Meta’s Manus AI Enters the Chat: Is Media Buying “Dead”?

Summary

Marketing Operators Podcast Summary

Episode Title: Meta’s Manus AI Enters the Chat: Is Media Buying “Dead”?

Hosts: Connor Rolain, Connor MacDonald, Cody Plofker
Date: March 10, 2026

---

Main Theme & Overview

This episode dives into the rapid rise of Manus AI—Meta’s newly integrated autonomous AI agent for media buying—and explores whether advancements like this mean the “end” of the traditional media buyer role. The hosts unpack Manus’s real-world capabilities, its impact on media buying teams, and what the evolution of AI means for both agencies and in-house marketing teams. Throughout, they consider how job roles are shifting, what skills and intuition remain valuable, and how to future-proof yourself in the age of AI-enabled marketing.

---

Key Discussion Points & Insights

1. The Hype and Reality of Manus AI

• What is Manus AI?
  • Autonomous AI agent platform, launched in March 2025
  • Achieved $100M ARR before Meta acquired it in December 2025
  • Now, fully integrated into Meta’s Ads Manager, allowing autonomous ad account actions ([07:04])
• Market Reaction & “Media Buying is Dead” Narrative
  • The phrase sparked both panic and excitement—especially amongst digital marketing Twitter.
  • Manus' integration marks a turning point for automation but perhaps less revolution than advertised.

Quote:
“Ultimately the people saying media buying is dead are probably the people not really familiar with how dead media buying already was.”
—Connor, [14:23]

2. Should You Trust Meta’s Own AI with Your Ads?

• Skepticism Toward Platform-Native AI
  • Cody expresses hesitation about letting a Meta-trained agent handle ad optimization, fearing inherent bias toward increased spend ([08:16]).
  • The analogy with Google Ads’ simplified local business interface: these tools may automate low-level decisions but aren’t built for nuanced performance marketing ([16:54]).

Quote:
“I don't know that I want an agent from Meta telling me how I should be optimizing ... you know, it's, like, ASC or CBO uses AI, but there's still a worry about bias.”
—Cody, [08:16]

3. Automation’s Impact on the Media Buyer Role

• Greater Efficiency, Not (Yet) Full Replacement
  • For small/local businesses: Manus-like tools are a breakthrough.
  • For DTC brands and performance marketers: Only routine, “low-end” tasks are likely to be replaced soon.
  • Senior strategists and “profit engineers" get more leverage; average/bad media buyers are most at risk ([21:00], [34:38]).
• Creative & Gut Intuition Remain Crucial
  • AI struggles with strategy, context, and creative direction.
  • The best marketers blend data with experience and intuition—increasing their value ([21:00], [22:39], [31:41]).

Quote:
“I think people are maybe scared because even though they realize it's not there today, clearly Meta is going to invest in this, on making it better and better and better.”
—Taylor, [23:03]

Quote:
“Average or bad media buying is dead... But the tools just give [the best media buyers] more leverage.”
—Cody, [15:22]

4. AI as a Team Force Multiplier: The Profit Engineer & Skills

• Rise of the ‘Profit Engineer’
  • AI gives one “10X generalist” much more bandwidth—especially as brands get leaner ([34:38], [61:08]).
  • Enablement is about building flexible workflows (often via tools like Claude skills) and training the org, not just adopting tools piecemeal.
• Claude Skills & Workflow Automation
  • Cody shares his journey developing a CRO program using Claude, training it on brand voice, product data, user research, and integrating its output into team workflows ([39:17]).
  • Training (and upskilling) your team on AI tools is now a required leadership task.

Quote:
“I just feel like I live in the constant state of that Charlie from It’s Always Sunny meme... connecting the maps, trying to pull everything together.”
—Cody, [35:03]

Quote:
“One really, really good person with a ton of skills and knowledge... will be able to do the job of what used to take multiple people.”
—Cody, [18:14]

5. Approach to Team Structure & AI Enablement

• No More Classic “Media Buyers”
  • At Hexclad and other modern orgs, roles are increasingly blended: analytics, creative, strategy, and buying now combined ([31:41]).
  • The new hire is someone who can build and train AI workflows and apply analytics and creative strategy, not just press buttons in Ads Manager.
• AI Adoption Curve: Spectrum, Not One-Size-Fits-All
  • Not every brand (especially established, physical-product brands) should rush to build in-house AI infrastructure.
  • Agencies will likely lead in best-practice AI adoption and pass on those efficiencies to their clients ([58:21]).
  • All teams should at least be exploring base AI tooling for “busy work” and leveraging agency/software partners for more advanced needs.

Quote:
“I think those roles outside of just the button pushing...you have to be able to do analytics, you have to be able to do creative strategy, you have to be able to work with a UX designer and developer... that's always been right.”
—Taylor, [31:41]

6. The Risk of Overbuilding (and How to Tactically Approach AI)

• Connor argues for a measured adoption, waiting for best practices rather than always being “first mover,” especially as some AI features are only a few months old ([52:55]).
• Leadership should still keep teams focused on the fundamentals—product, customer, and performance—while selectively layering on automations.

Quote:
“I think otherwise I'm going to, like, kind of stay. I think we're going to try to stay in our pocket a little bit, see how things trend.”
—Connor, [52:55]

---

Timestamps for Key Segments

• Opening: Manus AI, Meta, and Media Buying’s “Death” ([00:00]–[04:59])
• What is Manus and Who Was Using It? ([07:04]–[09:42])
• Should We Trust Meta’s AI? & Automation’s Limits ([08:16], [11:08], [14:23])
• AI Automating Low-End Media Buying—Who Is at Risk? ([15:22], [16:54])
• Intuition, Gut, and Human Judgment in AI Age ([19:49]–[24:27])
• Comparisons: Shopify Sidekick, Infrastructure vs. Tools ([26:30]–[30:34])
• Team Structure Evolution & Profit Engineer Concept ([31:41]–[34:38])
• Hands-On: Automated CRO with Claude Skills ([39:17]–[45:53])
• Training & Organizational Buy-In for AI ([45:53]–[50:26])
• Pacing AI Adoption & Strategic Enablement ([52:55]–[55:02], [58:21]–[61:08])
• Lean Teams, Project Management, and Specialized AI Roles ([61:08]–[66:12])

---

Memorable Quotes & Moments

• “Average or bad media buying is dead... but the tools just give the best media buyers more leverage."
  —Cody, [15:22]

• “The people saying media buying is dead are probably the people not really familiar with how dead media buying already was.”
  —Connor, [14:23]

• "What David...calls gut reaction, I interpret it as more like intuition to the next data point... that's where Manus just can't compete yet."
  —Taylor, [22:39]

• “If you can hire someone that can do, like, the 20% of things with these tool-building...I think those are the brands that could really leverage stuff like Cody's building.”
  —Taylor, [62:30]

• “You don't get any bonus points for precision. If you're just directionally correct, it could pay off massively.”
  —Connor, [68:54]

---

Takeaways for Operators

• Manus and similar AI will replace “busy work” but not the core of strategic, analytical, and creative marketing work—yet.
• Job roles are converging; high-performing teams will increasingly combine technical, analytical, and creative skillsets.
• AI tool adoption should fit company context. Agencies and SAAS are best equipped to build for the masses; brands should focus on their true differentiators and selectively leverage automation.
• Ongoing training, documentation, and infrastructure matter more than Band-Aid tool adoption.
• The future growth marketer (“profit engineer”) is a builder—comfortable with AI, organizing data, and driving cross-functional strategy.

---

Final Thought

“Media buying is not dead; it’s reborn—and the humans who thrive will be those who know how to build, interpret, and orchestrate these new, AI-powered tools.”

---

For more on building AI-powered workflows, operationalizing creative strategy, and preparing for marketing’s next disruption, subscribe to Marketing Operators.

Transcript

[Connor]

Media buying as we know it is dead. It's been a big week on the Internet. So Manus came out as an autonomous AI agent platform, hit like 100 million ARR really quickly. And what happened this past week and the reason why everybody's talking about it is the Manus AI integration now goes directly to ads Manager.

[Cody]

The role of media buyer has evolved. I do think the lower end of the market should have concern. You don't need as big and bloated of a team.

[Taylor]

I think people are maybe scared because even though they realize it's not there today, clearly Meta is going to invest in this on making it better and better and better.

[Cody]

The DTC 3.0. Maybe this becomes DTC 4.0. Do think you could run a $50 million business with three employees?

[Connor]

My question is how important is gut in making decisions today? And where is and isn't it used correctly and does that change in a world of AI?

[Taylor]

I can't tell you how many ads we've decided to go invest more in where on the surface level and Meta, it looked bad. But then you look at all these other data points, you're like, oh no, this is actually a winning ad and we should scale it.

[Cody]

What I realized from last year is like just have to continue to do the training and really strike while it's hot. 6 months ago as the models just weren't good enough. Now people, it's like, oh, this copy is like really good.

