some context. But speaking of talking in the car, dude, I was driving to Legoland today and I was talking to my CFO. His name's Paul.

My daughter, once you realize it's Paul, because he knows he's a finance guy. He's like, hey, Paul, do you want to make some charts and diagrams and work on today's projections? Yeah. that's funny how do you react he cracks up but he knows because she'll try to call like she knows the password for my phone both my kids do and like they'll text too they'll be like what's this month's numbers and like I don't talk like that so they kind of know because it'd be like hey what's the growth of this month or what's year over year I'd be more specific that's exactly how you talk

How are we doing with numbers? I wouldn't say this month's numbers. I would say, did we hit this month's numbers or targets? I wouldn't just say like, what's this month's number that's too generic.

But either way, you know, they get a kick out of it. Yeah, like they know the type of people I'm talking to and their roles and they'll say different things. So like with Paul, it'll be charts and diagrams and projections. With Mark, it'll be, did we hit the sales numbers?

Because I always ask him that. I ask him for country to country. My kids don't know all the countries, so they'll throw out random country names in Europe. And some of them are cities because they don't know the difference between a country and city for all of them yet.

They're young. And then like, if it's Mike Camo, they'll be, they'll talk about like, Hey dude, how much money are we going to make this month? Or taking salary this month?

That's funny. By the way, You know, it's, did I tell you, I was sitting on that AI operator's call. So again, you mainly, you have a lot of people that are working at these larger companies, right?

You know, nine figures plus. And initially all the CEOs were like, how do we cut headcount? How do we cut headcount? How do we cut headcount? But all these other people are like, well, no, they had to kind of,

walk the CEOs away from that language. And it's more about how do we enhance what we're doing. And I think that's really important, right? The messaging, the way you communicate something is really important too.

And so I'm not saying Neil wants to cut headcount necessarily, but it just reminded me of how quickly we change opinion on this AI stuff. And one thing I want to bring up, I was listening to Gavin Baker talking today. So you know Gavin Baker, the investor, right, Neil? Yeah.

So he's actually a popular investor. You know Brad Gerstner, right? He's kind of in that circle. They podcast the stuff and they're really good investors.

So Gavin Baker was saying this. And I didn't know the number was this small. How many people in the world would you say are actually super AI pilled? I don't know.

Fraction of a percent. Yeah, so it's like 0.007%, okay? So according to what he said. So he says that it's likely that 500,000 people in the world are super AI pilled right now, if not like maybe even 250,000.

So if you divide that by 7 billion, it's 0.007%, right? It's a very small number. And Neil and I were talking about this yesterday because even though I would consider that

I'm in some pretty AI-pilled groups. I like to talk about this stuff. Neil's in a lot of these groups as well. Both Neil and I, my CTO, other people, and I was talking to someone else today.

He works at one of these companies. We can say less than 10 people we know that are super AI-pilled. So what does that mean exactly? If 250,000, 500,000 people are AI-pilled, what that means is we're gonna need way more compute, way more electricity, way more a lot of things, and agencies are not dead, okay?

The harsh reality of AI right now is maybe we're just in the second inning, maybe, and that's it, right? All this talk about, oh, agencies are going to be dead. No one needs ad managers anymore, this and that. You can say we're biased, but I just, I don't think you can say that for a lot of these roles.

And a lot of the roles that people don't want to do anymore, we're just, they're going to be reskilled into other areas. So harsh reality, only 250 to 500K, that puts things into perspective.

Yeah. And that means companies are going to have to pay others who are experts. This is why things like service-based businesses or agencies are not dead. And I see them lasting for a very long time because people pay other experts and you could say, well, why don't they just hire those people internally?

Well, I can ask you the same question. Why don't people hire amazing SEO people internally 10, 15 years ago? You know, Some do. Some pay external contractors from Upwork and places like that.

Some hire agencies. Some hire companies where they place headcount into their organizations temporarily, like Tata or whatever it may be. There's another one from India. What is it called?

Infosys or something like that. It could be off on the name, where they'll place temporary employees within your organization. So there's a lot of options, but I believe if you're really good at what you do,

people will continually pay you. If you suck at what you do, people may temporarily pay you. Yes. And I also think that it's never going to change that hard talent is hard to come by, right?

Because hard talent is always like a percentage of people. And that's never going to change. So anyway, I just wanted to call that piece out because there's a lot of doom and gloom all the time. And sometimes I even call that stuff out.

So that is it for today. Please don't forget to rate, view, subscribe. And yeah, we'll talk to you tomorrow.