Build a Better B2B Growth Engine with Uzair Dada from Iron Horse

You're listening to The Dave Gerhardt Show. Exit. 1, 2, 3, 4. Exit. 1, 2, 3, 4. Exit. Hey, it's me, Dave. That's me. I'm Dave. Before we get into today's episode, I want to give a quick shout out to our friends at Webflow who are sponsoring this episode. Webflow is the agentic web marketing platform for high-performing brands. Their platform lets you build fully custom websites that rank in AI search and make your website your biggest growth engine without being dependent on a developer. I know that because we use Webflow at ExitFive. No developers here. This episode is part of a 3-part AEO series that we're doing with Webflow over the next few weeks. I've been having a bunch of conversations with practitioners who are deep in the weeds on AI SEO and web optimization. We talk about what's working, what's changing with real examples you can go and apply to your job right now. In marketing, Webflow also put together this, uh, free AEO assessment. You can check it out after this episode. We'll link to it in the notes here. AI is changing how buyers discover and evaluate your brand, and this resource will tell you exactly where you stand. You'll get a personalized score showing how your site performs in AI search today, visibility into how likely your brand is to be cited by answer engines, and a breakdown across your content, technical foundations, authority, and how to measure where you show up so you know where to focus. We'll put the link in the notes here. Thanks again for Webflow to create this series with us all about AEO, one of the hottest topics we've had in marketing, a big appetite for it. And I know this is going to be a great interview. So let's get into this episode. So this one company was spending big on LinkedIn, Meta ads, and Google ads targeting over a million people. But when they actually ran the numbers, their real audience that they should have been going after was only 20,000 people. That's 98% waste money that could have been reinvested elsewhere. Uzair Data is the CEO of Iron Horse. A growth marketing agency he's been running for 26 years. Just to be clear, he wasn't the guy wasting the money. He found it. He's focused on enterprise B2B and he's in the room with CMOs at companies doing $100 million plus in revenue. And what he's seeing right now is actually a massive gap between what's possible with AI and what companies can actually implement in the enterprise. Most of the IT and security teams are saying like, no, no, no, you can't do this yet. So we got into his growth framework, which boils the entire B2B engine down to 3 steps: get discovered, get chosen, and then close. And we talked about why most companies are advertising the wrong audiences because they never did the foundational work upfront. We talk about how to actually show up in LLM results and why the second query often matters more than the first. We talked about why he blocks every Friday afternoon from 2 to 7 to build with AI and how that habit has spread across his entire company and his take on why brand versus demand is a false choice that marketers invented just to argue about. I love that one. The back half of the pod, we get into AEO, what's actually working for enterprise marketers right now, and a reality check on AI adoption that most of LinkedIn doesn't want to hear. So let's get into it. This is my conversation with Uzair from Ironhorse. Okay, Uzair, good to meet you, sir. We just were in the back in the green room just reminiscing about the good old days, and a lot has changed. You were talking about Drift, and I, I have a duffel bag that I got for like my 2-year anniversary there that I still take everywhere. And every now and then somebody be like, oh, that Drift logo is amazing.

It is amazing. I was just telling you, I was reminiscing about the black on black t-shirt. I still wear it, you know. It's cool.

That's a good one. That's a good one. And you know what's funny? It's so like, so many of the lessons from that era, I actually think— if it looks like I'm distracted, I type notes during this pod, so don't get mad at me. I was doing, hosting a webinar one time, and the lady was like, look at the host, he's not even looking at the screen. I'm like, do you what it's like to have all the— you know, you gotta— we gotta run a tight ship. So Uzair, you're the CEO of Iron Horse, B2B demand gen agency. Is it true you've been working on this company for 20+ years? Am I reading that right?

Yeah, I started Iron Horse 26 years ago in January. So yeah, it's been a minute. It's been a minute.

Dang, what keeps you doing that for so long? I feel like, especially in the world of marketing, it's like, I'm gonna do something for 2 years, I'm gonna do something next, I'm gonna do something for 2 years. What's led to the longevity and especially in the world of B2B demand gen?

Yeah, I think to me it's the love for sales and marketing as a core consistent theme that's existed. And then sort of I think we've gone through like 4 or 5 transformations in that 25-year period and the ability— and I started it, so ownership has some privileges, you know, you can pivot and change and do what you want to do and do what people you like doing it. So I think that's the thing that keeps you going and Definitely. I think we're living in a super interesting time that kind of dwarfs everything we've seen over the last 20 years. So it's kind of fun.

Is it a big agency? I actually have no idea.

No, we're small. We're about 35 people, all based in North America and super, super focused on the growth side. That's kind of all we're focused on. Okay.

I got a bunch of questions I'm gonna ask you about the growth side, but just, I'm already going off script because I'm sure you've seen this. The narrative now is like best time ever to start an agency because the headaches of having an agency are gone. Is that BS? Do you buy into that? Are agencies hot now because you can do a bunch of work with AI? What's your point of view as someone who's actually been doing this for a while?

To me, what is changing is kind of reducing latency that came from all the functional handoffs. So I seriously feel like every time I try to do stuff, before I had to go to different people to get answers. Not that those people are not relevant in that conversation anymore, but the ability for me to just get shit done on my own without waiting for the latency and waiting for the design team to finish something.

Or totally, dude, I feel, I feel you the most on this one. This is what I'm saying is like my career, I came up and I like got promoted to be the VP of marketing because I was the good at marketing. Yeah. And actually a lot of the things that then I'm like, wait, my whole job now is like team management, performance management, HR. I'm like, don't you want me to be the guy who's like writing the landing page copy?