[Connor]

This is going to be the topic of 2026, no doubt. All right, we are back with another episode of Marketing Operators. We got all the boys here, no guests, a full. Connor, Connor, Cody. Pod. Connor.

[Cody]

How are you doing?

[Taylor]

I'm doing really well. Yeah, no complaints. I got like six drinks sitting here. I got a bone broth, a coffee, a yerba mate and a water. So I'm set. I don't need to leave, leave my desk for the next like three hours. I'm good to go.

[Connor]

So let me ask you this because it's 8am Mountain right now. So do you just. Do you just sit down in the morning and you just acquire all the drinks that you'll need for the next like four hours, five hours.

[Taylor]

Get up, hammer some electrolytes. And I want to, once I get to the office. Yeah, I start. I start getting my stuff ready for the next few hours? Yeah. Well, Cody Cody inspired me because in the early episodes I felt like Cody always had some fun, exciting drink in Nancy. See Cody Cody's always locked up with some drinks too. So I felt Inspired by Cody.

[Cody]

So what's your electrolyte of choice?

[Taylor]

I've been on the Cadence. So we got the yerba, we got the bone broth. We got the coffee right here. We got the water over here and the simple modern, of course. So, yeah, I'm ready to go. And my electrolyte of choice. I've been on the Cadence recently.

[Connor]

Interesting. For those listening, we just. We just got a live camera walkthrough of. Of Connor's desk.

[Taylor]

Of the corner of my desk.

[Connor]

Yeah, I've just been doing. I've been doing element.

[Cody]

Yeah, I'm an element guy. I. I just got this bottle. He's used to those big, like Green Mountain bottles and stuff. But I'll. Cause I'm terrible at just drinking water. So I'll just once a day just put it. Put it in, like a big thing.

[Connor]

Yeah, totally. And I am one. I'm a one drink at a time guy. I don't. I'm not accumulating a ton of different drinks. I'm not switching between drinks. I am going my only, my only. And this is really what people tune in for is this sort of. This sort of. These sort of hot takes and information.

[Taylor]

How are we optimizing our drinks?

[Connor]

I've been doing way more mud water.

[Taylor]

Oh, okay.

[Connor]

Back on the mud water cake. I was drinking mud water like 8 years ago. Picked it back up recently. Fantastic. Instead of, like. Because I am a person that will do like four or five cups of coffee a day, so I've switched like the last three to mud water. Fantastic.

[Taylor]

I love. I love the. The mud water product. Do. Have you tried their one? They made one that's like somewhere in the middle between coffee.

[Connor]

That's my favorite one.

[Taylor]

Yeah, it's great. It's like 45 grams of caffeine or milligrams of caffeine. So you're not. Yeah, it's perfect if you need something, but you don't want to get like full coffee going.

[Connor]

Exactly.

[Cody]

I. I like Shane a lot. We should have him on one day.

[Connor]

Dude, Shane's the man.

[Taylor]

Shane's awesome. That's a great idea. And. And by the way, I do have some Trevi electrolytes too. If. If you're listening, Mike, you saw the simple modern water bottle I got. I got the Trevi as well. So I don't want Mike to think that I'm not. I'm not, you know, Trevi user as well.

[Cody]

I have, like, they came out with this Minnie Mouse one. So my daughter is 3, obsessed with Minnie Mouse. It was a Minnie Mouse Trevi. So she loves Trevi. Got it. I bought it whenever it came out. She got a Minnie Mouse Trevi. It was cherry flavor and cotton candy flavor. It's like her four favorite things in the world. I have never bought something so fast. When I saw it on Twitter, I literally bought it in like five minutes.

[Connor]

The Disney. The Disney collab on hydration products is funny. You know, you hydrate, hydrate like Donald Duck or something like, right?

[Cody]

No, with kids, if they see that, like, you, you. You don't have a choice. It's a strong conversion mechanism.

[Connor]

Yeah, yeah, yeah. You see DDC brands, you know, working with, like, Chuck Liddell to, like, provide credibility for. For three year olds needing some electrolytes.

[Taylor]

It's.

[Connor]

It's Minnie Mouse.

[Cody]

Yeah.

[Taylor]

Cadence is actually, we should. We should have the guys from Cadence on. They've done a nice job of, like, I love how they've positioned their brand. Like, it seems like they're really positioned their product to be for the serious athlete, like the, you know, the person that's training a lot. And I feel like they're the. Now you're, like, even. Even driving now. They came out with their, like, sport version, which I think is also very, very smart to, like, position it for the person that's exercising a lot. But, yeah, I think Cadence does a really nice job with their branding and positioning.

[Connor]

100. All right, well, let's get into it, guys. Before we begin, I want to thank our sponsors, Motion Rich panel, pression, after sale, and revoir.

[Taylor]

So I don't know about you guys, we have an amazing stack of creative strategists at HeistClav. But I will tell you what, it is very, very hard to find those people. It takes a lot of searching, a lot of interviews, and it's ultimately just very hard to train great creative strategists to do such a complex skill.

[Cody]

And honestly, the role's only getting a lot more complex with AI, with platforms changing. Andromeda, it's a hard job and there's a lot on these people's plates, but it's also such an important one for our businesses.

[Connor]

And that's why I always send people to Motion. Right now, they are offering a more systematic way of learning creative strategy through their free Creative Strategy bootcamp coming up in March.

[Taylor]

Yeah, I mean, this is like a more formal version of their events. It's like a true curriculum. It's live. And honestly, it's just the first of its kind. I don't think anything exists like it right now.

[Cody]

Yeah, I definitely wish I had this when I was coming up. First of all, it's free. But you're also learning directly from operators at really great brands like Drool. Com, Caraway Harry's, things like that.

[Connor]

And that's the real advantage in a complex job like creative strategy. You don't get better from watching random videos, you get better from seeing how multiple elite operators think.

[Taylor]

Yeah, I fully agree. And you really get value from being hands on, which is exactly what this is from motion. There are real sprints, there are real assignments. You're learning things and then you're doing. And that is the best way to develop a skillset.

[Cody]

If you're trying to break into the role, sharpen your skills, or if you're a founder who wants to send someone to your team through proper training. This is probably the most structured path that I've ever seen.

[Connor]

It's starts March 17th. It's free and I imagine it fills up.

[Cody]

We'll drop the link in the show notes.

[Connor]

It's been a big week on the Internet. Everybody's talking about AI Manus. We thought we'd run through it. I was doing some research in preparation for this episode. We did is media buying dead, like four and a half months ago? Something like that, Five months ago. Now we can talk about like, is media buying really dead? So maybe that, that's how, that's how we'll tee up this, this next segment. So just as a quick overview, I've got some notes here on like what Manus is, right? So Manus came out as an autonomous AI agent platform, hit like 100 million ARR really quickly. Launched March 2025. December Meta snaps it up in 2025. And what happened this past week and the reason why everybody's talking about it is the Manus AI integration now goes directly to Ads Manager. So just over the last seven, 10 days, you've had the ability to now interface with an AI agent directly in ads, can make adjustments and edits and like take action within your ad account. So I'm curious one high level, if any of you guys have any hot takes or trends around how people are approaching Manus? Or more specifically, if there are brands today who you think it's more advantageous for to be exploring solutions like Manness. So maybe Cody, I'll kick that to you.

[Cody]

My hot take. And sorry guys, if you work at Meta Love, love everyone there. I don't know that I want an agent from Meta telling me how I should be, you know, optimizing and Things like that, you know, it's, I mean, I guess maybe it's really not that different from, you know, an ASC or a cbo. Like that is essentially using AI to, to make decisions. But I don't know, I, I think that part worries me a little bit. And I think, you know, it's like, it's like that's why everyone likes MTA or incrementality. It's like it is a little unbiased that I, I, I would be hesitant to have something like that, you know, making decisions or doing the analysis. And I'm not saying they're going to like lie on the analysis, but there, there is a way to, you know, to, to, to juice it and give recommendations that could encourage you to spend more.

[Connor]

Connor, what do you think?

[Taylor]

Yeah, I, well, I wanted to ask you a question, Cody. Cause I remember you went in there, like, I, I couldn't tell if you were, if you were full trolling when you tweeted this, but like, about the recommendations that it.

[Cody]

Oh yeah, I didn't actually do that. Yeah, it was just a joke.

[Taylor]

Okay.

[Connor]

Yeah.

[Taylor]

I mean, you know, actually I, can I ask a question before, before I answer your question. So Manus got to 125 million in ARR prior to Meta buying them. Right. So who, who is paying Manus, like, who was using Manus like originally, like in the previous year or two, prior to Meta buying it? Like, is this, I'm trying to understand their business model a little bit here and like, who is using the, the technology?

[Connor]

It was an agent and it would like, it could take action with like some, it had just a bunch of integrations. It could integrate into like your calendar, your email. You could theoretically just do a bunch of interesting, have it do interesting autonomous tasks connected to all these different things.

[Taylor]

Got it. Okay. So it was consumers paying for it, and that's how they got to that, that hundred.

