This website uses cookies

Read our Privacy policy and Terms of use for more information.

The martech industry has spent the better part of a decade building increasingly sophisticated tools for people it barely knows. The vendors write the press releases, the analysts write the quadrants, and somewhere in the middle a solo marketing ops person is sitting on their martech island at 11pm wondering if anyone else is dealing with the exact same mess. Phil Gamache decided to find those people and give them a microphone instead of a framework.

Gamache is the host of Humans of Martech and brings 13 years of in-house experience across companies including WordPress.com and Close.com, spanning B2B, B2C, and B2B2C. He has been building the show for six years, not because he had all the answers, but because he was honest enough to admit he didn't, and smart enough to realize that honesty was the product.

Why "AI-powered" stopped meaning anything

There is a tell when a martech vendor has run out of ideas. The homepage starts by talking about supercharging revenue with agent workflows without ever showing the product in action. The G2 review snapshot from someone at an unspecified company remains. The logo wall features a company where one person in one department used a free trial. Gamache names this pattern without ceremony.

The differentiating factor now is not artificial intelligence as a feature claim. It is whether you can show what the product actually does and back it up with the humans who will help you implement it. The sales rep matters. The CSM handoff matters. The dedicated Slack channel versus the Adobe-style support queue matters enormously to a practitioner who has been burned before.

Buying the tool is just a tiny part of that journey, making the most of it and getting it implemented. There's a human story there.

The vendors winning that argument are not winning on feature parity. They are winning because the humans on their team made a buyer feel like someone would actually show up after the contract was signed.

Why the best podcast guests have stories, not answers

Gamache screens guests by asking what irritates them. It is not a trick. It is a filter. The guests who come back with a specific situation, something that broke or was reassembled against the odds—are the ones worth an hour of anyone's time. The guests who come back with a polished opinion are not.

This distinction matters beyond podcast production. The entire knowledge-sharing problem in martech runs on the same fault line. Someone posts on LinkedIn about building something with a single prompt, and the industry experiences collective FOMO. Someone else goes on a podcast and admits they are still debating packaged versus composable CDP at a company people have heard of, and suddenly a practitioner somewhere feels like their own uncertainty is not a personal failure.

The person at Canva is still figuring out the same thing I'm figuring out. I'm at the same level — like I'm on the same wavelength as that person from this really cool company.

That feeling — being seen by someone whose brand carries more status than yours — is not soft. It is the actual mechanism by which a field improves. The practitioner who hears that goes back to their company with more authority to say the quiet part out loud.

The production stack no marketing team has built yet

Gamache runs a one-person content operation that outpaces most in-house teams in both volume and editorial quality. The workflow is worth understanding in specifics.

He feeds Claude a guest's LinkedIn posts, their pre-interview answers, and any published writing, and gets back a dynamic persona. Claude runs a mock interview with that persona, improves its own questions across three rounds, and surfaces the angles most likely to generate a real conversation. Gamache reviews those, sometimes sends them to the guest to pick favorites, and builds the episode thesis from there.

Post-production runs through a similar pipeline. Transcripts are chunked, cleaned, and assembled into a long-form blog post that becomes the engine for every downstream asset: clips, shorts, newsletter drafts, and Midjourney image prompts. The newsletter still gets heavy human intervention to stay in his voice, which is the right call. The Midjourney work he does does not automate at all, by deliberate choice.

You should never automate things that bring you joy. You love doing a part of it, even though it takes a bit of time. Why would you automate something that you love doing?

That is a better heuristic for AI implementation than anything currently appearing in vendor white papers. Automate the joyless parts. Keep the craft.

The job market is hard, and the industry keeps lying about it

A few years ago, the advice was simple. If you are a talented marketing ops or rev ops person sitting in a bad situation with no process, no authority, and tickets piling up — leave. The market will catch you. Gamache was one of the people who gave that advice, and it was accurate at the time. It is less accurate now.

The honest version of the current situation is that genuinely good people are sitting three or four months into a layoff with nothing to show for it. The AI narrative has made this worse in two ways at once: either it is used to justify headcount reduction or to justify demanding five times the output from whoever is left.

If you're not getting laid off, you're like five times more stressed and expected to have five times the bandwidth.

The industry is not going to solve the structural problem by talking about it more loudly on LinkedIn. But naming it honestly at least stops practitioners from internalizing a broken market as a personal failing.

Humans of Martech has quietly shifted toward helping people navigate this, with episode topics moving from “which CDP to buy” to “how to survive an interview question about what AI tools you have been using.” That shift is worth noticing.

What revenue attribution actually requires

When a CEO asks where to invest the marketing budget next, the honest answer is not a tool recommendation. It is a question back: what is the philosophical consensus at your company about how marketing gets measured in the first place?