Does it? Exactly. So to me, I feel like unshackled. That's the term. That's the term, but I truly feel like it. I can just go do now. And I started blocking like my Friday afternoons, my build day. So I'll have 6 hours from like 2 till 7 that I'll just crank and build stuff, which I could not even fathom doing.

Okay. What have you done on a Friday? My problem, by the way, is I block a Friday and then I just use it to do all the things that I didn't get to do during I started getting pretty religious about it over the last 4-ish weeks. All right, good. So tell me, what are you doing? What's the plan for Friday?

Yeah, so kind of the last few things, and we're sort of jumping ahead, so I'll go into it. The kind of the big thing that I was trying to think about is like, I spent a lot of time meeting a lot of companies. I kind of sort of context switch 8 times a day because I'm talking to different customers, prospects, partners, whatever else. And before, you were running from meeting to meeting to meeting. You were barely researched, you were barely rehearsed about kind of what you want to do, what you want to get accomplished. And you sort of went in there and sort of kind of had your domain, but you winged it. The ability for you now to get prepped, research, not search, because we all were doing search, not researching, analyze, and have the ability to digest that is so much simpler and easier because you can connect things. And harness information. So one of the things I did, and I'm happy to show it to you if you want to see it, is build out sort of this kind of notion of an account dossier. So what it does is it sort of takes into account a lot of different things we have. So it pulls out stuff from HubSpot, pulls out stuff from Clay and ZoomInfo, pulls out stuff from my Granola notes, pull out stuff from my Gmail and G Drive stuff. And based on sort of what we do, gives me kind of a summary of the what I should be thinking about going into the conversation based on the org or the person I'm talking to. My ability to be impactful, my ability to be contextual, my ability to be curated in that conversation literally has gone up by like 10x. Okay.

How did you build that? How did you stitch all those systems together?

You wanna see it? Sure. All right. So I'm gonna share my screen. We're gonna go deep into some lab crazy stuff. So let me pull up. So I'll show you what it is first, and then I'll tell you kind of how I got there. Can you see that? Yeah.

Uh-huh.

So this is just kind of Altira, one of the companies we were talking to. So this is the output. So this is Rev 1. I'm now on Rev 4 in beta build mode, but this is sort of what we are using across the board. So it kind of looks at who the company is, kind of pulls out all the information from public data or, or whatever is available. Kind of gets into strategy context of what their core strategy is, because we as marketers love to go into marketing strategy without understanding what the company does, right? So let's focus on what the company does. Then kind of looks at from a messaging and go-to-market perspective, what are they doing? It's got a bunch of things that we look at in here. It looks at sort of the website and both technically and from sort of a kind of UX/UX perspective, it looks at the tech stack. Looks at sort of your AEO, SEO, CRO type technology structure, what you're doing. It's got a bunch of thinking behind it. Looks at kind of what their people are doing from a paid perspective. Looks at what people are doing on the tech stack side. Is it connected? Is it not? Is it modern? Is it not? It looks at sort of the marketing team and how the marketing team is structured, who's there. And then it kind of goes into areas that may be interesting for us to talk about. From there it says, What may be interesting for us to offer them as a company or talk to them about? So it's this, this is a really cool structure. And then it ends with things that we might wanna dive into more. And all this is coming from a host of different sources. Makes sense?

Yeah. Tell me some of the sources, like how are you mapping org charts and getting website traffic and paid data?

Totally. So this is sort of the stack that it's sort of built on. So you can see at the bottom, like all the data sources it's pulling from. So the org stuff and everything else is coming from ZoomInfo, from Clay. It's pulling in a lot of different data sources that we have. So a lot of our first-party data all sits here, and then it's using just a boat ton of sort of the surface area is using Claude and then kind of hooking everything up through Claude to go build this. And then the orchestration layer is all this kind of the manifestation of it. But sort of, it has a lot of things that have gone in and that it has a lot of governance architecture built in, in terms of how we wanna see our brand guidelines, our methodologies in some sort of, some shape. So this is Rev 1, right? And from Rev 1, where we went being good smart marketers and having a lot of time on Friday afternoons is 2 to 7. I literally build out a roadmap of what I wanna do because I saw it. I'm like, this is so effing cool. I wanna do more. So I built out this crazy roadmap that I'm working on now as the project of all the different things that I think we could evolve across all of these areas. So this is sort of our Iron Horse marketing brain. Nice. That we're sort of building out. From there, I kind of went down to, well, let's not build until we've thought through it. So I actually started building a product requirements doc. And so I've got a full-blown PRD that I'm sort of working on, sort of tweaking because you can build very easily, but then it's a shitty output. And so kind of started building out the PRD. PRD led to sort of what's my tech architecture? From the architecture, I was like, okay, this is really cool, but why am I just building it for me? I should really build it for us, our company. So from there I said, hey, what would it look like if I build this as an app that everyone in the company could use? So then I prototyped a UX, and from a prototype of a UX, I now have a local working prototype on my website. On my desktop that right now looks at all upcoming meetings the day before and makes sure if the company is new or not for me or anyone on our BD and account management team.

And sorry, are you on all sales calls? Is that why?

I'm not on all sales calls.

No. Okay.

But I'm saying me and anyone else. So I basically built it, but now I'm sharing it across the org.

Yep.

So since I built it, it's still locally running on my core instance. We haven't deployed it across everybody. That's the next step.

You got it. You want to make this a web app that anybody at Iron Horse can go access. Exactly.

So it's got SSO integration with our Google Enterprise account, all that good stuff.

So did this all originate with Cloud? I'm sure there's multiple tools here, but was this originated in, you said Cloud, is it Cloud Code to help?