[Connor]

Probably some of like prosumer. Yeah, right.

[Taylor]

Got it, got it.

[Connor]

Yeah.

[Taylor]

I mean, I'm excited to get into like some of the things that, you know, I think it can do and can't do. I put some notes in here, but

[Cody]

like, have either guys used it? Have you guys played around with it?

[Taylor]

I have not, I have not played around with it.

[Connor]

I was using it last night a little bit.

[Taylor]

I just think we're like, we actually, we tried a tool like this last year where it was like automated media buying and it didn't do poorly. It was really hard to say, like, if it outperformed or underperformed Having, like, a real person run the account, like, it was like, we're looking at period over period comps, and you're trying to look like the year over years in the different periods, both, like, blended and in the account. And there's just so much changing. And there wasn't. It wasn't like, obvious that the. The automated was doing better. Listen, I think there are certain things that we'll hopefully be able to do, but I don't think it. I don't think. I think it's like a utopian outlook to be like, the media buyer's dead in. In the next 30 days. I don't think that's. That's where this is headed.

[Connor]

The dystopian outlook, actually, sorry, the D, at least for our little podcast in our corner of Twitter, is like, this was. This was like DEFCON 1.

[Taylor]

Well, maybe for, like, maybe utopian for a brand owner and dystopian if you are. If you are a media buyer. But, like, I think there are certain things that, like, if you could go to it and say, hey, like, I want you to keep budgets at this overall level, but I want you to, like, optimize, like, within the account at that level, I think that could be really interesting. I think one of the use cases I was thinking of was category spend. Like, if I, If I said, hey, I. I want you to always maintain like A. Like a 70, 15, 15 split across, like this category, A, B and C. And like, that moves up and down. Like, I think things like that could be really interesting because that's like serving an overall business need. And if you could take like, the manual nature out of that, I think that would be an efficiency gain. I think building ads would be like. I think that's one that we still haven't solved. And if. If you could just like, upload all your media and copy and have it build the ads and like, almost like voice, like, almost like text prompts, if you, like, drop them in this campaign instead of having to, like, glitch your way through the ads manager. But that's. That's where I think is like, I don't even think we're there yet, but I think that that's what I see as being like, what could practically add value to a brand or media buyer through this. Through this tooling right now.

[Connor]

All right, so I've got some interesting numbers to share from Ridge from our beloved sponsor, Ridge Panel. We switched our support stack at Ridge to rich panel about 15 months ago, and our cost per ticket has dropped over 70%. That means same team, same volume and over $500,000 in annual savings. CSATs have not changed. We've been sitting at 96% week after week. So the automations did not come at the cost of customer experience. Our last platform talked a lot about AI, but nothing was really changing under the hood. Rich Panel is genuinely AI first. They came in, they rebuilt our workflows and we were live in under two weeks. Basically no lift on our side. And they're about to roll out a returns portal which for us is huge because if they can do the same thing for returns as they did for customer service, that would be sick. If you want to cut down your support costs now and save on returns platform with an AI first platform, talk to rich panel, go to richpanel.com demo tell them Connor from Marketing Operator sent you and they'll take good care of you. Thank you. And I want to come back to like hitting what are the tasks that we think something like manuscript will be good to utilize for but like to. To go a little bit higher level? I guess I was somewhat surprised. Cody's original point was I'm not sure I want like a meta trained agent to be optimizing my meta ads. We spend a lot of time on the podcast thinking about like not that, not that our goals are completely at odds with what meta wants us to do. I actually think it's like probably one of the more brand aligned platforms versus something.

[Cody]

I agree with that. I agree.

[Connor]

You know what I mean? But the, the, the most optimal form of media buying is typically not just strictly best practice. Do exactly what meta tells you to do. That feels like a pretty not hot take to me. Like pretty like oh yeah. That's exactly what I would have expected to hear from most people. So I was kind of surprised by like the visceral reaction that people had to like, you know, lots of people, oh, media buying said media buying as we know it is dead. Do you think. Why do you think that is? Why do you think that is? I have a theory and I'm happy to start. If you guys want to hear.

[Cody]

Please start.

[Connor]

Ultimately the people saying media buying is dead are probably the people not really familiar with how dead media buying already was. Like I heard it actually I heard it from our friend of the pod boring marketer who came on talking about extremely early, talking about Claude Code last year. He said media buying as we know it is dead. And I think media buying as we've known it has basically been dead and that what media buying has become is like finding ways to truly align the meta platform with what your brand is trying to drive. Whether that's incremental results, moving through inventory, you know, growing different lines of the business that might not be a short term return but like actually be thinking about a return on a much longer timeframe. Those are the things that media buying has slowly become over time. And Manus won't do that, can't do that. It probably ends up being way less sexy tasks like Connor Rowling mentioned, like it'll be building ads for us up

[Cody]

until now, until this point and probably still to this day it's like average or bad media buying is dead. It's the same thing with like you know, I felt like this until now. And we'll talk about some like the website stuff I'm doing where like design it's like you know, you can, you can make static ads, you can make landing pages. Like you can replace like it's almost like you could replace the lower level. Like the like, like David Herman has nothing to worry about, right? Like he's you know, really great media buyers. Not dead. The tools just give them more leverage and, and help automate some things. But like even with asc, like if you just think how things went, it used to be abo individual audiences and you had to actually media buy and, and pick those things and it was like it has gotten more automated and the role of the media buyer has evolved. Like it's still like you can't write current today train an AI to be an expert media buyer. I do think the lower end of the market is, is should have concerns though. Like I think that a lot of not great media buyers and they're probably a lot of not great media buyers. Like I think it can make better decisions like that. So that's how I felt about you know, E Com Web. That being said, one of the reasons I think people are so in a frenzy about this stuff and excited, myself included is like these models have just gotten so good that I think it's debatable whether they're there now or not. A lot of the stuff they are coming for because they have just gotten so good and in some cases is producing superior work to what people are doing. Some cases, not everywhere still need people totally.

[Connor]

So on the point of the, the low end media buying maybe being like completely abstracted away, it reminded me a little bit of I don't even know if they still have it and we never talk about it but I always thought it was interesting that Google Ads used Google Ads Ads like the one that we use that you media buy out of. Like any brand spending any amount of money is using like the pro version of Google Ads and then they have the like simple version for like local businesses. Have you guys ever like set up a new ad account and got pushed into like. Yeah, the like it's got all the bubbly ui, way fewer buttons, way fewer metrics. It's basically just like the spend money button that you just press and it turns it on and like Manus is probably. That's a better comp for what Manus is. It's actually like oh yeah, if I'm just my local barber shop. Like I'm actually never even going to touch ads manager. I'm just going to interface with this, this chat that I can give it some images, give it, I don't even need to give it copy. I'll give it my website. It's going to generate ads, it's going to generate copy and I'm going to be spending money. Tom is fantastic for the barbershop business but like in many ways, at least for the foreseeable future in my opinion is it's not going to be making any meaningful changes and definitely not in an automated way from like a performance DTC brand.

[Cody]

I do think, I do think and I'll talk about this a little bit but like I think Mercury must be in retrograde because I'm going to agree with Taylor Holiday twice in one week. But like I don't agree with the whole like profit engine. Where I don't agree is on the creative perspective. Like I just am not as bullish on like there's always going to be a need for real content and I think there's going to be a bigger need for real content. So like I don't think one person's going to just do that but I do think one. It's like the 10x engineer concept and we're seeing it like internally and it's going to change how even we hire. It's like one really, really good person with a ton of skills and knowledge and stuff will be able to do the job of what used to take multiple people. And so I think one like you can just need less media buyers and I think that's true today than it used to be even to the point of uploading data analysis. Like that stuff is really what we should be trying to automate. We use ad manage. I don't know how much but it has definitely helped my team. I'm you know, happy to like ask Them if you're curious. But like we use that for our ad uploaded, right? And so that has helped to automate a lot of it where we would need more junior level people. So like that's one component. It's maybe not a hundred percent, but maybe it's saved. You know it saved 80% of the time of uploading ads. And then there's obviously analysis stuff. So I just think one person can, can currently and then in the future, you know, do do the role of five people. But, but that one person has to be then almost higher level more like you, you only get usage out of these tools if this one person has the context and the strategy as, as well. But they can just also be the executor 100%.

[Connor]

Yeah, I mean that's, that's definitely the, the future that we're turning towards. I will say just because I'm sure Taylor will watch this episode. He's, he's very bullish on people continuing to produce the content. I rewatched our, our. I rewatched Taylor's reaction to our is media buying bad episode and he was very clear. He thinks all the like leverage or a lot of the leverage accrues to this one person at like the strategic layer and not necessarily like the content production layer. So let the record show I'm defending Taylor Holiday and his positions. Okay, let me, let me ask you guys this. David Herman, another friend of the pod, in response to Manus, he goes the difference in my opinion on having Manus or any AI run your media buying always comes down to the gut reactions versus data reactions. AI can do the job of a basic media buying agency and probably outperform it in some extent. That's kind of that, that low end that Cody's hitting. But his ultimate point is like the ability to execute on some sort of. He says gut reactions is important and I'm curious how you guys take that. Like my question is how important is gut in making decisions today and where is and isn't it used correctly and does that change in a world of AI?