That conversation is harder than any technology implementation, and it has to happen before any technology decision is made. Gamache is currently advising a company on this. The framework he describes is not complicated. It starts with what measurement method is currently in use and whether there is a culture of experimentation to support more sophisticated approaches.

A 15-person B2B startup has a completely different conversation than a 100,000-person B2C operation. Last-touch attribution at a small team with limited resources is not a failure. It is proportionate to what the data can actually support.

You have to decide on the right mix of triangulation signals and methods. If you start with just the last touch, no one's going to fault you.

The problem is not that teams use imperfect models. The problem is that teams skip the philosophical argument entirely and go straight to buying a measurement platform to avoid having it. The platform does not resolve the disagreement. It just makes it more expensive.

Three Takeaways

Six years of Humans of Martech, distilled into three things worth taking back to work.

  1. Show the product working with a human on screen. Not case study snapshots, not logo walls. If a vendor cannot show a real person using the tool and getting a real result, the AI-powered positioning is carrying weight it was never meant to hold.

  2. Automate ruthlessly except where the work actually matters to you. Gamache has automated transcript cleanup, blog post generation, clip identification, and newsletter drafts. He still does the Midjourney prompting by hand because it brings him joy and produces better output. That is a workflow principle worth stealing directly.

  3. Do the philosophical work before buying the platform. No measurement tool resolves a team that has not agreed on what marketing is supposed to prove. The internal conversation is the hard part. The technology is just what you buy after you've had it.

Phil Gamache is the host of Humans of Martech and a martech and growth advisor.


00:00:01 — 00:01:52

One year of making sense of martech. And for our anniversary episode, I booked my main competitor. Skokomish has been doing this longer. Episodes of Humans of MarTech are built on the same belief that drives the show. The most interesting thing about our industry is not the tech, it's the people trying to make it work.

The ones who are on their martech island at 11 p.m., wondering if anyone else is dealing with the exact same mess. We're peers. We're jokingly frenemies and genuinely good friends. We're not just trying to shape what the industry looks like based off of the tools. We're trying to change how it treats the humans inside of it.

And this felt like the right conversation to mark year one. Phil has quietly become one of the most trusted voices in this industry, not because he has all the answers, but because he's been honest about not having them. And he's created a safe space where other smart people can be honest, too. And that is far rarer than it sounds.

But a little bit about him first. So Phil, not only, of course, is the host of Humans of MarTech, where conversations are happening with the sharpest minds in the industry. He's also really earned his stripes across 13 years of in-house experience for marketing and growth at companies like WordPress and close casual and spanning B2B, B2C and B2B B2C.

He's also confessed that if money were no object, he'd probably be running a fantasy hockey newsletter. I've got some questions. He did rivalry ones up and still is coming to terms with how niche that actually is. Welcome, Phil. Really glad you're here. Yeah, thanks for having me. Really excited, big fan.

Obviously, in the long time coming, I feel like I'm excited to to chat with you. Finally, on the record, we chat all of them off the record, but aren't all cleaned up in Polish. This will be in terms of like, uh, censoring ourselves a little bit. No such thing.

00:01:53 — 00:02:18

Yeah. Um, for folks who don't know, we first met formally when I was getting interviewed when I had just left Grammarly. And that was also the same week you were going full independent and quit your full time job. So we've kind of been parallel on similar journeys this whole time without realizing it for a while.

Yeah, we're both flying through the sky on our on our jetpacks. Yeah, it's still true.

00:02:20 — 00:03:05

Well, let's dive in and start with some rapid fire. What was your first martech tool? Uh, you'll love this answer. I know you're a big fan of the tool. Part of it was my first marketing big marketing tool. Like, I guess, like, Photoshop is really the answer if you can't Photoshop, but it doesn't really count.

I thought about that. Yeah, I've been on Photoshop nearly 20 years. Yeah, but yeah, Pardot, uh, worked at a tiny little startup, and I'm in Ottawa in Canada. And like, in the the depths of a suburb, we had this, like, little engineering firm and they actually had a ton of like top of funnel. And we had a big newsletter and they gave me the keys to Pardot right out of university and as my startup marketing automation never looked back.

Love it! I'm glad you started with the hardest and the best.

00:03:06 — 00:03:36

And at the time, I don't know if that's the case anymore, but I feel like the best looking UI, the most enjoyable one. It's still more enjoyable than Marketo. Oh yeah, let's be honest, no debate. Okay, so your podcast, you don't do it for consulting leads or sponsors or really anything like that. What is the real reason why humans of MarTech exists?