It did. It did. Yeah. So it's Cloud Cowork was where it started.

Cowork. Okay.

Yeah. So we started in Cloud Cowork as a sort of a, just a brainstorming idea on a fun Friday afternoon that became really cool. And then sort of just in since then has just taken a life of its own.

Yeah. This is cool because I think there's a lot of AI use cases that like play really well on on social media but are kind of like dumb and don't end up getting used. But if you think of something like this, it is— man, like, I remember even just as recently as a couple years ago being like, oh my God, having access to Gong calls was like a game changer.

Huge, huge. I still think that's one of the coolest things that is underutilized.

Well, look, I was, you know, before I got on with you, I— we were just in a team meeting And what's really interesting is I feel like because I know that everything is being recorded for a purpose that we can use now, I'll be like my Dave self, but then I'm like, hey, hold on, I actually want to like speak this into the record. And I'm like, here's how I would position this. We should go do this thing because I know that after that call, like Allison's going to take that transcript from— we use Fathom and she's going to take it and create an action with it. And so it's like it is this entirely different way of working of like putting it onto the call like this. But this idea, it's like every company, every role now can have this kind of like customer brain or company operating system of sorts where like, let's collect all the stuff, you know, website data, product data, customer data, calls, transcripts. You know, you're an agency, so you want to have your methodology applied there. It's like, man, this is— that's cool beyond like, look at a silly picture of Dave from Nano Banana.

Yeah. And our whole notion was So the first thing was like, it was just a prep thing for me.

Yeah.

Then it became a prep thing for everybody.

Sure.

Now it became automated, so it does it for people. And now the next iteration of this is there is a huge disconnect between BD and sales teams having customer conversations and then handing it off to the account teams to deliver. And there was a big disconnect.

Well, it's like, you know, software product development 101. Almost always the good tools come from when you solve your own problem first versus being like, what cool app can I make? Versus like, I, even in my conversations and with our team, I'm like, I'm trying to put my blinders on to like what other people are doing with AI. 'Cause it's part of me is like, I want to know that stuff. And so it gives me inspiration. But the other part is like, let's just kind of have this first principles thinking of like, how do we default to like, what problems do I have in my business? My business is unique to your business, to others, but what can I solve in my business with AI? And then what you're doing is like, okay, I'm literally, I'm on a call right now with the CEO who's been running a growth marketing agency, super successful for 25 years. And you're— here's how to drive AI adoption within an org. You're the one leading the charge. You're not sitting back, like, not doing this stuff. I think that's really important.

That's a huge lesson. So, you know, one of the things that kind of from there, we were patient zero, client one, for sure.

Sure.

And one of the things we did for adoption, remember, we've been sort of at it for like 2 and a half years, and there were a few people that were incubating, and this is dogfooding. And then we try to say, how do we systematize this across the org and drive adoption? And it was starts and kind of stops. And one of the things that led to the unlock was about 4 or 5 months ago, we institutionalized for our leadership meeting. Every leadership meeting starts every 2 weeks with every leader showing what they've built in the last 2 weeks. So we don't start with where the company is, where the business is. We're saying, what did you build? And we spend like 45 minutes to an hour looking at what people are building and thinking. So There was a little bit of competitiveness came in, a little bit of pressure, peer pressure came in, but that was amazing cuz that get everyone starting riffing off each other. Now the next layer is those leaders now are doing the same thing in their management meetings for their own teams. And that's been awesome because it was not just bottoms up, which we tried doing where we sort of said we have to drink the Kool-Aid, we have to believe, and if we don't, it's not going to stick. And that's been a big unlock for us. And really, I think a really cool way to sort of institutionalize that across WORX.

Okay, so I have a bunch of— I want to talk about AEO with you. I want to talk about your growth marketing system. I think that's really relevant. But yeah, just using this as a jumping off point. So that's a useful example. You showed most of my audience is, you know, B2B marketers. Yeah, in-house. Some of them also work in an agency. They're listening to this because they want to get smarter about marketing.

Yeah.

Based on that example, how would you talk to them? Maybe it's the same story that you talk to your company, but how would you tell them, as someone who's seen the shifts in B2B over the last 25 years, what do I need to be paying attention to? What matters? How do I separate the noise from the reality right now? Like, what should I be thinking about when it comes to what's changing in marketing and AI?

Well, I think the biggest thing to me, it always starts with literally Understanding who your customer is and what are they asking.

Actually, wait, are you a techno optimist? Do you believe marketing's gonna go away? Is everything going to be replaced by AI? Is SaaS going to go away? Like, I'd love to get that. Let's frame this first with like, what's your opinion there?

I think the new technology is giving marketing and marketers superpowers. So I don't believe marketers go away. I think marketers evolve.

But you see these new apps and stuff that come out, and I saw one on X like 2 weeks ago, it's like replace your CMO with Claude, you know, replace your CMO with this thing, or like replace your CMO with Claude Code. Is that a bunch of bullshit, do you think?

I love the one which was replace McKinsey with these 50 Claude prompts. It was awesome. It was the latest one I saw the other day. So, you know, the clickbait is at next level now. It's like freaking awesome.

Okay, so how do you calmly like sort through that? How do we help people?