[Cody]

So I think again like David Herman is known to be like best of the best, right? And so it's. He has intuition, pattern recognition for probably 15 years of doing this that like people don't have. And so like that's the standing. So again like Manus is not coming after him. Right? You know what I mean? Like he could be, he could probably be across way more AD accounts if he wanted to with mana and ad manage and stuff like that. Like obviously he can build teams. The, the one thing I so I 100% agree. Right. Like for example, I know he'll look at things like you know, he's not just looking at roas, he's looking at time on site and things like that and like trying to kind of put pieces together and he'll look in North Beam and ga to see like hey, maybe this ad is great for the customer journey but it's not necessarily pulling in, you know, I mean I feel like theoretically you can train things on that. But I think there's probably a lot of those things that are, you know, decisions that gets made. The one maybe counter is like I think with anything with AI it's like it's almost like compared to perfection. It's like anytime a waymo gets in a crash it's like that makes big news but it's like people getting crashes all the time. Like even at companies like there's there, there's stuff even at nine figure companies where like media buyers don't always make the right decisions and we just kind of don't necessarily know it. But like there's a lot of things where like we're even trying to change how we are approaching media buying and the metrics we look at because we think some of the things that we've done are potentially hurting our account for future. So it's like yeah, like yeah Matt, man, I shut off that ad that it shouldn't have or it scaled that one that I shouldn't have. But like that also happens at companies all the time as well. So it's just not a, it's not a fair comparison.

[Taylor]

I think what David, like what he calls gut reaction, I interpret it as more like intuition to the next data point. So like Cody, you started hitting on it but like you know, you know, if we're looking at ad performance for example, we're going to be bouncing around between the ad account and North Beam and a media mix modeling tool and server side like tracking tool. Like that's what I interpret as gut reaction where there's like this intuition you develop over time where you know like which data point to look at next. So it's not like I think when he says gut here I didn't interpret it as like vibe, vibe marketing or vibe performance marketing. I interpret it as like no, like having the intuition to know what to look at next and how to piece it all together. And that's where, that's where man is just like I don't think maybe it'll get there. Maybe it won't. I think that. I think people are maybe scared because even though they realize it's not there today, clearly Meta is going to invest in this, on making it better and better and better. But I fully agree with David. I mean, you can't replace that intuition of knowing which data points to look at, which platforms, and then ultimately what do you do because of that? I can't tell you how many ads we've decided to go invest more in. We're on the surface level in Meta, it looked bad, but then you look at all these other data points, you're like, oh no, this is actually a winning ad and we should scale it, we should do more about it, that we're so far from that, from that. Be that person that can like, do that and then coordinate the next steps with the entire team from, from being replaced by this or like so far away. But, like, who knows, maybe in three years, five years it can, it can integrate with all these different data sources. You can tell it like, hey, I care about these KPIs from Northbeam and these KPIs from Prussian and these KPIs from Google Analytics. Like, maybe that's, maybe that's coming, but that's. It seems like we're very far away from that.

[Cody]

I also think the same way, there's M.M.T.A. tools and MMMS and incredibly, I think there will be a suite of software tools, whether it's new ones or existing ones, that will kind of do the manus. I mean, I know Connor, you said you tested one like a year ago, but I think that'll be like an offering that we'll see more and more of. And I think they'll be built on, like, they might be a little bit more customizable, I guess, and built on the underlying data that, you know, these softwares have. So I'm excited to see that happen.

[Taylor]

Operators, quick gut check here. Q1 is when everyone realizes the same thing at once. Traffic gets more expensive, growth slows down. So the question isn't how do I drive more demand, it's how do I make more money from the demand I already have? That is exactly why we use after sell by rocked. Most brands think upsells are about being aggressive, but they really aren't. They are about timing. And the best time to upsell someone is when they're already in buying mode, which is when it's right after someone buys. So after sell, it lets you put the right offer at the right moment. With one click, there's no reentering Payment info. There's no extra checkout steps. Brands using after sales see around a 30% lift in AOV and when you're running real volume, that adds up fast. But here's the part most people miss. It's not just upsells that after sale ads. Once you're live, you unlock the entire rocked monetization suite. Rock thanks monetizes your thank you page with premium non competing offers. Think Disney plus hello Fresh and brands are seeing $0.30 to $0.50 in pure profit per order Rock Pay plus as a clean wallet placement at checkout and kicks back another 10 cents to 15 cents in profit per order without hurting conversion. And in some cases it actually improves conversion. No inventory, no new ads, no operational lift, just margin. This isn't growth hacking, it's just found money. If Q1 is about tightening margins and getting paid more for the traffic you already earned, go to aftercell.com operators, activate Rock thanks or Rock Pay plus and you'll get the full after sale suite free for a year or an extended 60 day trial for post purchase upsells.

[Connor]

So this was, this was another like good comp I thought for Manus was Shopify Sidekick where like Shopify Sidekick is extremely interesting. It's like Livic within Shopify. It has all your Shopify data. Um, you can have it pull like pretty basic sort of analyses and, and data points and things like that. But it lacks, at least from my experience, it lacks a lot of the context to do any more granular or comprehensive analyses. So like you know, we've been using Sarah's for our data warehouse for a while and Sarah's is all the context that we want. Right? We have our full, we have all of our Shopify data, all of our Amazon data. We also have a big like product catalog master SKU table with all the product attributes that we want. So if we want to do an analysis of this product category for this time period on this channel, Sarah's has all that data and it's that context that actually gives us. It's way more actionable if we just think about like I want to understand what is being sold on my store and Sidekick just lacks that right now. And it's like there are many examples where Sidekick can be extremely valuable, maybe simpler data analyses. I can imagine a future where Sidekick is doing sort of rudimentary E. Comm Ops related tasks within Shopify, but it's lacking that larger context. And I do agree that's kind of where Manus is at. It's going to lack all the context of your geolift studies and like what your objectives are as a business and ultimately that a lot of those like more meaningful, impactful media buying decisions will become agentic or will have AI tools to help support it. But it's going to live outside of the meta sphere. I think for the vast majority of like large, more complex, you know, performance oriented brands.

[Cody]

Yeah, yeah, no I agree with that. I think the, the future seems like it's like APIs and like not necessarily. Yeah, like getting all of the stuff into one place. Whether that's a Claude, they're obviously certainly trying to go there and actually this was one of your predictions by the way.

[Connor]

I know I put it in the doc.

[Cody]

Oh man. I just realized that you were. We thought it was going to. I don't know when you, when did you think it was going to happen? Because I feel like it's happening faster

[Connor]

than this was my prediction and this is one of the reasons I bring it up is I'm biased as far as like where I'm spending my time on AI is like is with this in mind but my prediction was brands realize the key to unlocking AI across an organization is not through one off tools but with infrastructure and documentation. Knowledge work looks more like cursor with

[Cody]

model agnostic structure where damn, you were right about that.

[Connor]

We can sit on top of all the other tools and that's where we want to be like unlocking value with AI versus within meta with Manus, within Shopify with Sidekick. So I think that's where it trends.

[Cody]

You're. You were spot on. I mean I'm seeing it like we just all share about it. Like we just got our whole team on cloud for that reason. Like it seems like it's going there but like I'll give you an example. Like intelligence, right? Like that's one where it's like, you know, we're testing a lot and they have an API or an MCP and like we are in myself again because I'm busy. We're able to go deeper on the insights and the data than we would have normally. And I know you've talked about this on the pod for like a few years of you know, exporting all of the orders and putting that into, you know, Excel and having an analyst go through that. Like I can't do that but so now I can and so I can just go so much deeper and like not accept surface level. Oh, this was better conversion rate. But it's like why what was like our order distribution did our you know, like stuff that you would normally have to know how to, you know, write code or SQL to do. So I just think like that. And then again, you can pull in other sources if you have other data warehouse connected. So I just, I agree with you. Having it all in one place with the shared context and then obviously you can build skills and things like that to kind of analyze it how you want. I think that's the future. I mean, I built a meta ads MCP because I was hitting the rate limits on a free one I was using. But so that has, that is in there. And then like, you know, you can, you can have other business contexts in there. You can have your forecast in there. Right, Totally. You know, so, yeah, I 100% agree. You were very, very right on that.

[Connor]

We'll see. We'll see. It was a very. I give myself the advantage of sort of a vague, just directional prediction. There's going to be like, no, no, like hard, hard, like yes or no.

[Taylor]

Which is true.

[Cody]

When Stephen A. Smith 2026 predictions, I think AI will improve and take over. Yeah, yeah, yeah, yeah, yeah. Right.