Well, the humans of MarTech exists for the humans. The the lame answer there, but bada boom boom, it

00:03:37 — 00:05:37

gotta be the humans all like. Obviously we do it now for for sponsors like we do it through sponsors. Before I thought we were going to do this because it was going to lead to credibility and consulting gigs and like I do it for the people that are lost in this crazy galaxy of marketing tools and AI tools, or the people that, like, just lost a job and are trying to figure out what is next in my career.

I was in a dead end job. I stumbled upon your life cycle martech saga and dived into it. Gave me the courage to start consulting, and now I consult, live on my own account and make my own decisions. Left that like dead end job and I'm so much happier now. Thank you for like putting out this content. It's actually people like seeing value out of this saying that's why we do it for the humans.

I love that you called staying too long at your first job the biggest mistake of your career. Since you do talk to people a lot about their careers, what would you tell that version of yourself today? Yeah, I'd go back and I'd tell Junior Phil to, like, not spend as much time at that one company. Like, maybe I was a bit too comfortable and thought that I could, like, make it big and Ottawa tech scene.

And I would tell Phil that, like, there's there's such a bigger world out there, there's a bigger martech ocean. And, um, I don't know if it would lead to a different outcome, really. But that's the advice I give to a lot of folks that I mentors like. Don't get too comfortable in the gig that you are right now.

Like you always have to be hunting and thinking about what's next, whether it's a side project or like this, like base of advisors you have like what is next? If you get laid off, no one's safe these days in this space. So, um, yeah, that's the advice I give myself, like four years at a job, first job at a university.

It's kind of okay, but I don't know if it's the advice I'd give folks who still do today that makes sense. And it feels so antithetical to what we grew up on. Like, no, you stay at a company forever, and at least for years. I mean, I was told by my grandparents at four years, you know, what you're doing. I'm like, I disagree.

00:05:38 — 00:08:00

I knew it a long time before that. So one of your unique things, when you're screening guests and people to come onto the podcast, you ask what irritates them. I don't remember being asked that, but maybe you did. What does this answer tell you about whether they're worthy of a conversation or they'll bring insights to your audience.

Yeah, I can't remember if, like when we started rolling out these, like, pre-interview questions. I stole this from Lenny. Like, Lenny does the same thing. Um, like, regardless of who the person is, he has, like, these list of questions that he sends over. I don't do this for, like, everyone recently, like, I've gone down this train of just, like, inviting, reaching out to folks that have written something really cool that, like, piqued my interest, challenged what someone else said on the podcast, and they come in and they kind of debate it.

Or I've had a couple authors that released a book recently, and like I read the book ahead of the conversation. We dive into like one chapter really deep. But for some folks that, like, haven't written a lot of stuff or, you know, we're just looking to come on the podcast and we're not really sure what the topic is going to be.

I send over that list of questions, and immediately the thing that jumps out with like conversations that I think are going to be interesting is when they have stories in the answers, like a lot of people answer, but the best interviews are the ones where people have a story and they're just like this one thing happened that my company, we had GTM ops under Rev Ops, but then we moved over under finance and like everything was in shambles.

And so they want to like tell that story about how they brought the band back together or how they ditched MTA. But then they had to crawl back and reintroduce it because they had nothing to give them real time signals on all the performance stuff they were doing. Like, there's so many of those people that come on that are just, I want to talk about this one thing, but they don't attach it with any like stories of on the ground, in the trenches.

So that's the folks that have like the, the coolest answers are the one that have the real stories. The human stories. Yeah. I mean, it sounds like those who are willing to be relatable by telling the truth. Yeah. Anyone else make sense? It also makes for better content and better interview, for sure. Okay, you've said that AI powered is no longer differentiating factor.

It's the default. So what is the differentiating factor? Now that we're in the year of 2026. Yeah. The differentiating factor is that you have to be AI pilled.

00:08:01 — 00:22:37

I know it's your favorite word right now. I have to like slide it in. Yeah I'm saying like token Max okay. Those are real triggered words at this point. Let's be frank. The privilege of like, being really selective with the consulting that I do right now, and I do a little bit of that. I have this, like, bird's eye view of martech.

And I love giving people some of this advice. Like, your site is really cool. I think you're like features that the menu is on point, but there's nowhere on this site where I actually see the product in action tells me. And like, I don't care about those G2 reviews, snapshots of like so-and-so at unknown company like that doesn't tell me anything, even the logos you have on the site.

Like, maybe it's someone in this one tiny little department at notion that's using the tool, and it's not really like anyone meaningful or a meaningful department at the company or whatever. So like, show me how. What does that mean to like, scale or supercharge revenue with your agent workflows? Like, what does that actually mean?

