Sean (2:08)

What's up Operators? So we have a great episode today. We have Craig, he ran AI Crocs. We have Matt, he is the CIO at Bissell. So we have two legacy company experiences talking about how to use AI practically your day to day in your org right now, how to get buy in where AI is, where it isn't, what to focus on the worst demos they've ever had. We're going to talk about all that stuff. How are you guys? Welcome. Welcome to the operators network.

Matt (2:37)

Doing great. Thanks for having me.

Sean (2:39)

Cool.

Craig (2:39)

Likewise, man. Good to see you.

Sean (2:40)

Yeah. And also, if you like more of this conversation, Craig runs a weekly free off the record conversation where he brings in cool CEOs to talk. He'll plug it at the end, but he doesn't record them. So you have to be there in person. You have to be on zoom with your camera on. So if that's kind of your speed, he's there for you.

Craig (2:57)

The, the challenge is there's one other piece which is the only way to get into the group is to be invited by a founder who's already in it. So if you want, ask your friends who participates to see if you can get a. A plus one.

Sean (3:06)

Dude, this is like the Skull and Bone society, but for learning nerdy e commerce stuff. I'm. I'm here for it, man. Okay, so let's talk through. Today's episode is what can brands do today with AI? Where is AI and how to get buy in across organizations? Both these guys have huge company experience. They'll talk about failed demos, they'll talk about how you can't get everyone to use it at the same level. But how do you find the winners? How do you find the winning programs? How do you actually get buy in? So if you're ready for it, let's dive in. Craig, what's the worst AI implementation you've ever seen? Oh, man.

Craig (3:40)

Well, I think it's important to kind of, if you're a game, take a step back and just kind of outline the role. Right. So when I led AI at Crocs, the mandate was to do two things. It was to go network externally, talk to other companies, Mondelez Moderna, understand the ways they were leaning in. It was to talk to venture capital firms, to work closely with the teams at A16Z and Sequoia to understand what they were investing in. And then it was to talk to AI startups. Right. I probably spoke, no joke, to 200 different sort of AI startups in the landscape over the two years, leading AI for the company. And then the other part was to network internally. Both brands, Crocs and Hey dude, all departments to find problems. And if AI could help to solve one of those pain points. We had budget, project management, resources, executive visibility to kind of spin up pilots, Right? And we must have launched between 40 and 50 pilots over my time there. And I think what I learned is why that famous statistic that just came out like that, 95% of AI pilots like fail. And there's a number of reasons why that's the case, but I live that sort of like full stop.

Sean (4:40)

But Matt, do you have a similar experience, maybe explain your role as it sits day to day right now, the

Matt (4:45)

chief investment officer role, I'm in charge of both the finance department. So the CFO reports to me, it reports to me and then all of our various subsidiaries or acquisitions that we've made report to me. So I've been spending a lot of time in AI and helping figure out where we want to make our capital allocations or investments and which tools. When I hear you ask what's been the biggest disaster my finance brain goes to, where have I made this huge investment that didn't pay off? And I think what's unique about, you know, the AI wave here, you know, especially relative to traditional SaaS, is that we haven't seen those big flops because the business model or the go to market model is, has been, you know, either let's do a POC or we're going to do outcome based pricing. And so you really don't have to make those big, those big kind of lock in investment decisions. And so that's something I've really frankly enjoyed about the latest kind of technology boom. We've had sure plenty of failed demos or demos that didn't live up to the hype of what we wanted to do, but there was never some big investment around it where it was like this one way door that you couldn't come back from.