I think that there's a lot of cool stuff that's happening, but The things that are changing are no longer is dollar currency the differentiator between good companies and bad companies, because tooling is becoming democratized, right? So everyone's got access to pretty much the same tooling. Now, naturally, if you have more credits and more things, other things are possible. Everyone's got similar surface area to use, which is pretty awesome. But the difference between good to average is taste and judgment. And people who have really interesting POV inside the companies or at other agencies or wherever that say, here's what's unique to us, here's what I can do that is different than others. And taking those new surface areas and applying them are doing incredibly cool stuff. And I think that's the difference to me. That's the big unlock. So it's the ability to not get scared. But jumping in. One of the things I told my team is if you think you're doing good, you're not failing enough. Literally, because good is changing every day, every week. And so you sort of have to be in it and be immersive. It is scary initially, but once you get over that, it's like my whole notion is, oh my God, I couldn't do this yesterday. And that to me is such a rush of being able to be doing something that was not— it's not that it's just faster, it's something I could not do.

Okay, but if that is the case, why would you— if I'm a founder, like, why would I hire a CMO? Why would I hire a marketing team? What unique value can they play? Isn't it just going to be my AI agent for my company is just going to own talking to the rest of the world about what we do, to that AI agent, and all the buying decisions are going to just be made through that agent and none of my job's not gonna matter.

No, but I think the question is, how are you showing up? Where are you showing up? What are you doing to go show up? Even if it's showing up for agents, it's not that you just automatically just appear, right? It's how you communicate. Do you have a unified voice of how you show up, how you talk about yourself? There's a lot of technicality that goes in. I think we're sort of— marketing is becoming akin to somewhat engineering. We're becoming builders, we're becoming marketing engineers. I won't call it go-to-market engineers 'cause it's more than it, but it's much more structured, much more systematic, much more first principle systems thinking around making it happen. So I don't think it's a, I just was doing something that was a single function that carries over. We're sort of becoming much more cross-functional and the taste and judgment, like a cool example is there's a company outta Australia, a studio called Eye Candy, AI Candy, and they're AI video studio. And I saw that stuff and I was talking to one of my friends who's a creative director and I said, this is incredible. We're doomed because if AI can do that, we don't need creative directors anymore. And she goes, you know who's enabling that? It is the best creative directors in the world that are actually telling the AI what to do. So tooling is tooling, but the taste and judgment is what makes awesome happen.

It's an interesting per— and by the way, I'm just provoking because I totally, totally— I believe this. I want to have marketers to continue to have jobs. I want to help make the case. And I actually think that when I put my phone down and I'm not on X or LinkedIn and I have a minute, I think a lot of this stuff is actually 3 to 5 to 10 years. I saw some new data today that like the majority of the country, yeah, America is actually revolting against AI. And so there is going to be some like— maybe not majority was like, you 46%, but I'm a marketer, I round up. So, and then if you think about like, okay, our audience is B2B, B2B buying is pretty complex. Even if your AI agent is gonna do a bunch of the initial research, I do feel like you're not gonna, oh, that wasn't me, I didn't make that decision. My AI agent brought in, you know, IBM. So like, don't fire me if this doesn't work.

What is that?

Yeah, no, but to your point, so I was gonna say like, RIP to Sora because ChatGPT shut this down, but I was not good at making those funny— there was that stretch of time where everyone was kind of making those funny little like AI Sora-generated videos. The people that were really good at that were like people who just have ridiculous ideas. And like, I am not— I cannot think of a cat on the moon with Elon Musk, you know, like I don't think of those types of things. So maybe that is the point about having taste is like coming up with ways you're going to be sitting in your like super lab being able to prompt anything. But that is the skill.

I think that is a skill. And as a result, the noise level is just going to 500. And so that taste and judgment becomes even more important. So who are going to be those curators? Yeah. Who are they going to be? Those experienced people that can sort of use this and harness coolness that comes out from it. So I feel like our role is going to evolve even more. And I think if you think about The way GTM is happening, and as you said, agents are going to talk to your buying agents, are going to talk to your selling agents, yada yada. Like the journey from getting discovered to getting chosen is changing. And to me, sort of things we have to do is kind of go back to the simplicity. I think we as marketers have overcomplicated the shit out of marketing. Over the years. Oh, and I agree, we just argue about stupid shit that's just completely—

okay, I've never talked to you before, but like a story that I tell all the time is like, I live in a small town in Vermont and it's pretty rural area, but there's this one sandwich shop down the street and man, she crushes it. And she does good marketing, they have good Instagram content, and I'm like, you know how she measures the success of their business? Like Did more people come in the store or not?

Exactly. Exactly. Yeah.

And like, who knows why? I don't know if it's like because of all the, you know, martech vendors and attribution and the fact you can make the case B2B buying is more complex. And so it is. But I don't think she cares which touch it was. And I do think that, yes, like, God, I love the simplification. It's like that bell curve of like when you start out in your career, You try to measure everything and justify everything and sound way smarter than you, and you realize like, actually, no, it's kind of like my friend Pranav, who's the founder of this company called Paramarkt. He's like, don't treat customers like— let's not stop treating people like they're morons. Like, if you do good marketing, people are gonna find it and show up, and they're gonna tell you that they heard about you.

Yeah, to me, I, I use this a lot. I say brand and demand is a figment of marketers' imagination.

Right, right. Is that because— so that's the punchline, but why is it? Because they're not different.

Well, no, they are different, but we sort of over-index on this or that, right? Not this or that. It's a structured approach. You know, another good one that I just saw, Scott Brinker came out with this sort of composable stack thing the other day, and John Miller posted around attribution, and he goes, don't use attribution to prove marketing, use attribution to improve marketing. And I thought that was so great, right?

So like perfectly, he's good at that little nuance in the words right now. Yeah, I did a podcast, uh, the last one that came out before this one will come out later, it was with this guy at Ramp, Drew, on the data science team over there. And he said basically like measurement is the number one thing that kills good marketing before it even has a chance.