Show me the product. Like, show me demos of people using it. Hide the sensitive data, whatever. But like, tell that story, the product story. So that's that's part of it. And the other part is AI powered is one thing, but I feel like we're seeing this renewed focus of human powered. I don't know about you, but like when I buy martech part of it as a product decision for sure.

But there's such a human factor in which RFP I end up going with based on who the sales rep was, how were my conversations, the vibes with that salesperson, the conversations with the potential CSM that's going to take over that account on your website, showcase the human part of it. Who are the humans that built the product?

What was their story? Why did they build it? Who's the account team that's going to get you? Set up the success team. Are you going to have Marketo Adobe style support or do you have like dedicated slack channels and people that are going to help you out? So I feel like those two things show in the product. What does AI actually mean visually?

And then like talk about the people that are going to help you implement this thing, because buying the tool is just like a tiny little part of that journey, making the most of it, getting it implemented. There's a human story there. Yes, I cosign everything and I am well known for being a sales reps. Worst nightmare or absolute dream customer.

And I actively will say I need a new salesperson if they are not up to snuff, let's just say it doesn't go over very well, but it is required because we want to put the best foot forward for both sides. Okay, I want to dive in with you because something that sets you apart isn't who you have on your podcast, isn't how many episodes, but it's the format.

I love that you took Humans of New York that approach to how our industry is talking in frameworks and vendor decks, and instead you're pointing the mic at the messy human reality beneath the surface. I know I've been a long time follower of humans of New York, and so I just I love how that has become part of what you've built.

So where did that instinct come from to treat martech practitioners as people worth actually knowing as humans, not just experts, to extract insights and frameworks and advice from? Yeah, I think those things come better as stories when people share the learnings along the way. Like, how did you get to that tactic?

Because that tactic can't be applied blanket at like every company that you work at. So knowing the story, the business model, the failures along the way, where it broke, where it didn't broke, where it didn't break. And so like, I feel like that's part of it. Everyone sees this on LinkedIn. Like people sharing this one thing, I built this thing with one prompt and like everyone feels a sense of FOMO and imposter syndrome.

Like I'm not good enough in my role not doing this, and I'm not doing that. And part of like the synthesis of Sharon Human Stories is sharing the feeling that, you know, you're not doing that bad, like you're doing okay. You look at these people, at these amazing brands, they're in a very similar spot as you like.

They don't know what they're doing either. They're still figuring out as they're going. They've added this. They've added that they're still debating packaged or platform CDPs. They're still talking about MTA or not. They don't know how to report marketing, like you're not doing that bad, you're doing okay.

You're figuring stuff out. You're asking the questions, you're curious. You're trying to learn like you're already better than most people. So I feel like that's the humanizing part of it is like, no one's perfect. People on LinkedIn like to make you think that, you know, they have it all figured out.

But reality is like most people don't. So we talk about the tactics and the strategies that come from the learnings and the journey along the way. So it's like our Humans of New York version of it, if you will. But yeah, that's my take. I wholeheartedly agree, and that's why I was ever drawn to when we met. I was like, you get it?

Because does everyone wants to be seen. We want to be heard. We want to be validated because we're often on our own little isolated martech island internally or Mars Island, and only your few stakeholders understand. Or that's if you're lucky, frankly. And so I think the most beautiful part of our industry is the community.

And I'm grateful you're a big part in it. Yeah, and I'm only a part of it because of the people willing to share those stories. And so I feel like there's there's a shout out here to everyone that is willing to not just be heads down, lost in the sauce doing the work, but also willing to spend time sharing with other people and contributing to the community and helping younger folks coming up learn from where they've been and take that advice.

Like, there's there's so many people still today, six years into this, thousands of people listening. I still have a hard time sometimes, like getting a hold of someone to come on the podcast because they don't invest in personal branding. They work at this big consumer tech brand and like, they're not worried about finding another job if they lose it.

They've never done any public speaking. They're just like heads down doing the work. And then they clock out at the end of the day. But along the way, they've learned from people, people that have shared freely, that have given back to the community and they decide to not give back. And maybe it's like an introverted thing or not, but tons of respect for the folks that have gotten somewhere in their career and are at a stage where they have learnings that they're willing to like, share with other people, and they have the courage to come on behind the mic and share it with other folks.

Not everyone is willing to do that, and it's totally fine. But yeah, kudos to the folks that are willing to to share that for sure. I mean, I relate to that deeply, and that is why I want to create my own platform, too, because I was that person that was not responding to anything, not because I didn't want to, but because I literally never logged on to LinkedIn, because I was working.