Sean (5:49)

Yeah, and I guess that is the most exciting thing about right now is that if you fail an ERP implementation, it's 6 months or 8 months and $250,000 in burnt capital where because ChatGPT rolled out on a per user $10 a month model, like they kind of are the pricing ceiling for everybody like lovable is like okay, it's $30 a month. So everyone's like going this like prosumer model, like going for individual think is great. Craig, you were going to add, I

Craig (6:15)

would say two things and building on exactly what, what you both just said, Matt, like with a, a big corporate lens, obviously kind of where some of the big investments are happening in AI right now, you know, executives immediately go to oh, we got to work with McKinsey on this big consulting project or something like that. That is where a lot of money I feel is Being wasted, having been on the inside of some of these, you know, consulting engagements, it is not kind of worth it, to be honest. And all of the value, Matt, comes from what you just outlined, which is like, look for 10, $20 per user per month. That's where the value lies. Just get good at everyday AI tools and you've both figured out AI, but also saved yourself. And the reality is, and then I'll be done. Like there are so many pilots that you could spin up with one off vendors that three weeks after you launch it, that's just a feature in ChatGPT or a feature in Gemini. What that startup was doing is now just possible in this everyday AI tool. So I think there are areas where you could waste money, but for the most part, you're absolutely right. Just spend 20 bucks a month and you're good.

Matt (7:15)

Yeah. To your point, Craig, we, I actually have a, we have a no kind of consulting policy on this. We have that. We have not used anyone like, you know, the company you mentioned. I won't throw anyone under the bus, but, you know, we don't want to do the big consultant projects and kind of assessments. We have been burned on those and other types of digital investments. And so yeah, we, we intentionally avoided those, those big projects.

Craig (7:37)

I should flag. We didn't work with them. I didn't name the one. We did.

Sean (7:40)

Yeah, the best thing you do is just get the people's hands. So at outreach right now, we also haven't taken any demos. Right. Like, we just try to pay for all of any SA, any AI SaaS tool that people want, we will just pay for because it is so cheap. We're using Claude, ChatGPT and Gemini and we give access to all three of those to everyone at the team. And I would say the amount of people who use them every day is like 30%. The amount of people who use all of those every day is like 10%. So that's really what I want to address here today is how do you get that number higher? Because specifically the latest version of Claude, it blows my mind how good it is at business functions. Like, I'm doing my, you know, monthly reporting, I'm just dumping raw data in there and everyone's like, oh, it hallucinates. The latest version of Claude does not hallucinate. It puts exactly what I want out. It seems near perfect.

Postscript Representative (8:33)

Every SaaS company says they are AI powered, but very few can explain what it actually does for the revenue of my brand. This is why postscript's approach stood out to us. They don't just build AI for demos or people buzzwords. They built it to drive real incremental revenue. Postscripts AI called Shopper. It shows up inside of SMS at moments with real buyer intent when shoppers are likely asking questions, hesitating, maybe even about to drop off. Shopper can answer product questions instantly, answer questions about fit, availability, recommendations, order issues, the kinds of stuff that people usually bounce for. This means more conversions, higher aov, less lost demand. So you are driving more revenue and doing it more efficiently. Check out Shopper from postscript. We use it at Pela, which is why I am telling you to check it out.

Sean (9:19)

So I want my team to be blown away by what's capable, but when I have 10 or 20% of people buying into it every day, that's where I feel like I failed. So do you guys, how do you guys address that? You guys have huge organizations. How have you guys changed that and actually got the adoption?

Craig (9:35)

Yeah, I think I'll answer it in two parts and probably pause on on the second, but I, I think you might be, I don't want to say looking at it wrong, but I think goal number one is drive adoption of the tool, get everybody comfortable, excited about leveraging it and seeing the power. Right? Everybody has to have your aha moment and I can talk about ways to drive that. I don't think the goal, at least how I approach it, is to have everybody using all the tools all the time. I think you want to get people expert at one. The one that I've decided is, hey, ChatGPT has 800 million users per, you know, per week. It's the one that everybody knows in their personal life. Let's bring that into our normal business process so people are comfortable using it at very different strategic, cross functional points to see the power. And then once everybody's really comfortable with that, they have a foundation of how to use everyday AI tools. And then you start to layer on Gemini for areas where ChatGPT isn't strongest, Claude for certain numbers, etc. I think the idea of just hey, let's turn it on and go use it. Oftentimes people might not be using it in the right way. And then they'll say like, oh, what, what's all the fuss about? Right? Like I, the analogy that I have is my son Coley is three and when we did potty training, it wasn't just like, hey, here's the toilet man. Go go at it. Like there's a lot of like hand holding that, that has to go through the process. So anyways, first things first, I have a number of ways that I would suggest rolling it out based upon best practices I've, I've seen. But I don't think it's, hey, just go use these. I think it's actually being really strategic and intentional around enablement use cases, celebrating. And then the tools are so powerful you can't help but roll them out at every part of, you know, of your role.