[Connor]

When it's Stephen A. Smith runs for President in 2028, that my 2025 prediction will like, have fully come to fruition. But, but yeah, this, this one, it'll be like sort of a grayer area.

[Cody]

Never wrong, just early.

[Connor]

Yeah, 100%. Okay, so maybe. Connor, I'll kick this one to you. We've hit this point all the time. It was what the episode five, six months ago was about with his media buying dead. But at Hexclad, and in that episode you were talking a lot about how you guys don't have a media buyer title. At Hexclad, it's a little bit broader. There's more responsibilities that you expect out of the people who happen to be media buying. Over the last six months, with all the AI tooling, has anything changed now or do you foresee it changing, you know, by the end of the year?

[Taylor]

Like, the reason we don't have like classic media buying titles because we don't have a single person who their only responsibility is, is media buying. So like, if you look at our, our YouTube buyer, I mean, he's buying all of our, all of our YouTube ads across all of our markets. He's doing regular analytics reports on what's working, what's not working, and then he's ultimately doing a lot of creative strategy with our creative strategy team to produce what's next. So, like, that may, like that person becomes even more valuable in this like era of manners. Like, I think that's one of the, that's one of the notes that I wrote down. I forget this was in, this is in your GROK summary, but it was like elevating those who focus on broader growth tactics. Like yes, 110%. Like, but that, that's no different than it was a year ago. Like that's already the direction that we've been trending. I just think it like kind of re emphasizes that. And like same with our paid social buyer. Like, yes, he's pulling all the levers in the ad account, but he's doing detailed analytics reports every single week. We're talking about those analytics and what we're seeing and like what we should do because of it. And then he's going and facilitating like the next steps on that. So I don't think this changes any of that. Like we still don't have like a true media buyer title for that reason. I just think it, you know, maybe bolds it a little bit more and the importance and I think that's what the, like, yeah, I just, I think those roles outside of just the button pushing, like you have to be able to do more than just pull the levers. Like you have to be able to do analytics, you have to be able to do creative strategy. You have to be able to work with a UX designer and developer to be like, hey, based on this data, I think we should do this landing page. Like if you're just getting less and less valuable if you can't do those things. But that's, that's always been right.

[Cody]

It further takes you away from media buying.

[Taylor]

Yeah, exactly.

[Cody]

Like profit engineer.

[Taylor]

Right, Exactly. And I think, you know, the profit engineer thing, I love that I do think is a little utopian to think that you're going to have like one to two people that can like basically set up this infrastructure of AI to kind of like do everything that a team of 10 is doing. Like I think you can absolutely eliminate roles. I think you can be leaner. But I don't think we're there yet. I don't think we're to the point where like a single person is running growth for like you know, a hundred million dollar brand. That seems, that seems like crazy to me. But yeah, so I don't think anything's really changed on, on our end on how we're thinking about these like roles that do media buying. But they're not, but they're not, you know, cut and dry media buyers like you would find at Maybe a, a meta ads agency or something like that.

[Connor]

Totally. Okay, so I think that's a good segue. So we, we hit Manus. We hit sort of like the more recent trends in, in media buying and some of the meta tools. You, you said the, the keyword here. Profit engineer. Cody's been, Cody's been a profit engineer. Pilled. So can we talk about this tweet and then jam on like more of the broader implications of AI and how it gives leverage to, you know, a smaller group of employees within a brand?

[Cody]

Yeah, I mean I'm, I'm, I'm totally pilled on it. Like it's funny because like yeah. Six months ago when we had what's not. Not Borrow marketer or whatever his Amazon. Like yeah. I was like I had no idea cursor could be non code stuff. And I was like that's so technical. But I guess I have just, I, I just feel like I live in the constant state of that Charlie from It's Always Sunny meme. Like that's just like how I feel constantly. You know what I'm talking about, like

[Connor]

where he's connecting the maps or whatever.

[Cody]

Yeah.

[Connor]

So can you, for those listening, can you just talk like talk through this tweet and then maybe give the background a little bit of like how you got to this realization.

[Cody]

Yeah. So Taylor Holiday did a reacts episode, I believe it was to us talking about our team structure, our growth, team structure, things like that. And I think his take was you guys are bloated and you have too many people and one person is going to be able to do all of this. I don't want to like overgenerize or something. Yeah. But it was something like that and you know, I didn't necessarily agree at the time but like it was going in that direction. I think again my, my, my pushback is maybe more on the creative stuff, but I think a lot of it is totally right. I think we've seen it. So I would say E commerce is definitely behind tech. Right. Like where it sees as like cloud code came out a year ago but like really started to get like really popular and cursor and Windsurf and all that stuff, like more you know, like earlier than it hit like our circle. Right. And I think because it's obviously better at code based things and other things now, people are waking up that cloud code can do so much more than code. But what I think people have seen there is like there's just been a convergence of roles and pm so product managers can now design.

[Connor]

Right.

[Cody]

And engineers can now product manage and you know, things like that. And so just these roles are kind of starting to come together. That used to be very separate skill sets where it's like you know how to code, you know how to, you know, whatever project manager does, run experiments, spec, talk to users, write briefs or you know, how to design. But like you almost don't have those same things and those are just kind of converging into one, you know. Um, and I think the same thing is starting to happen to us now. I think in different. I can share why in. In different orders. Like I think, you know, creative is actually lost.

[Taylor]

We're.

[Cody]

We're doing a lot of like a ton on creative that I, I think there's a ton. Then Dara talked about it or like you can do that's not making creative. Like I'm still so bullish and lo fi content, things like that, real content. But you can still do a lot of reviews, analysis, briefing, all that stuff with AI. But I think like, you know, E. Com makes a lot of sense. Retention makes a lot of sense because you're not creating video assets that you're trying to replicate people. But I think, you know, in this example, you know, E. Comm 0 like if we take 0 like that's actually the one I've been really working on. Right now I'm doing it by myself and I can kind of share the workflow but able to do what a multiple person, team or agency used to do and I would say at the same or better, better quality and output and it's not a hundred percent there yet, but it's. It's close.

[Connor]

All right, this episode is brought to you by beloved sponsor Revo. What is Revo? They are a retention platform for Shopify plus customers. They do accounts, loyalty, referrals, membership, cashback, wallet passes and so much more. Stuart Chaney's the founder. He's very public on Twitter, building best in class retention features powered by AI, quicker and with higher quality than anybody else in the business. We are lucky to have them as a sponsor. Revo is working with some of the best brands in DTC like Ridge, Hexclad, True Classic, Dr. Squatch, Portland Leather, Fenty Beauty and hundreds more that are on Shopify plus and want to provide better experiences for their customers. I think retention is going to be a huge trend in 2026 and beyond. All the effort that we put into incrementality and acquisition will eventually get diverted to retention so that we can continue driving incremental Sales and Revo as an account solution will play a big part in that. If you want to become a Revo customer, go to Revo IO. And right now all Shopify plus brands get wallet passes 100% free on the platform. Tell the marketing operator sent you. I would love to hear a bit about the workflow and then my kind of like the next question is if you have any, if you have thoughts on like who is the type of brand who should be trying to build the systems that you're working on right now. And maybe the answer is everybody. But like that's one path that I'd love to go down. And then the second one is you sent me a couple of loons that I watched beforehand and I'd love to talk more granularly about like Claude skills and the idea of that. So maybe you could start high level, talk us through like the, the CRO workflow and then we'll segue into the questions I just mentioned.

[Cody]