I was doing all of the work. And it's, it's and also I'm introvert, but really it's when you're so heads down because you're doing good work, there's so many people that are forgotten or not known. And I think we all deserve to be elevated to where we're supposed to be, whether we realize it or not. Okay. You do a lot of prep, and I know this very well.

You send the guests a loose kind of template ahead of time. I do something similar, but you're clearly after something more than just the polished answers. So how do you get someone to drop the professional armor and actually talk. We end every episode on like, this happiness note. Like, uh, um, what's your personal system for staying aligned with?

What? What makes you happy? And I tease out that question by naming, like, the professional roles and the home life roles of everyone on the show, and sometimes they don't speak to publicly. So I don't really know what their home life roles are. And so before we press record, I tell them that, hey, for this last question about happiness, this is the professional roles that I have for you.

I know nothing about your personal life. What can I write down like, are you an avid gardener or are you a parent or grandparent? And and so we kind of like get them loosened up a little bit by sharing about their home life. What what else like, gives them energy, what makes them happy. And so we're writing like hobbies and their other like personal homelife roles.

And then we get into actual recording. So they feel like that maybe helps with like losing them up a little bit, getting into the mind frame of like, hey, you're a human. We're not just talking about like how to buy martech or how to lead marketing like we're interviewing you, the person as the operator. Not necessarily.

You know what you think other people should be. So I feel like that's like a good way to ground that conversation. Maybe it doesn't always like remove that professional armor because like you said, like we're we're both introverted. And I feel like some people that come on the show that don't do a lot of public speaking.

Like have those, like, prepped answers. And you can tell sometimes that, like, not that they're reading the script, but they're kind of like defaulting to what they wrote down and stuff like that. And so it doesn't always like come through, but I feel like that's, that's the least we can do a little bit.

And sometimes we like prep them for some of the topics, but we don't give them the exact questions so that sometimes it is like a bit of a surprise. But I love to do the prep thing and like the loose script so that they can actually think about like examples and stories you and I talked about, like some of the questions ahead of this, and I came up with like stories that I could bring for this question and that question on the spot in the moment, no way would have thought about any of those.

Maybe it makes me a bit more interesting and less like corporate armor, but I don't know. Just riff in there. After so many conversations, what is the most unexpectedly hopeful thing you've learned about our industry? I love chatting with people that are at unknown brands like startups, up and coming founders in martech, but obviously love getting an inside look at companies that are at like OpenAI or Uber or Canva, meta like anthropic, like some of the cool companies that I've had on the show.

But I think that the thing I love the most, and some of the feedback that I've gotten from folks, is that regardless of how cool these brands are, I feel better after listening because sometimes they talk about stuff that I've never heard of or I've never done. I'm like, oh crap, I need to go and learn that thing.

But sometimes they talk about something that I've actually figured out already, or they talk about something that I'm also figuring out. I'm like, whoa, the person at Canva is still figuring that thing out that I'm figuring out. I'm at the same level or like I'm on the same wavelength as that person from this really cool company.

So it's that feeling of, I don't know, being seen, being validated, being reassured that you're not the only person in this alone on this little island of Rev tech or marketing ops or revenue ops in your company, and maybe you feel like you don't have that authority to to like, speak about some of these things.

But then you hear from someone at this really cool company seeing this other thing that you've been saying, and now you have like a bit more ammo to bring back to your company. So I don't know, like no one has it all figured out regardless of like what company or brand they're at. We're all kind of like learning as as we go.

And you're not doing that bad. You're going to be okay. Speaking of not doing things so shabbily or badly, the singular Scott Brinker has told you that Humans of MarTech is one of his favorite out of hundreds of podcasts. What does it mean to you that the so-called Godfather of martech gravitates towards conversations that sit with the complexity rather than sell certainty.

And with you at the helm, I feel like Scott is the epitome of that. Like I grew up on reading chief MarTech com and it's like still my de facto newsletter. It means a lot. It means a ton. Throughout the years, like I've gotten to become friends with Scott. Like Scott's come on the podcast twice and we've collaborated on On There this stuff together.

He's such a voice of reason in the industry. Like, everyone loves God. Everyone thinks of Scott when they think of martech, so it means a ton. And I was just like pondering this. Like, imagine our industry if there were even just like three times more people today that were like Scott, that actually, like stepped out of their heads down work and shared what they were learning.

Like Scott does. Like, I can't imagine how amazing, more amazing the community would be, like how faster other people would be learning if there was like 7 or 10 other chief martech coms? People had other brands like Sharan stuff along the way. So grateful for for Scott, for everything he's done for me in my career and in the creator journey that I'm on for sure.

That's amazing. I think you're underselling yourself. I think you're one of those people as well. Okay, I want to dive into tech more specifically. And you have been using AI, leveraging it, especially because you've been on paternity leave a number of times. You've had all your babies on your chest.