Sean (11:12)

Similar experience or how are you getting by it?

Matt (11:15)

Similar experience. I mean we, we've also made the tools widely available across the organization and I wouldn't say we're having a problem from, from use. There's a lot of daily use but there's, there's a bit of like an imagination gap or you know, really trying to help the organization understand like the art of the feasible and what can really be done with these tools. You know, we actually, we required training for each of our employees to learn how to use the tools, put all the employees through training and then we actually survey them to say, you know, where do you think AI is going to be the most impactful on your role, on your function? To try to both get, get the organization to think about it, but also surface ideas that we could then go take and implement. And you could see pretty consistently in their responses that you know, everyone was thinking about it mostly as like a personal productivity hack as opposed to like very strategic. How am I going to embed this into my workflow and change the way I work in a really impactful way? And so, you know, we're approaching it both from a top down perspective and a bottoms up. The bottoms up is giving everyone access, letting them experiment. But then we're also picking, you know, our top two or three most impactful areas where we either know AI can provide value right now or it's highly strategic to the business and making them highly visible to the organization so that they can see real dollar impact, real change in how we work. You know, the first one we tackled was our customer service function and implementing. We used Sierra. We are one of the early customers there. And so you know, that was something where we, we really celebrated as Craig talked about and made it very visible to the org in terms of what, you know, what can happen when you, you do this the right way.

Sean (12:47)

Yeah, you hearing you guys talk, it really reminds me of like the meme where it's a iceberg and like the tip of the, like the top of the iceberg, the one you can see, it's like ChatGPT and then like, maybe it's Gemini, then maybe it's Nano Banana. And then like, the further you go, there's just like more like, harder to use and harder to enact tools. Like, there's going to be NNA down, there's going to be Higgs Field down there. Right. And it's like, we just want people to start using some of the iceberg before you actually start getting deeper in the stuff. So. So, Craig, you said you got tips to make people excited to learn about AI and actually practically put it across their organization. How are you doing that?

Craig (13:23)

So I'm close with a friend of mine, Bryce, who led AI at Moderna. Harvard Business Review literally pointed to him as the best AI rollout across the country and what he did at Moderna. And he shared in that review that the number one thing that he did that drove adoption, it's so simple, but was just having people transform their team's profile photo into a Pixar character. Right. Like just a couple of words. And maybe it wasn't exactly Pixar, but it's the first thing of just like, oh, my God, like, this took two seconds and now it's this fun, engaging way. Right? So first things first is just kind of getting everybody familiar kind of with. With the tools in that way. And then in terms of other. Just general best practices, which. Which ties back to something that Matt said about bottoms up and tops down. But I'll go there later. Right. I think it's a lot of what you do at Ridge, Sean, which is, hey, public Slack Channel, where people are sharing wins and announcements and things and ways that they're using the tool. OpenAI, when they do enterprise rollouts, they often say, like, look, coming out of this, we've spun up 50 different GPT that are unique to this company's way of working. And so when you have somebody who's celebrating like, oh, my God, I use this to transform the way that I write briefs. And now when I hand things off to my design partner, there's no more spin. Right. Like, so celebrating those kind of use cases, I think similarly, I hate committees. I hate them so much. But this is actually an instance in which, hey, if you have an AI steering committee, it doesn't need to be just like, oh, I used AI for this or used AI for that. I think that's the Slack Channel. I think you can be really pointed in celebrating the power users of the tool who are maybe junior and giving them the floor in front of the organization to share what they're doing and kind of celebrate them that way. My friend Will, who's the CEO at Harry's, talks about the ways that they have power users just present to the board about the transformation that's occurring. And I think this notion of celebrating and incentivizing and rewarding that adoption is really important. I've got a number of others, but I'll stop there. Those are just a couple of really basic ones.