Yeah, so it's just kind of, I was just like, you know, playing around with, for, for fun, but I'm like, let's see if we do something valuable. So I'm like, let me see if we can do any, you know, CRO stuff with this. And so built skills. Skills are, I don't know how to describe it, but essentially like shared instructions kind of. And you can train it on whatever context you want. And if, if you do something once and then it did it great, you could be like, save that. Remember that every time, you know, this comes up, I want you to do it. Or if it took you a few times to get an outcome, you could be like, hey, save that. Do it the first time. So yeah, I was just playing around with it. I was like, let me just start feeding it data and seeing what it can get access to. Like let me just see what it's capable. So essentially fed it. So there's an intelligems API or mcp. There's not a heat one. But I was able to just have Claude take over my browser if you haven't like that's such a fun place to start is just use Claude. Like I'm using the desktop app, but like use the Claude in Chrome and just watch it do stuff in your browser. It's just, it's like mind blowing what it can do. Also one of the things you just ask Claude, hey, this is what I want to do, like build me a plan. What data sources do you have access to? Which is really cool. So did that had a Slack channel with former, you know, CRO stuff, gave it access to that. Just went through all of that. We had a bunch of looms. I gave it transcripts of the looms. I had some stuff from Dylan Anders. So, like, transcripts so you can train down whatever you want. I gave it a little ourself of like, here's the framework, things like that. We're trying to do a lot of surveys, user research. So I fed it, you know, user research as well. Right now, the beginning, I just give static stuff. Like I would give it CSVs or whatever. But in the future, like we use Typeform. I'm gonna try to pipe that in so any new stuff gets updated. But essentially gave it that and was like, oh. And then. And then had it, you know, create a document for an ICE framework, right? A impact conference effort, you know. And so the first one was just like, hey, can we create a roadmap? Maybe it's just in a spreadsheet, like, let's see what a roadmap looks like. And it was pretty good. Like, it was. It was as good, especially for like the first try, one shot of like, what we've gotten from an agency. Like, it was good tests. It was built on the right framework, which is essentially looking at higher traffic pages, looking at, you know, just the right elements. It wasn't like, hey, I'm going to change this button color. Cause I trained it on those. So that was the first step. Got a roadmap. And I was like, all right, what, what, what can we do with this? I was trained in a lot of like Jones road brand tone and stuff like that. So I kind of just kept building out. I built a UX for the roadmap, which is like, that's not the important part. Like, anyone can build it. Like, it doesn't matter where it lives. It's really what. It's the strategy that it's trained on. And so just kind of building it out. So now we can go from ideas. I'm building like a way that like anyone on my team, we have like an E Comm ideas channel. Something gets tossed in there. Claude automatically adds it to the roadmap, scores it against ICE framework and so can be. Oh, that's an idea. Like, I did this in. You can also have these skills inside of your cloud account, where I had an idea, just wanted to test images on our PDP and. And see. And so it's like, what do you think? And it scored them. ICE framework again, like anything with AI, you gotta, like, you gotta make sure you're not just getting like slop outputs but hallucinations. But it was like, yeah, like based on these factors user research framework, it's above the fold, there's a lot of shade matching anxiety, you know, this is it. And I was like, all right, recommend the 5, recommend the 5 images we should do in order to it again used all of that stuff that it knew. It's just all of the stuff it's trained on. And then I was like, all right, write it, write a brief for me. And it wrote a brief and I sent it to my team and you know, well, we will now execute it manually. So like I think there's a lot of human in the loop things where it's like that probably is doing 50%. But what I've now been trying to push is design and dev and I mean I can't tell you how much time I put into it. So it's not the best use of my time. But I'm doing it, you know, five minutes between calls and then Claude goes, does it. Or it's more like I had, I had six hours on a plane last week, so I was doing it then, but can now get what I think is pretty much pixel perfect design and dev Where I'm able to build landers, I'm able to take an inspo of a lander, ask it to build an HTML file. I can either push that to Figma like so I did this one, I literally have a page live right now that went live today where I can, I can do that, I can push a page to Figma, get a designer on my team just to make some slight edits to it and then I can push that back to Shopify. And I did that yesterday and you know, still ironing out the kinks, but every time I'm doing it, it's just quicker and quicker because you train those skills and update them or going directly from, you know, HTML to do that. So I'm at the point where I can write Shopify code and again it, it has taken a ton of time to get there. So why I think that's a good use of my time is a. For learning, setting the culture right I'm and, and building the, this infrastructure, these skills to then pass off to somebody in my team. So the plan with CRO is we're going to hire one person that can learn this stuff or do it and one person we think will be able to run a CRO program for us. And maybe you need some design but instead of needing now a full Fledged UX designer. Like you can get a, a junior designer who's just, hey, that's the right font. That's the wrong font. That's like, I need to put this image in. But you don't need like as big and bloated of a, of a team. So I, that's one area where like profit engineer is happening. Like, we're gonna have one person running our entire program.

[Taylor]

So you built this tool because one thing I like is that like you as the leader of your org, like innovated and built this tool, but it sounds like you then successfully like integrated it with your team. Because what I heard you say is that you actually are now using the tool. So now anyone that has a CRO idea, they can actually like submit a ticket or something. It runs through this tool you built and then it like ice scores it and then I presume it puts it in that spreadsheet so that, so you've actually have you successfully. Because I think that's, that's the, that's the, that's phase two, right? It's like, all right, you built the tool, it's working how you want it to. But then, you know, as you all know, like an orga of our sizes, like it could be hard to get the entire team bought in and using it. Like, you have to train the team and answer questions and then the team has to use it. So that to me is like, what almost makes me the most excited is because, you know, Connor, that's what I was going to bring up when you, when you kind of relooked at your prediction is like, I wanted to see an example where a tool was built and then integrated with the team. And then the team's coming in and using it and connecting all these sources. So, Cody, that's like, that's pretty cool. Was that a, what was that process? Like, did you just have like a training session on, on how to like use it or like, how did you get that, that integration and buy in from your team?

[Cody]

So this one I'm still doing. What I'm gonna do is I'm gonna run this myself for a month and then I'm gonna iron out the kinks and then hand it off to somebody. Yeah, or, or hire somebody for it. Backing up a second. So we did like a big AI initiative like six months ago. Like talked about on the pod. I would say it went okay. We did like a bounty. Some people did really cool stuff. Some people just use like ChatGPT. But our ops are procurement manager. This was before it was even that good. Like vibe coded in Ops dashboard. That saved us a ton of time and money. So, like, some people did cool stuff, but it was kind of like optional, you know. And this year I was like, we're just gonna make it mandatory. It's gonna be part of just your, your, your evaluation, it's gonna be part of your review. It's not mandatory. Like, this is as me as a leader, like where the companies go and where I see things going. And it's just gonna be part of our basic expectations. But we also have to put our money where our mouth is. So we signed up everybody for Claude. I promoted our data analyst, director, data Nai. And so he's leading weekly mandatory trainings. Like we just did the first one teaching people, because you also have to give them the skills as well. You know, some of it is obviously the culture of like, hey, your jobs aren't going to be replaced. We're just trying to level you up and things like that, you know, and

[Taylor]

so teaching in those, like one of those trainings on like, what is it? Like, different every week. Are you like, are you ever, are you ever going to him and saying, hey, look at this tool I built. Like, you're training him and then he's training the whole team. Like, what is he doing during those weekly sessions?

[Cody]

So it's like we're gonna do like an eight week. Different people are at different levels. Like, some people hadn't heard of clouds. Some of it is just, hey, this is what it is. This is why we're using it. This is, you know, like, cowork is the new. Like that was the tipping point for me was like, hey, this is cloud code. But with an ux, anybody can do like one thing I did when I first, you know, when coworking, I got it, like I had a meeting, we had like a hour and a half category review and I and granola, which we use for notes, just got integrated with it. I just said, and what are the follow ups from that meeting and what can you help with? And it gave me a deck that had the answers and I sent it to my team and like it was on the effort and skill was like as low as possible, but the value was so high, right? Like it answered things. It like it looked through all of our reviews as well and analyze them. And I was like, there's no one on our team that can't do that.

[Taylor]

Right?

[Cody]

Like, I don't expect everyone to learn cloud code or learn cursor, but like, there's nobody on our team that can't do that. So, so what he's doing is like an eight week thing and it'll become optional over time where it's like, this is what cloud is, this is what connectors are, this is how to use them, this is how to use skills like just like, just teaching like 101 and then 201 and then 301. I would expect. Um, this is how I'm currently thinking about it is like a few people in the org to build stuff. Like I'm building stuff. He's building stuff probably four or five. Like our systems manager built a iPhone app the second day of having Claude that, that scans barcodes so our retail teams can use got it as like a custom Shopify app on like his second day having Claude and it'll save us like 10 grand a year on netsuite licenses. So like I expect some of that. So maybe four or five people who are actually going to vibe code stuff. I expect everyone else to use tools that we provide for them. Like I don't exactly expect everybody to vibe code apps and stuff like that, but I do expect like right now my head of growth is trying to create a skill for just, you know, she does a, three times a week just pacing, analysis, forecast analysis. Like so she's just trying to document everything so she can just make that quicker, you know. So I think it's been a combination. But definitely I think what I realized from last year is like just have to continue to do the training and really give the resources and like not be cheap about, you know, like we're not doing shared access and we're giving people like the, the credits that, that they need and we'll have fun. Like we're going to do some like hackathons and like get pizza for everybody and like just do some fun stuff. But it's, it's been, it's been. There was a lot of trial and error before and now learning. Um, and the other thing I'm trying to do is like strike while I'm. It's hot. Where like first of all it's, it's good. I, I think the challenge was six months ago is the models just weren't good enough. So it wasn't like the copy wasn't like it didn't make people's lives easier. Now people, it's like, oh, oh, like this copy is like really good. So like somebody came to me last week, it was like, hey, we're having, we're having delays with copy and I'M like, let me strike while it's hot. Let me get the skills that I built into it for everybody. You know, shoot a loom, show them how to, how to do it. And so it actually makes their jobs easier. Now I think that's one of the big keys

[Connor]

on the point of skills. I would love to hear just in that instance specifically, you know, somebody needs copywritten. They're not happy with what's getting done. What does the skill do that just, you know, typing into to Claude doesn't?