ChatGPT was taking off, and that turned your curiosity into a whole new content layer for the show. So at a high level, we walk us through your actual production stack. Where is AI actually helping and assisting you and where is it continually letting you down? We started six years ago just recording on audio and not even having show notes.

And you know, we had like a little less than a hundred people listening in. It was fun. We were doing it just for fun. Right? But yeah, had my first kid and it was still working in house at the time. So had decent pat leave benefits and baby was like sleeping on me like a baby sleep, poop and cry. I was like researching a lot, like prepping episodes.

And that's around the time that the earlier versions of of GPT came out. And I went on this like deep exploration of like, what does the role of marketing look like with LMS? Like, how are LMS chasing everything? And everyone was like wondering those questions, asking them at the same time. So I happen to like, write and post episodes about this at this perfect inflection point right before everyone else did.

And it was like saturated the we we can see kind of in the download spikes, like right around the time where GPT was kind of coming around and a couple weeks after that, like when I went back to work, big project we were working on in-house was debating the package CDP versus the composable CDP.

00:22:38 — 00:28:37

Something that most people don't like to argue about nowadays. I feel like the battle has been settled. We can talk about that. You'd be surprised. Depends on the company structure 100%. But like I did a whole series around that as well. Um, packaged versus composable CDP and like had both sides of the coin and those episodes did super well.

Also, I don't know if like it was the AI stuff that did better first or if it was CDP was all like around the same time, but super randomly it started having like founders of CDP companies reach out to to come on the podcast. I was like, what is what is happening here? What's going on? I even had a PM shout out to Katie, who is at census, and she invited me to host a webinar with Darrell Afonso and Scott Brinker and the CEO and founder of Sensus.

And I was just like, what is happening here? Like, I just happened to stumble upon this really interesting debate that a lot of people are going through right now. And the AI thing really got me into like, what is the future of marketing and marketing operations and what does that mean for me, like on my own production stack today?

Like, I went down this like recent journey in 2026 of like fully embracing cloud code and cursor as like the main command center for the pre-production and the post-production stuff. I've got a couple of cool systems and processes where I'm feeding a couple of different documents and like outputs of scraping different like documents of stuff that people have read that are coming on the show.

I feed it like LinkedIn, experienced LinkedIn posts, the pre-interview answers that I shared, and then I have like a pre-production system that suggests a bunch of different content angles for the episode. And so I have this like human checkpoint where I'm actually like picking the angle for this. I actually, like, reach out to the guest sometimes, like, hey, there's like three ideas here.

Which one do you enjoy the most? And then based on that cloud, creates this like dynamic version of the guest's dynamic profile based on how much information they've shared online. Sometimes it's really loose. Sometimes someone like Scott or like an author has like a full book to go off of, and then it has, like a mock interview with that dynamic persona of the person, and it improves its own questions, like through the mock interview.

It does like three rounds of it, and it shows me like the output of the questions that were improved based on the mock interview outputs. So I take that, I send it to the guest, we go back and forth a little bit on on the questions. They highlight the ones that they like the most. That becomes like the core thesis for the the pre-production side of it.

And then post is like where I do the most stuff, like right around like the early days, I was converting everything into a blog post version of it. In the old days I had this like long ChatGPT prompt. There's like a three step prompt to remove all AI slop out of it, or as much as possible. And I would just take like, all right.

Help me create this blog post passage from this question. This is the question I asked. And this was the guest's answer. And I was doing this all manually, like copy paste, sending that to ChatGPT, pacing that output, pacing in the second version of the prompt like, all right, improve this again. Remove this, remove that.

So I was doing it all manually stitching it together in this like long thing and my WordPress blog. So that part of it has just been all automated in in cloud code. Now, like I feed it a transcript, it automatically finds all the different questions, the answers, and it chunks it into different spots. It cleans it up where it's repetitive, it kind of cuts it.

It fixes like the missed transcriptions, and then it outputs like a section for that. And then by the end of it, it has like a long blog post goes through it in one shot, like synthesizes, summarizes, combines similar ideas, outputs a blog post, and that blog post becomes like the the center of all the assets that I'm going to repurpose from it.

So it becomes like the engine for the clips that are going to come after it, like the shorts. So I have a bunch of cloud skills that look at like the best, like most potential viral moments from it. I don't know what works still, and I, I think it's an iterative process. I did spend a lot of time like looking at like how many people in shorts, like, end up actually becoming subscribers and look at the long form stuff.

And I looked at like other people's reports. They're also and it's like not a great signal. It's a different audience. It's like short form, like really passive listeners versus the people that actually stick around and listen to full episode. But anyways, long story short, like blog post gets repurposed into a bunch of different stuff.