Sean (15:23)

Yeah. Matt said the word creativity gap, right? Or imagination gap. I think that's so true because we're so used to software doing the same thing forever. But this software is updating in real time all the time, right? So even if somebody did a demo six months ago and thought it was cool, like, what's capable now is just so much stronger. I bring up Claude getting really good at data and not hallucinating anymore. And it's like that happened in six months. So in six months it ended up being not a good data tool to the best one on the market. So I think the biggest tip for everybody is you just have to do live demos in front of everybody. So, like, get everyone on a call, make them have cameras on, spend an hour and build a game in real time, or build AI videos in real time and, like, go through all of the actual workflow. And it's slow and it's boring and it's messy. But watching somebody be like, okay, here's an image we have. We're going to make a whole video story about it. And doing that in real time in Higsfield blows people's mind. So I think that's like the biggest tip. Matt, any response to that?

Matt (16:23)

We're doing the exact same thing. You know, we've got a. A teams channel where everyone's just posting things that they're building, you know, that the organization can interact with each other. I've got questions around, you know, how to do X and cloud code or, you know, something else. So we want to foster a lot of organic engagement. We do a monthly town hall that, like you said, the power. We don't have any of the executives in it. The power. The power users, you know, present to other people that are just interested in AI, want to see what, you know, what we're building, what we're doing within the business. And like you said, live demos, videos. You know, we're constantly posting, you know, YouTube's from vibe marketers around. You know, look at this workflow they built. You know, I've been buying, you know, prompt libraries and just flooding them into the, you know, our shared folders. Just to get people to experiment. So a lot of this actual, you know, experience of using it, I haven't

Sean (17:10)

even thought about buying prompt libraries. That's hilarious. So it's like, yeah, I'm, I'm sure that's like a whole market. And like, I just, like, there's so much undiscovered land here. What's up, right?

Matt (17:20)

And it blows my mind. I mean, when you watch one of these YouTube videos and you see like, the, the complexity of these prompts and like, what it really takes to get the, the output or the output that you want, you know, I, I think that's one thing where a lot of people don't understand what a, what a good prompt looks like. And if you can just take, take that variability out of it. Because, you know, if they do use the tool and they don't write a prompt, they just write it off as though this doesn't work. You know, a lot of the time it's because they're not prompting it, right? They're not using. They're feeding it with the right context, the right data. And so trying to take as much of the kind of user error out of it as possible so that they have a better experience is something we're focusing on.

Craig (17:53)

I think that is the most important point. I think a lot of people are threatened by the technology and they want to point to it as not working or not ready for prime time. And when you get crummy outputs, people are like, oh, I don't get what the fuss is about. But you know that the reality is, is to Matt's point about prompting, like, just use a simple prompt optimizer. ChatGPT makes one available. Just say, hey, this is what I'm trying to do. This is what I want to say. And I'm like, oh, if you're trying to do that, just tell me this and then that's it. Right? But I think a lot of people don't want this to necessarily succeed in a weird way, for whatever reason. So, you know, pointing to high quality outputs, I think is critical. I think you've also both talked about sharing things publicly, and I think one thing that's beneficial about that is you're able to see the ways that other people use and interact with the tool, and that can be a light bulb moment for you in your own interaction. But Matt, you talked about tops down and bottoms up. And I just want to share my, my buddy Ben, who's a phenomenal leader, I think he's led one of the best rollouts across the country. Same with my, my friend Danny at Tabletics. And they've done something similar where they've trained a small group of individuals, you know, really entrepreneurial, curious, empathetic people who've been at the company for a while and have high trust and then they've sent them out across the organization to go find problems and pain points that AI could help solve across different departments. But most critically what they've done is they've then surfaced that back up to leadership to say look for you as a marketer. Here are 20 things within your department where we feel AI or automation can help make an impact today. But most critically, you need to tell us the three that you want us to work on. Right? And so this idea of bottoms up but then top down mandate and requirement is so important because the labs and others will tell you like this stuff lives and dies with the executive champion. If they're not all about it and making it happen, then it's not going to necessarily go anywhere or people aren't going to celebrate it. So I do think that that bottoms up tops down mix is critical in our org.