[Cody]

Yeah. So the, the skills are trained on, you know, context. This is why it's so high leverage. It's like it's so much time upfront to build proper stuff, but it just pays off tremendously. And you know, I said in the limit. Like some of the stuff that I built skills for for one department is actually proving valuable for other stuff. So some of it is brand tone. Right. And so you can. The trick that I've learned with Claude is just asking Claude, hey, this is what I want to do. What do you recommend? Right. And so you got to train it on your brand tone is super important, right? Cause you've all seen like AI copy. That's not, um, and so just give it a ton of stuff. So if you have Brandon docs already, give it that. Right. But if you, if you don't give it a lot of stuff that you like and ask it to analyze it and it'll create a markdown file. Markdown is essentially just like the, the, the language or the format that, you know, these things need things in. And so you just really create all these markdown files. And then from the markdown files, which is like the reference and the, the training documents, you just create a skill that's essentially do it this way. And so trained on brand tone, stuff like that. It's trained on products, it's trained on reviews. Again, you can have AI analyze reviews, pull out sentiments, themes, things like that. We have our outer signal Persona data in there and have enriched that a lot and then channels. So you probably want to speak differently on meta versus on a PDP or things like that. Right. Both from a tone, from a length perspective. Right. You have to give it constraints and also from a. Right, like levels of awareness. Like one, one challenge that, that, you know, I had before is like, we have a really good copywriter, but she's, she's more of a brand copywriter. And so like writing landing pages took way too many revisions. And so now I just put, I just created a framework and a format for listicles. And it's built on all the brand stuff, Personas, but it's also trained on how we think we should be writing Listicles. You can select the level of awareness.

[Taylor]

Right, right.

[Cody]

Problem aware. And so, like, you can kind of just do these things up front and then hand it off to somebody and be like, you know, use this tool, use these skills. And it's like, it's like training, you know, you're kind of training, and it kind of like gives it within that constraint. So that's. That's kind of an essence of what it's built on. Okay. I love it.

[Connor]

I was thinking about, you know, this podcast last night, and I was trying to figure out, like, where. Where do I as the CMO of Ridge, like, want my team or where do I want to be spending my time and where do I want to fit into, you know, the adoption curve of. Of AI tools? Because you're very. You're striking while the iron's hot, right? You guys are. You're spending a lot of time building new infrastructure. I can't, like, I think my approach, I was thinking about it yesterday was like, I think I deliberately want to be a slower adopter on this. Like, I do think there's value in being on the forefront, and I'm curious if you have any pushback on this. There's a ton of value in being on the forefront. When you guys have the, like, homegrown CRO high leverage system built out, I'm going to be extremely jealous. But I think otherwise I'm going to, like, kind of stay. I think we're going to try to stay in our pocket a little bit, see how things trend. You know, Claude skills were announced in, like, November or something. Like, we're not even like a full financial quarter into having Claude skills. So I'm like, before I go like, super deep on any given infrastructure or any given, like, school of thinking for how AI should be adopted at Ridge, I think we're gonna. I think we're gonna like, remain in a holding pattern for a bit. Now the only thing that I think is an exception to that is, like, I. I've talked about it on the podcast, we use Claude now for like a. A copywriting workflow for how we go from executive product summary to marketing summary to, like, the theme of the marketing campaign to copy. And then I can typically, quickly, typically be pretty quickly turned into like, E. Com briefs, but that is all done with, like, project knowledge documents within the Claude dashboard. And I would like to spend a little Bit more time revamping that system with like the, the, the, the more later tools like the skills and things like that, more formal development and try to get buy in across the org there. So I'm just trying to figure out and I'm curious if like again you have any pushback or feedback from me in terms of like. And this goes back to my original question, what brands do you think should be adopting this and how should they be thinking about it? Do you think they should be approaching it like you, is it a spectrum? Would you. Do you think people should be starting at any given place?

[Cody]

I think it should be a spectrum. I think it's like if you want to and whatever. Like, you know, like I know Mike Beckham is vibe coding stuff. I know Matt is doing like crazy stuff, but Matt's also like, you know, an ex developer, right? Like I think it depends on your skill set and stuff. Like again, just to be super, super clear, like we're also a physical product company, right? Like this is like I don't, I don't have delusions of being a tech company or, or, or whatever, you know, like, don't lose sight of the main goal and I'm not. Right. I just like don't talk about product dev on Twitter and stuff like that. I don't spend that much time on it because it's not my skill set. But like I'm very, very, very focused on, on, on product, right? On retail expansion, on, you know, creative, on influencer. Like stuff that like AI, you know, maybe can touch the, the behind the scenes. But you're not actually doing so like that. Yeah, like, like if we were a tech company, you were saying, I'm not going to really think about AI. I'd be like, that's, that's stupid. But we're. You're a physical product company, you know, so I think that makes sense. I, I think everyone's got to kind of pick their own and yeah, like I don't know that I expect everyone to have the skills and knowledge and should try to build these things. I think this is where it's kind of the whole God, I'm gonna say Taylor Holiday was right three times Sean thing, right? But it's like, right, it's the whole tech companies coming together, you know, it's like that's probably. It is like can you evaluate your, your tech stack and your agencies and like are there people that you can bring in to kind of build tools for you or you know, use tools and do it? That's, that's probably it. And just like, I don't know, maybe I think somebody in your org should be paying attention like and, and, and trying to build stuff and try to push the limits and do it in a way where it's not distracting everybody else. And like kind of what I had been doing the last few months is like somebody should just be doing outside pushing limits. And if you then are like, okay, this is actually valuable, like let's get a strategy to do it.

[Connor]

Yeah, 100%.

[Taylor]

If you're still making budget decisions based on platform dashboards, you're not optimizing your gambling this episode is brought to you by Pression AI, the measurement and forecasting platform built for scaling omnichannel brands. Because here' the problem. Attribution tells a story that's convenient, but not always the story that's true. Your best campaigns, they don't just drive direct conversions, they create ripple effects, they lift branded search, they drive organic traffic, direct traffic and ultimately they improve performance across your entire funnel. Pression is built to capture the full impact of your marketing, including delayed conversions, cross channel effects, and the long tail performance that keeps working even after you've stopped spending. Repression really changes. The game is forecasting because the real question isn't what worked last month. The real questions are what happens if we scale this campaign by 10%? Are we already saturated or is there still an efficiency pocket we're missing? If we cut spend, will performance hold or will it collapse? And are we balancing top of funnel and bottom of funnel the right way or starving one side of the system with prescient? You can forecast at the campaign level, model different spend scenarios and understand the real relationship between spend and outcomes. So you stop wasting budget on outdated assumptions. Because a dollar in December is not the same as a dollar in August. So instead of defending your numbers in a budget meeting, you can walk in with a plan and if you're ready to stop guessing and start forecasting, visit prussianai.com Operators pressure AI like GPS for performance marketers.

[Connor]

Okay, because I wanted to hit that point too. Sean Software enabled agencies.

[Taylor]

Yeah. Wait, what's the end?

[Connor]

I have no idea. Anyway, the idea being that agencies are primed to deliver outcomes by adopting AI and have super high leverage employees. Because like and he Taylor actually says this months ago when he was reacting to the POD was like what they are doing is they spend all their time building these systems like they are offering media buying as a service. So figuring out how to leverage Manus is extremely valuable because then they're going to do it for hundreds of clients every year. And that's the same with like how, how are they uniting it with their forecast, all this, everything they do is so redundant. Like that is their thing. 40% of revenue is going to the people that you pay every day. So if you can get higher leverage out of that, like that that needs to be their core competency. And D2C brands are just different. Like we are in a. We are like frankly almost better positioned to be just leveraging those agencies or, or waiting for the best practices and the best in class tools to come out so that we can get that leverage ourselves rather than trying to gain any sort of edge implementing it on our own. So I just think that's, that's an important point because I think everybody gets a little like riled up with like the new, the new meta of like AI tools. And I should be spending all my time in cowork when in reality brands times are still best spent getting better at products, getting better at digital marketing and not necessarily like in investing and not even investing but just experimenting in like the infrastructure AI layer. It's just like a very slippery slope. I think that's how I've been feeling currently and that's, I'm like, oh yeah, I think we actually need to take a step back, sit in our pocket a little bit, wait for the, the best in class practices to be developed and then we'll kind of make a move there.

[Cody]

I think there's a case for like everyone to be on cowork because even just like that's just your basic like computer work. Like hey, I automate this just quick data analysis stuff like that. So like I do think there's a case there for like everyone to just like use a tool like that and use cloud and like just automate busy work. 100% agrees. I think agencies and softwares are just going to become one and, and you know, leveraging tools but you don't have to build it is probably what most brands should be doing. And then I, the one difference, like you know, the DTC 3.0, maybe this becomes DTC 4.0. Like that is a potential unique competitive advantage. And like I do think you could run a $50 million business with three employees or you know, even larger ones. So like that's where like if that is your core competency, your competitive advantage, like obviously you still have to make great product, great ads, but like if you can kind of use these things like that actually makes sense for a brand like that. But for a brand Like Ridge, it doesn't make sense.