100%. And so we sort of, we've been talking about alignment for God knows how long between sales and marketing. We've been talking about attribution and kind of how things happen. We've talking about kind of KPIs that have been pulled up from the marketing teams and sales teams into the board levels and they ask for stupid MQLs and SQLs and other things. It's just, especially as you go to enterprise and market, Dave, it's just so broken. It's just, those are fundamentals that have to be revisited. And to me, I think where people need to start is sort of need to truly start with good data sets and alignment on what matters.

And I think, yeah, maybe, so you have, I was gonna transition it to ask you about this. So you have this kind of growth operating system. Your focus is specifically around enterprise. Do you want to just kind of talk me through your view of the world? I think this would relate to everyone listening to this.

Yeah, totally. So, you know, the way I kind of started talking about it is to me, I'm saying if in our conversations, what I'm having with CMOs is, and practitioners, is simplify what you do and kind of anchor everything underneath. So start with get discovered, which is 75 to 80% of your journey, which is truly right now pretty much off domain. So it's your AEO, SEO stuff. It's your demand gen stuff. It's your AR, PR, community strategy stuff all combined. Historically, all we did with when we wanted to do more stuff was spend more on paid media. It doesn't work anymore. The second part, which is sort of your new middle funnel, is your website. That's where the selection happens. And so if 70 to 80% is off domain and the next 10 to 15% is your sort of website. The website needs to have a bridge from outside to inside. That means we need to make discoverability super easy. That means that we need to make conversion optimization kind of point built into how we structure those journeys. It all starts with understanding your audience. It all starts with refining your audience. And today, most companies are very lazy about doing that. And as a result, everything waterfalls and it's diluted. So we have a company we were working with that was sort of spending a lot of money on LinkedIn, Meta, and Google Ads. And for context, was over a million. When we sort of did analysis, taking intent signals from their first-party stuff and using Clay and Zoom, and that they should be advertising to was actually 20,000 contacts. Just think of that for a second. 98% wastage.

And I bet you that inside that company, the marketing team was stressed out, overcommitted. So many competing priorities, you know, tale as old as time.

Yeah, their budget was healthy, but their budget, if you broke it down to sort of, and from an accounts perspective that they were trying to reach, their spend was $16 per account per quarter. I'm like, dude, that's like price for one click on LinkedIn. You can't do it. You're trying to spray and pray and it's not going to work. So if you just simplify that and we started, let's go back to your starting conversation of Gong transcripts. Between your sales and customer success transcripts and meetings, you have a gold mine of insight. Gold mine, right? If you use that to sort of decipher what conversations are happening, what questions are being asked, what things you need to be talking about, that goes into the conversation arc that companies need to be building.

Yeah, this made me think of— we were talking about Drift offline before, but one of the best pieces of content that we created when I was there was we found the top 5, because it was a new company, kind of challenger brand trying to, you know, replace the status quo, whatever. We work with our sales and CS team. We said, give me the top 10 reasons that people either don't buy Drift or cancel.

Yeah.

And we published them.

Yeah. Yeah.

As marketing, no spin, just like boom, boom, boom. Question, answer, question, answer, question, answer. And it was like, it's not some crazy sexy campaign. It's like help people cut through the noise and be like, yeah, how, okay, how is this different than Intercom or Marketo or whatever? Okay, cool. Yeah. All right. I'm gonna take a demo.

You were feeding LLMs before LLMs were a thing, man. You were using—

I know Claude was trained. Claude was trained on me for sure.

Facts. No, it's true though. Like that's a fundamental scenario, right? And You were brave enough to do it. Most companies are very scared about having the conversations that need to be had, whether they're competitive conversations or someone calling my child ugly outside and I don't want to confront it, right? Sort of, they just shy away from having tough conversations. And I think that's what you need to be doing in this new world. So anyway, going back to sort of the get discovered part, the get chosen part is all about sort of website optimization. And then the last part is sales, which is all about context.

Sorry, so is this like a 3— is it get discovered, get chosen, sales, and then close?

Get close, close.

Get discovered, get chosen, close.

Yes. And to me, those are the 3 things that get discovered. It's just sort of your— to me, the artificial wedge of a pizza slice that we built as a funnel has really become more flat. And the channels that we have, or the tools we have to influence it, are broader than what we perceived it to be 3 to 6 months ago.

Yeah, I'm just a fan of simplification always, like we talked about, you know. And the first boss that I ever had that made me understand marketing, like, put me on the whiteboard and drew like the AIDA funnel, you know, awareness, interest, desire, action. And I was like, yeah, oh, I get what marketing does now. And I like this, like, Okay, your job now is to get discovered, get chosen, and close. And also you can look at each one of those stages and be like, well, we're pretty good. A lot of people know who we are. We don't need help on the discovery piece right now. Actually, the biggest bottleneck is getting chosen. And so that's where we're going to spend time. I like it. Simple.

Yeah. And underneath it, there's the tooling is changing, right? We've got a lot of interesting AI-native apps coming in. You got the Clays of the world, the Profounds of the world, the Webflows of the world that are doing some really, really cool stuff that make doing some of this stuff a lot easier. But when you go into enterprise, those are exactly the constraints. You're dealing with a lot of legacy stuff that you need to figure out what to keep, what to replace, what to upgrade. And so I think the stack is changing. And then the cool stuff is that everyone's pondering about is what's going to be your surface. Is it Claude Co-Work? Everything else is just a data source underneath. That you sort of use and plug in.

Sorry, what do you mean? What's your surface?

How are you going to interface? What's going to be your interface to kind of go get the work done tomorrow? Today, a lot of what I do starts and stops in Claude Cowork.

Got it.