Also goes into like this newsletter engine that's still pretty heavy human intervention to kind of like clean that up and put it into my own voice. I'm always thinking about not token maxing, but like, what? What else could I be doing? Because it's like, it's so fun. I have a ton of fun with it. Like oftentimes our little one is like in bed sleeping and my wife's about crash and she's like looking over at me, and I've got this terminal window and cursor and she's just like, what are you working on?

That looks like so gross and depressing. And I'm like, no way, man. I'm building a search app right now on my site, and I've finally connected to a railway database and I can see what people are searching for. And it's giving me like, content ideas. And now I can like surface episodes that chunked and talked about that based on the rag pipeline that I build.

So anyways, I could go on and on here, but it's so fun. Brought to you by our sponsors. If there's one theme that's followed me at every stop in my martech career, it's trying to get good data into the hands of marketers. That's why I'm so excited to tell you about our sponsor, High Touch, the leading composable CDP and AI decisioning platform companies like Domino's, Chime and Ritzy and PetSmart trust High touch to power their data.

And here's the kicker 90% of customers have a real use case live in production within their first week. That means you can implement a world class CDP in months rather than the usual years long headache. That's why top brands choose high touch to personalize every customer interaction at scale. See what high touch can do for you at high touch.

00:28:39 — 00:29:44

And now back to the hot seat. Based off of what you're saying. I feel like you are leaps and bounds ahead of every company's content approval generation creation process. So I'm just going to take the moment to say, guys, you don't have to be in-house to know what you're doing. I'm just gonna put it out there because while I know a number of companies doing advanced work, nothing is that advanced.

Yeah, I've had the pleasure of chatting with some cool folks that are, like, deep into not to throw like, rule buzzwords out here, but like, content engineering, like my old co-founder, old co-host John Taylor. He like, left the podcast to build out his SEO agency, and now he's pretty much just become like a content engineer.

He's created his own, like, miles ahead of anyone. And like I learn a lot from those, like marketing engineers that have been marketing engineers for a decade. The ones that just got created. So what are you talking about? In April, the startup, like one startup created, uh, marketing engineers. Jacqueline.

You didn't, you didn't. It's so profound.

00:29:47 — 00:43:31

But, yeah, it's fun to think, like, sometimes you chat with folks and you're like, all right, I'm doing really good. I'm ahead of those folks. Like, I'm not losing too much behind. And then we chat with, like, Austin Hays of the world and like, the Keith Jones and you're just like, all right. Wait. Sorry.

What? Harness engineering. What? What are we talking about? What is that thing? One for me. The what? Like what is opening I doing right now? Symphony orchestration for LM like. And so there's always going to be those folks that are in like the 0.1%. And if you're still like learning from them and you're ahead of like most folks, I think as long as you're having fun, I think that's the most important part.

I'm having a ton of fun. Definitely. It brings you joy. You're going to keep on exploring and doing it. Okay. We've talked through how you set up all of your workflows. However, you neglected Midjourney. Everyone adores your process. For Midjourney, I want to know, like, how did you start that idea? There's a whole breakdown already on your podcast of how you do it, but I want to hear where that came from in terms of your inspiration for cover art.

Yeah. I actually didn't think of Midjourney when you said like the the AI stack is for me, it's it's AI for sure, obviously, but it's such a manual process. There is an art to prompting in Midjourney, even though it's not like human art, but I feel like it is what sets like some of the brand apart for for the show.

Like a lot of folks have said that, especially in the earlier days, like especially the cover art. Like, people are just like, well, it's so cool nowadays it's less cool. Anyone can do it. And like, it's great. No one's quite got the style and kind of consistency. You do? Yeah. It's it's fun. Like Midjourney, I feel like is one of my favorite creative outlet.

Part of this, once we've got it is like another output that I get Claude to do. So Claude takes like the header of each section of the blog post, which is essentially like all the questions and the topics that we cover, all the H2 and the blog posts that become their own clips. And so Claude outputs like five potential Midjourney prompts for that one idea.

I go through like some of my favorites. Sometimes Claude is really bad at doing that. I try to come up with like, cool visual, like metaphors for something like, we're talking about source of truth, and or we're talking about marketing planning and can only do so many lighthouses in the horizon. There are so many, like compasses out in the wild, you know?

So I'm like quickly running out of analogies, but it's fun. I feel like it's it's my favorite part. And before, I used to be just like like a blog, right? Like the header for the blog. But now it becomes like the header for the clip. Like it's something that shows up in the short. It's something that I use sometimes when I'm like repurposing an episode and just like sharing about it on LinkedIn.