[Connor]

Totally. Yeah. No, and that's a great point too. If I was starting a brand today, maybe it's something that I'd be investing in earlier because like the other way to think about is, you know, whatever we have single digit percentage points of total revenue going to headcount and it's like, let's just say we could be so much leaner. Like there's not that much fat on the bone there. But like if you can avoid, if you can avoid any sort of bloat and just remain as efficient as possible. And D2C, you know, DTC 3.1 or 3.2 or whatever is actually like yeah, you're running a business with 1%, 2% of revenue going to headcount is seems more and more possible.

[Cody]

I mean that's what I would do. I would, I'd run a subscription brand. I'd spend 60% of revenue on marketing and have like two people.

[Connor]

Yeah, totally.

[Taylor]

I mean it changes that person like that skill set because that's what my head immediately goes to. The like you know, $10 million brand, 5 million dollar brand that maybe has like one really talented like generalist growth manager. Like if you can hire someone that can do like the 20% of things with the, with these like these tool building and like that's how you like I, I go back to the episode with Dan. Obviously they're different, right? They're like probably close to a nine figure brand now. But I think like that that person you hire has a very different skill set that like growth manager is a very different skill set today than maybe they had. You know, maybe you hired a media buyer five years ago that could also do a lot of the other things but it was still required on like him collaborating or her collaborating with other people. Like now I would be hiring someone that is doing like, is really good at what Cody's doing and he can actually like create efficiencies and, and oversee all these channels without having to hire you know, a media buyer or retention marketer or a designer. It's like yeah, maybe now you can hire half those people and the other half he or she is, is building tooling for it. And I think those like five ten million dollar brands where it's like founder, growth marketing manager, like you know, kind of like a single person overseeing the different functions of the business. I think those are the brands that if they know what to build and they know like what the needle, like what, which needles to move that I think those are the brands that could really leverage stuff like Cody's building.

[Cody]

And by the way, case in point of work, I think there's a lot of drop shippers that are doing that. All of these fraud sites that have come up that everyone's getting hit with, part of the reason because they're just spinning up pages.

[Connor]

That's DDC 4.0.

[Cody]

Yeah. No inventory. What do they call it? No ship. I think they call it no ship. It's like no guys, it's called fraud. Yeah, but like, no but like seriously like they are, they are essentially arbitraging this ability to spin up this stuff really quickly, run meta ads to it. And you know, obviously your, your, your CAC doesn't have to be that great if you have no, no personnel, but also no fulfillment costs. But like I think it just shows like how quickly like you can actually, you know, build stuff and, and, and, and have it working. So I, I, I agree with, for us I am most likely going to hire people for this function to build this stuff. Like I think for us it'll be as big of a thing that as we have a few pilots and we see that they're really effective. It's like, okay, let's go do this for everybody. Let's build a recruitment pipeline that you know, now our, our senior director people use and now we don't need to hire more for her and you know, let's do this for, for E Comm. And so we don't need to hire, you know, as many. So like I do think we will. But I agree with you, most brand shouldn't be.

[Connor]

Let me ask one question for you guys. Do you guys have dedicated project management teams at each of your brands? Like is it like a centralized resource? I know we've talked about this in the past. Hexclad's got it. My question is going to be. Because that's what it sounds more like. I also, sometimes I feel like I hear people being, I saw this on Twitter, like oh well, just everybody should be vibe coding and to some degree everybody could be vibe coding. But the idea that everybody's going to be capable or have the skills or understand when and where it makes sense feels very low. And that like AI enablement within an organization ends up looking more like project management where it's a centralized skill and resource. These people are going to have a more familiarity with the best way to build things and the best tools to integrate and then they can kind of move around the org. So I like love the idea of the dedicated Hires?

[Cody]

Oh yeah, yeah. So, so it's, it's essentially I'm doing a little bit just, just, just for fun on like projects I care about. It's our director, data and AI. And then we're gonna probably bring on like somebody under him. Maybe it's an engineer, like a overseas engineer, or maybe it's just like a young kid who's just really into vibe coding. But it's like, yeah, like they, they kind of do the FDE forward deployed model and they just meet with people and it's like, what are, what are you doing? What can I build for you to make your job easier? Like, I do think at a, at a large enough organization that makes sense and that's worthwhile. I mean if you think about it, that's a, that's a hundred K higher. Probably maybe 150 if you get somebody like senior, like the amount of value that you can create, I mean is, is pro. It's actually probably a no brainer if, if you wanted to go in that direction.

[Taylor]

I talked to this, this COO like a year ago and the way that he described his role is that he had half his role which was like his always on basic. Not basic, but his blocking and tackling as the coo. And then the other half of his role is he was kind of like parachute into different parts of their org and like figure out which processes and systems weren't working that well and then like build it up to make it better. That is kind of what I think is now maybe hopefully getting like that part of his role is hopefully getting infinitely easier by having like this, I don't know what you call it, like AI engineer or vibe code lead. That's like this guy on this guy's hip and he can be listening in and like understanding what the problems are and where the system's broken. And now instead of the COO having to set up some like manual thing, it's like no, him and the, this, this kind of vibe coder, you know, AI engineer are building a solution which I think is really, really exciting. I, I think it makes the solving of things a lot, a lot easier than it probably was a year ago or even, even three months ago.

[Connor]

100%. So Cody, you'll have to report back. We need more pod fodder. So all the AI tools, hacks and guidance you can give, I think the better.

[Cody]

All right, I've been like also bone chipping, so I'll let you guys know how my looks maxing is going.

[Taylor]

Bone chipping. Cody, Cody's going. So I. Cody, I gotta say, joke. I'm inspired by the founder mode you're putting into all the vibe coding right now. It's very, it's very inspiring and it's making me want to sit down and think, like, not only in Hexclad, but also in my personal life. Like, like, I'm training for a big trail run right now. I'm like, how can I build a tool to like, track my progress over time where it just, like, integrates with my, like, Garmin and it can tell me if my, like, pace is getting better like my 5k time. Like, that's. I'm like thinking, like, how can I do this in a bunch of areas? So I'm excited to sit down for a few hours on the weekend and just build stuff.

[Cody]

I find it fun. It's like, it's like a video game. Like, my, my wife the other day was like, oh, should we watch tv? And I'm like, I was about to. You ever, like, when you're playing video games, like, tell them. I'm like, no, no, Mom, I'm about to get to a level.

[Taylor]

Yeah. Yeah.

[Cody]

I was like, how I felt and I was like, no, I'm about to get a break. Like, it just is like a fun thing, but it's also like, you're hopefully building something valuable. But yeah, I don't know. I, I, you know, like, the whole, like, tick tock shop trend, like, I was late to that and I really regret it and kind of kicked myself for it. And I just don't want to be late to trends. And, and it might mean that I'm jumping on some fads that weren't. And like, I kind of decided I would rather that. Right. I'd rather risk kind of like jumping on some stuff that is a little bit of a waste of time versus missing out on some. I just don't want to do what I did there and have it be a lot of lost value creation.

[Connor]

Totally. Yeah. You don't, you don't get any bonus points for precision. Right. So it's like if you're just. If you're directionally correct, it could pay off massively.

[Cody]

Yeah. And then the other thing is, like, the CEO's job get on my soapbox a little bit. Like, the CEO's job is supposed to be like, on trends and, and future and like, thinking ahead further than anybody else. Like, when I'm, when I'm too in the business, people give me for like, oh, that's not the best user and then like I've removed myself from stuff and actually have time to vibe code if I want. And then it's like you shouldn't be doing that with your time. It's like maybe, maybe the reason I can actually do that and I'm doing it more at night and stuff is because like I'm not day to day. I'm not thinking about what our ad account looks like today. You know, you're, you're trying to be like hey, what's totally what is, what is the future for us and like how are we going to re reshape our workforce? Like that is a big thing that we are thinking about right now because we just got to five years and it's like the team that got us here is probably not going to be the exact one. So like there are restructures that have to happen and things like that. And you know, I just think, I think, yeah, that that's why I think the CEO should do it. I think also like the leverage, right. So like naval talks about you know, code being one of the leverage areas but like you can now take the CEOs or head of growths or whoever's like knowledge and actually be able to pass it down through the organization so much more.

[Connor]

Right.

[Cody]

You know, like with so much leverage that like I can build something once in a day and then have my 50 person team use us, whatever. And, and I know that sounds like a little arrogant of like hey, I have this knowledge that everyone should be doing but like hopefully your senior team does have that.

[Connor]

You're just trying to scale yourself, bro.

[Cody]

Kind of.

[Taylor]

Yeah.

[Connor]

Look, I think that's a great point to end it on. It's a good ep.

[Cody]

We'll have to.

[Connor]

This is going to be the topic of 2026, no doubt. So we'll have to circle back in

[Cody]

the future this and creative testing. Yeah, yeah, yeah.

[Connor]

Awesome. Thank you for listening to another episode of Marketing Operators. Thank you to our sponsors, Motion Rich Panel, Prescient After Sell and Revo. As always, please like subscribe, share with your friends and family. We will see you next week.