And so the question is, do I need to have HubSpot as my surface that I'm going to directly interact with, or I'm going to just use the HubSpot API as an MCP server through Claude as an interface to go do what I want to do?

Yeah, you know, it's funny, I just saw this question earlier. When I was going through my emails, but someone in our community said, I want to create a Claude skill built around my HubSpot MCP server so I can run reports on campaign performance. Anybody have experience with this analytics specifically in HubSpot? It's a chink in my armor. I'm hoping I can set this up. And so it's like, there is kind of this world now where it's like, maybe all these SaaS companies aren't going to go away, but like, finally my dream of never having to log into Salesforce again can come true because I can just interface with Claude, right?

Yeah. And I think there's still a lot of companies. I think there'll be specialized operators in these tools that do some things, but I think it'll become way more accessible to the masses than it ever was before because the abstraction layer is there to actually use the information. So I think some of these will become sort of more the context layer and the decisioning layer. And then I think your interface of consumption will change. Yeah, for different people.

So I had this in my prep was like to talk to you about this like growth operating system. Did we hit on that?

Yeah, I think so. I mean, I think there's a growth— OS has a lot of cool stuff in it.

So, you know, actually, could you just tell me like maybe in the words of like your customers, like your customers are enterprise B2B marketing teams, like what are they doing right now? What is working? What are they not doing? What is not working? What do they care about? I'd love to crack into some of that.

Yeah, so lot of them, for them is, I wanna be transforming my work using these AI-enabled tools. How do I do this at scale? So really the first thing is functionally breaking down what they're trying to do and saying, where could this be impactful? Second is lots of experimentation, nothing sort of widely rolled out, but saying, how do I now enable it? And the place where they're running into some challenges is How do I govern this in an organization so that there is not a bunch of random stuff that can come to bite me? So starting to talk to IT.

Sorry. So, so the biggest bottleneck is not, can we create cool stuff with AI? It's actually the governance piece of this.

100%.

Yeah. So we have this community of CMOs, like most of them 100 million plus in revenue. We did an event for them in Arizona.

Yeah.

And the number one thing that resonated was like the disconnect between you open up LinkedIn and someone— or open Claude— oh, look at that, look at all this crazy stuff. This is the future. I get a message from CMO of a $700 million cybersecurity company. She's like, dude, we absolutely could never, ever, ever in a million years implement this. So it might not as well even exist. And I'm like, this is the reality that I think is so important to talk about.

100%. I sit on a board of a public bank and, you know, talking to our CISO, they're looking at everything. The cool part is they've got a really good, I'll call it a harness that they've built for ensuring everything kind of goes through this continuous integration, continuous deployment process that is sanctioned. So yeah, you can go build, but then it has to go through sort of a deployment cycle like we do today. And I think that ensures that I'm not going rogue. I'm governed from a regulatory perspective. If those are things that I have to be thinking about, that I've got my data being used the right way and it's not going out. God forbid someone used a Claude API key in some random app and now someone else has access to key and you have a half million dollar bill sitting on your laptop, right? So that whole structure, and it's real, that can happen so easily.

I know I accidentally like on my first that early at Drift, I've accidentally spent like $17 grand on AdWords over the weekend and it was like the worst day of my life.

Great example. Great example. You can do that in an hour. So yeah, those are real things that are happening. So I think the governance part and the cool part is a lot of the LLM companies are building those governance architectures. Claude and the Anthropic guys announced that I'd say Google and Gemini are probably the furthest ahead in terms of—

So the goal eventually would be like the same reason we would use— I'm not a tech guy, but so like the same reason we would use like Okta to like, you know, sign into all this stuff. If they build these features and then the vendor has all the security stuff, then okay, cool, marketing is using this cloud instance implemented by our IT team, like we're good. Is that what it would be?

Yeah, and the challenge becomes I build something, I showed you my dossier, I shared it with someone, someone says, okay, let me riff on that. If there's 500 people in an org and 50 of them want to create a different version, how do I manage that? Right. And then you made some cool changes to it and I want to make sure that it gets updated in a global instance. How does that happen?

Right. And then like what level of customer information is in there for different levels of seniority and people on the team? And yeah, man, that's crazy.

But how do I even roll up? So if you did something cool, who manages it? Who updates it? Like, who is the owner? Like, none of that is sorted out today. It's really much more of a, I'd say, individual sport than team sport today. I think that's the stuff that needs to get sorted out. So yeah, I think the individual innovation is happening at crazy scale.

Yeah.

The organizational side is still much more of a labs experiment, at least externally.

Did you see, um, what Ramp published this thing like earlier this week where they created their own internal AI system where they basically digest— it's kind of like your dossier thing, but for the whole company, where they're like, we built this ourselves, we have all of our company and customer information there, any employee can like submit a skill. I wonder if more companies will start to have something like that.

100%. So I think the whole notion around sort of building a company brain, which has got all the different skills in a managed manner available. So whether everything as simple as the brand team giving brand guidelines and how to use them, in a consumable manner, to the design team publishing standards, to the performance media team saying, this is how we do certain things so that anyone who's thinking and iterating is sort of using a common knowledge base. And now you have those owners that when you were saying, you know, hey, what does the marketing team need to do? Some of those things are, they're still the product builders and owners of those things. And then you have an agent or architecture that helps you refresh and update that on an ongoing basis. But yeah, I think that's going to become more and more common. And the other thing that we're doing is, and I alluded to it earlier, we're building, but we're not spending as much time refining. And as a result, a bunch of slop is coming out. So I think 90% of what we build right now is going to get thrown out. So how do you build a structure for building? And that's kind of where my— at first I built this cool dossier and I was like, okay, a bunch of shit in here is crap. Yeah. Makes no sense. So then I went into the classic product mode. I said, I need to define the requirements for what I wanted to do. And that's how you have to go back to your basics of systems thinking again.