Sean (19:49)

10% of people, super users, like 10% of people will do like they won't do any work that doesn't involve AI. So if I ask them to do something, the first thing they're doing is going to AI and figuring out what I mean and then optimizing the workflow from there. And then 25% of users on top of that are, you know, they're going to use it in something but it's really getting that like the, the, the majority of the block, 60% of people to break it into their workflow. And a lot of it is just like look, these people are professionals. They've been working for 10, 20 years. Like now they have to do this, this other thing when they know how to do it the, like the way they've always done it. And getting that to break I think is really hard. So Matt, you brought up CX as like the first thing you guys do. But the next tip a practical for the audience. What's the 8020? Where can people get AI value right now? Max, how about your experience? And we can kind of expand out from there because it's an obvious first one. What's up operators? Welcome to the Rich Panel ad read. Rich Panel has been a sponsor for over 12 months. I've been a paying customer for over 12 months and guess what, I just renewed the pay again for another year. We have cut our SaaS bill in half. And automation dropped our cost per ticket by 70%. Our CSAT has also improved from 88% which is still really good, to 96%. Best in class. All powered by Rich Panel. I told them last year, hey, you guys need to do the same thing with returns. And now Rich Panel has a returns portal. It's built to cut down your tickets and convert more refunds into exchanges. They do the heavy lifting, data import, self service, retention flows, team training, all of it. And it'll be live in two weeks. If you want to save 30% guaranteed on help desk and now returns, book a demo.

Matt (21:27)

You know, coding and CX is where everyone's focused and you know, we're, we don't have as much of a coding need and so that's why we picked cx. And you know, it's, it's been a success for us. We've been really happy with the output and now it's about scaling it and you know, the fact that we have this tool in place enables us to do frankly just more and better customer experience than we, than we did before. You know, for example, we're going to launch in Poland this year and you know it, prior to having this AI solution we might have had, you know, the only customer service solution for our Poland customers is to call our, you know, an English speaking, you know, agent. Right. And we couldn't, you know, we can't translate languages. And so this is enabling us to deliver a much better experience in each of the countries for each of the products that we serve. And so we'll continue to grow that, that product, you know, the obvious one and where I'm sure a lot of your listeners are focusing on is, is just our, our marketing, you know, content and just increasing the velocity and variability of our content. You know, that's always a bottleneck I think for any consumer product company. And so being able to do more of that is going to be a huge, a huge win for us. And then long term, I think what I'm most excited about is how it's going to change our product development cycles. So, so both using the, using the models to both collect all the insights that we have internally that we built over, you know, 150 years of being in operation to scraping all the external points in terms of reviews and customer feedback, feeding that into our product development cycle and then also just doing the actual, you know, automation of, of things like Cat. And I haven't seen that necessarily get to the point where it's ready yet, but I think that's going to be a big unlock is just the, in physical product development. It's something that Jeff Bezos is working on with his new startup and I think that's going to be really transformative in the, in the physical product space.