But yeah, like I had Tata on former Nike, who does like a couple of AI courses on Maven. I had her on the podcast. She says one thing that really stuck with me because, like, I've always been thinking about, why am I not automating this part out of Midjourney? Like, I spend tons of money, tons of credits on Midjourney and like a lot of time and focus and energy.

But she said something like, you should never automate things that bring you joy. You love doing a part of it, even though it takes a bit of time. Why would you automate something that you love doing? And so it made me reflect on that and like, made me pause on like, maybe I don't need to automate everything on a midjourney, but Midjourney is by far and above the the best image generator.

Not for like concepts and for words. GPT today dall-e there engine is beyond Midjourney, but the features in Midjourney like being able to like change just like this one spot in the image, or have like a style reference or character references. These are all things that are like years ahead of the training models on any of these other images, and it's what's allowed me to have a consistent style, at least, like across episodes, like I come up with like the cover and then that becomes like the style reference prompt for all the other header images that that come out of it.

Well, I think as a guest, the experience is also exciting too. You feel special because you're like, oh, it's just a custom cover just for me based off of my interest experience. And I think it's a it's a nice experience for all across the board. Brings you joy brings customer and audience and everyone joy.

Win win win. Exactly. At what point did the show stop being a side project for you and start feeling like your main thing, and that you needed to commit it? Probably around the time where, like the AI exploration. The CDP battle episodes really picked off like we weren't putting out episodes for a couple hundred people.

All of a sudden, at that point, we were chatting with like a couple thousand people and we're like, whoa, what happened here? Like, we're actually generating downloads here. And at that point it morphed into like, all right. Our audience is like really, really niche. There's not a ton of other creators in this space, in this one niche thing, like how many people talk about CDPs and composability and AI or orchestration.

There's not a ton, very many. I went on this like journey. I was still working in-house at that time, but I was like, hey, maybe, maybe there's a world where like, there's a way to monetize this through through sponsorships because like, you know, there's brand awareness play here. Like, not everyone knows about all the CDP vendors, and we're talking about CDP, like maybe a CDP vendor would be interested in, you know, sponsoring a couple episodes, doing it on a test basis.

And so some of the first ones that we just like pitched this to wear Nak. Nak was like the the first sponsor that we had. There's still a sponsor today four years into it. But like their head of growth and their CMO at the time, where it's like, hey, we're doing this content around marketing, operations and martech.

And they were just like, yeah, never sponsored a podcast, but we'll give it a shot. We like the content, but we love you guys more than anything. We're happy to like, support you creating content for this community. We're building a product for this community. I think it's like a perfect match. And we had iterable CMO on the show.

I know you know very well Adri and she came on the show. We were just like three ad or we're like, hey, you think like iterable would ever consider sponsoring our tiny little podcast? Like, we've actually had some pretty cool growth. Like, we're not like, speaking into the void anymore. And she was like, yeah, chat with this one person.

Let's give it a shot for a quarter or see what happens. And then like iterable and NAC, like two really cool brands were some of the first ones to sponsor us. And like now these other vendors, it's like whoa. All right. Like, if iterable and NAC are willing to like throw money at this like customer IO came knocking like they sponsored for a full year after that and Mowing Gauge is sponsored after that.

We had a couple CDPs come on like Growth Hoop as a sponsor now. So yeah, it's been a journey and I feel like that was like that moment where like we had one really cool martech vendor, reputable brand, just be like, yeah, we're willing to pay you to put our logo on your content. That was like the moment where I was like, all right, maybe we have something here.

Like, we're never gonna be Lenny and have like 100,000 listeners, but we live in this niche of a niche, and it's a fun niche and it's a high quality audience. And I think there's really cool content to make here with the audience and, and partner with some vendors that we like and trust and vet their products.

And so at that point, I was just like, yes, my co-host and I like he ended up going his own route, like building his own thing. And I was just like, all right, I need to not focus on in-house anymore and like, just double down on the content and and see what this like solo creator life does. And now Daryl Afonso, who used to be like a long time listener, you reached out and he was like, hey, look, I noticed you have an opening for co-host.

Like, I'd love to. Come on, just for fun, see what happens. And I was like, all right. Like, everyone knows Daryl, like, for sure, but I'd be silly. He was, like, one of the most downloaded episodes in in the history of our show when he came on. So, yeah, he's been on for a couple of years now. He's been super fun partnering with him on this.

Now I want to take a step back. We've been in the weeds. We've been talking about how you navigate your own tech stack as it relates to your product and podcast, and your kind of mark on the industry as one of the more honest public voices about the gap between what martech promises and what it actually delivers.

And you don't tear the industry down, but you think it deserves bet

Reply

Avatar

or to participate