Yeah, I've enjoyed using— I think maybe this goes back to what we talked about at the beginning of the podcast about being like a creator, or, you know, how you can do a bunch— however you set out, you can do a bunch of stuff on your own. Yeah, I really like the act of like writing the brief. Like, it's really fun because I love marketing. And so it's like I'm gonna spend an hour or 2 hours gathering all the information, like getting the examples, getting the swipe files, getting the copy, and then I get to like press go and like my little supercomputer is going to build something for me. Like, but I think to your point, without the taste, and like Dan, who's my CEO, like always gives me a hard time because I'm critical of like how fast we're doing stuff. He's like, well dude, you have 15 years of specific experience in B2B marketing. Not everyone is going to be able to do it as quick as you. But I'm like, this is the whole point about taste. It's like, I can do it now. I'm creating, uh, you know, copy or my newsletter and I'm doing with Claude. I'm not just going to Claude, Claude, write me a newsletter and I'm sending that out to 50,000 people. I'm actually spending a couple hours on each one, each newsletter, but I'm just doing it differently now. I'm like, so I talked to this guy Uzair, right? And he's really interesting and he's done this and that. And on my podcast he mentioned this and I wanna write about this in this week's newsletter. Can you do some research on him? Grab this from here, grab the transcript here, and then come back to me with questions. And like, that's how I write now. And it's like, totally. But that can't be— that's what's cool. Like, typically it'd be like, oh, you outsource something like that to the intern. Now the taste can be at such a high level because of the person who's doing it with the experience plus AI. That's the thing I think is really cool.

I think really cool. The scary part of that is because the output coming out, even the first time, looks cool. We don't do the diligence that we just talked about. So we just say, oh, it's good enough and forward.

Yeah, dude, listen, this is the first batch of hiring that we've done since Claude has been mainstream. And I literally was going through applications the other night and someone submitted like the actual Claude thing. It says here, you should tell the Exit5 team about blank. And so someone just got caught copying and pasting like the wrong thing.

Yeah, that's awesome.

What about AEO? Everyone is asking about AEO, SEO, what have you learned? Everyone wants the magic pill to like move the needle. Obviously it fits in your get discovered, get chosen, you know, world, but any learnings from your clients there?

Yeah, I mean, the first one is it's not an or, it's an and. That's the very simple one, right? Like sort of like everyone's like, I don't need to worry about SEO, I just need to do AEO. And everyone's like, no, if you don't have good SEO, your AEO will suck. So that's the first one. Second is back to your best insight comes from understanding of your customers. So if there's one thing I would ask people to do to start, I'd say take your transcripts from your Gong calls, your Zoom calls, from whatever else you have, both from your sales and customer success, and understand what the questions are that are being asked. Once you understand those questions, then actually go through, and this will be shocking to people, a manual exercise of taking those questions and going across different LLMs and seeing what the answers are and how you're getting cited.

We're back to basics. We're back to Google search, like what, you know, Google search in incognito mode.

Across the board and have different people do it inside and outside the organization to see how you get cited. And the cool part is you can see Are you showing up? Who else is showing up? What other third-party sources are getting cited? Because 85% or 80% of the credit is coming from third-party sources. So your content is important, but how it's cited and sourced outside is even more important. So, but we all jump into buying tools. We want to go get Ahrefs, we want to get Profound, we get the new Semrush, whatever you want to get, but we don't understand what the hell we're doing. And so I'd say that's the number one foundation if you want to sort of—

so, you know, is understand the question, have an opinion on strategy and what you need to solve before you go pick the tool.

What a concept, right? Right.

I mean, yeah, but you have to have confidence. You have to have confidence to do that because it's much easier just like pick a vendor and the vendor solves all the problems. It's harder to be like, Block everything out, maybe read a couple articles, right? Go do the searches, see how you show up, and then try to move the needle. I, man, I, I knew I'd like the vibes from you, but I love simple answers like that. It's like, look at it with your own eyes.

You have to, you have to, until you see it and understand it and look at side by side, it doesn't sort of happen. And then I think that again, the back to basic principles, focus, what are those 4, 5, 6 things that you could influence. Another really cool, interesting thing that I sort of, I was talking to one of our chief AI officer and he was saying most people are focused on showing up in the first query in the LLM, which is, hey, what are the best CRM platforms? That's great. That lists the top 10, but our attention span goes from that. We don't even look at the top 10. We then go, I am a manufacturing company that does roofing that is $500 million based out of this. Here's my tech stack. I have a team of 5 people. What should I be using? Right? So the notion of the ability to iterate and get to a decision based on what's relevant to me, even though you're showing up in the top query, you're probably not showing up in the answer.

Well, it's interesting. It's like the first one is like the listicle, right? And it's like, are you going to spend $100 grand and implement some tool because some shitty listicle cited Marketo as number 3 on the list? Yes. Or are you going to ask the more specific query, which you just did, and then believe that answer? You know, it's like, it's different. It's, uh, I need to know the time and temperature for cooking salmon in the oven. Okay, like, I'll take that one, right? But it's different if I'm gonna actually buy something here. Yeah, yeah, interesting. So that's like the long tail stuff almost matters more now.

Matters way more. And I think the question is, and the other part that's kind of happening and the rules are changing is there's a recency bias, which means that if your data is not 90 days or older, it sort of deprecates more. So it's no longer just keeping it current for that time, but you have to continuously get better and better and better.

So, sorry, when you say your data, you mean like how the LLMs are finding information about you?

Yes.

What's the best way to keep that up to date?

So there's a few things. There's the technical stuff that the CMSs are doing. So your schema markup needs to be updated and all those things. So your new platforms like Webflow and stuff make that easier. And then there is the content part. And then you have companies like AirOps and Profound coming in that are building essentially agentic workflows that'll go and say, okay, my organic search based on this content has dropped by 15% in the last month. Go research this topic, see what else is being talked about, look at what my competitors are doing, refresh a blog on that, update it on my— I'm gonna talk an example of on my Contentful page or Webflow page and set it up for review so I can have that orchestration layer now fairly automated, but it's still human in the loop. Now you can decide not to put a human loop, which I don't think you should today. But the notion of being able to research that's data-driven, be able to go create a revised content, and revised content is no longer what we all used to do is change the date and change the title and the content was refreshed. It's truly rethinking, reshaping the topic. But there is a lot of workflow automation happening with tooling right now. That's kind of interesting. Profound just raised like a huge Series C round. Their agentic workflows are kind of interesting. So there's a bunch of cool stuff that's coming out that I think will help with that today.

Yeah, it's interesting. Like, I feel this pressure to learn everything, and even you saying like, oh, well, if your CMS is, you know, you run your website in Webflow, Webflow's gonna have tools to make sure you— and then I see someone on Twitter, it's like, make sure you have the LLMs, you know, text file or whatever. And I'm not a web guy, so it's probably very simple to do that, but yeah. I wonder if all these, if everything's just gonna converge. And I'm even like, man, how deep do I even need to go with Claude Code or Codex or whatever as an example? Because the way these models, everything's moving so fast. You don't think in 6 months from now I'm gonna be able to just like write a prompt and have a website done? I think so.

I think so. I think you'll be able to, but I think that ability for, it's sort of the mechanics, right? If you understand the engine, then you can fine-tune it.

Yeah, fair. I did. I shared, I kind of threw some shade at like Vibe Coding a couple weeks ago and I got crushed for that. And that was kind of the feedback, which, and my thing was like, why spend my whole weekend Vibe Coding like a scheduling app when I could just pay $12 a month and use Calendly? And fair, I just was being a troll obviously. Yeah. But someone else was like, you know, but the value in that is like learning how these things work and learning like what goes on under the hood so you can do better with the AI tools. And that is totally fair.

Yeah, I think back to everything coming to taste and judgment, the more people use it, the more people use it better, right? And it's sort of that iteration is amazing and you get more and more excited. It was like analytics. When someone showed you a report, you saw that and you had 10 more questions. Why don't you have those 10 questions before you asked me that first dashboard that you wanted to build, Dave? Like, so it's the same scenario. The more we see, the more you see the art of the possible, the possible becomes more.

Sure. Okay, I'mma— let's leave with some parting wisdom. You could give a piece of advice to B2B marketing leaders that listen to this. In a time where there is a lot of uncertainty, there is a lot of disruption, there is a lot of hype, every day I open my apps and I feel like I can't do enough or can't keep up enough. Who's there? What would you say to someone?

Yeah, I'll say two things. One is find a problem you want to go solve and fix that problem or try to fix that problem, because I think you can do it. My example was a good simple example that we did. So be patient zero, be user zero of solving a problem. And the second one that I just think that is so paramount and waterfalls into everything is use the data you have to understand your customers. Because if you can understand what they're asking of you, everything else becomes connected. And we I feel in an organization today have become big and bloated and disjointed from really understanding what they're asking for and answering those questions. And I think we have the technology, we have the resources, and we have soon becoming more and more so the connectivity to enable that change.

I love it. Stay curious, solve the anxiety by doing something, right? Go take the next step forward, go build something. And I know a lot of people listening have not even taken the first step. I was on a session the other day and there was a demo of like how to set up Claude Code and I could not, there was hundreds of questions and everyone just completely overwhelmed by it, trying to figure it out in the chat. And I think the better opportunity is to like pick one thing, like you just said, I can think of what is something I do all the time and can I use AI to like make me smarter in that area? And for me, it's always been learn by doing. I, Don't learn because you told me. Look at your exa— your example's perfect. You don't need to be doing this. You got people that run shit for you. You don't need to be building this, you know, dossier, but it was an amazing way for you to get in the tools and learn, right?

Yeah. And it was the frustration of I was doing it, I was getting it, but it was bit. And I was like, how far can I go? And my new North Star and my thing is like, can you become Superman? Like, you know, is this a, is this a possible, how far can you go leveraging the awesomeness of the team that's there? But really not be physically dependent on them all the time. And I think that is possible.

All right. Thanks for coming on. We're going to wrap. Good job. I'll see you out there. Goodbye.

Sounds good.

Hey, thanks for listening to this podcast. If you like this episode, you know what? I'm not even going to ask you to subscribe and leave a review because I don't really care about that. I have something better for you. So we've built the number one private community for B2B marketers at Exit5, and you can go and check that out instead of leaving a rating or review. Go check it out right now on our website, exit5.com. Our mission at Exit5 is to help you grow your career in B2B marketing, and there's no better place to do that than with us at Exit5. There's nearly 5,000 members now in our community. People are in there posting every day, asking questions about things like marketing planning, ideas, inspiration, asking questions and getting feedback from your peers, building your own network of marketers who are doing the same thing you are so you can have a peer group, or maybe just venting about your boss when you need to get in there and get something off your chest. It's 100% free to join for 7 days, so you can go and check it out risk-free. And then there's a small annual fee to pay if you want to become a member for the year. Go check it out, learn more, and exit5.com, and I will see you over there in the community.