Sean (23:16)

Yeah. So three levels there. The first with Ridge Ridge is up 50% year over year. And with that, tickets naturally go up, but human answered tickets is probably at like a five year low. Right. So like some tickets left to go to humans, but as the business has grown, the number of tickets to actually reach a human operator is, is shrinking over time. So that means our CX team is, you know, 10 agents where it used to be, 25 agents. Right. So that's like a real practical use case right now. And everyone, I think everyone agrees that's going to be solved. It's like, it's like they're, they're doing voice now, but like we, we don't do voice, we just do written text and it's like, yeah, this is what ChatGPT was literally designed for. Craig, what were the trends you're seeing around CX before we move to marketing?

Craig (24:03)

Well, I think your point on, on voice and cx, I mean this stuff is getting so good that it is both inevitable and non. It's not controversial to say that text, audio, video will be indistinguishable from human produced content or voice like relatively quickly. That's, that's probably a 2026 thing. And so yeah, I mean, without a doubt, like you know, CX and all these other areas, like, you know, content, creative, like we just now have a new tool at our disposal to help us with the output of that work. Right. I often say this is just like Excel. Like it's just another tool to help people do work. I think there's a lot in the SDR and outbound sales automations and things like that where you hear teams. Just like you had said, hey, instead of 25 folks, we've got 10 folks doing it. I think sales and outbound sales is also that world right now too.

Sean (24:55)

And Ethiopia isn't talking to someone in person but like setting up calls, like answering questions, everything. It's like, dude, the, the ball isn't even moving there. The ball's there. It's just waiting for someone to pick it up.

Matt (25:07)

I think just assume that the, the technology will be good enough even if it's not good enough. Now just this, your, your default assumption should be it will be and it will be faster than you think. And so you have to start experimenting now. So that when it is there, you've got the kind of the pipes and the foundation in place to take advantage of it. But because of that process, I mean, Craig, you talked about needing executive sponsorship. I think that's been critical for us because there has been like, when we've really tried to embed it into material workflows, there has been this J curve to the investment where it is a worse experience for your associates. Like you said, Sean, they're used to doing things a certain way. They're probably pretty efficient at it. You're forcing them to use this new tool. It might not be working exactly how they want it. And so there is that, you know, learning curve where it actually is probably less productive than their current process, but then quickly will accelerate through to a much better experience. But you really need the executive sponsorship to clear the way and kind of maintain the kind of the conviction that this is going to work.

Craig (26:05)

Can I jump in with a great use case within that? Sorry, Sean. I get so excited to build on his point. So I'll stop. You go and then.

Sean (26:11)

No, no, no, dude, it's a convo, man. Craig, build on it. What's up?

Craig (26:15)

You know, we talk about kind of like slowing down and problematic in that way. I would try to be crazy strategic around some of like the big use cases where a simple solution could, could solve. One thing that I immediately go to is there's so much process loss when briefs are potentially handed off to another cross functional team. Right. And what gems or GPTs are, they're just effectively like books around what best in class output looks like. And so anytime you've got a brief that is handed off or something like this, like just run it through a book about what a best in class brief looks like and then it'll spit back out. The GPT will say like, hey, you know, head of category at Ridge. No, this brief is missing X, Y and Z for your partner on the design team. And when you have a use case like that where the two teams are able to work together and define what best in class output looks like, to then have that middle layer, that GPT, that's then a trigger for them to be like, oh my God. There are all these other use cases where something like this is relevant. So I'd be strategic there. I had a situation at Crocs where we built that for somebody and she got like legitimately emotional. She was just like, oh my God. My team spends hundreds of hours each quarter on things that never see the light of day. Not because the Work isn't good enough, but because the brief was off and all that we deliver didn't meet the goals. Right. So when you can be like surgical and identifying these high impact, low effort opportunities and then people just see how easy this thing is both to use and create and it starts the snowball.

Matt (27:43)

I think that's an underestimated value of AI, especially for larger organizations is, you know, everyone wants to talk about all the alpha that AI can create, but there's a lot of value in just the, like reducing the beta across large organizations in terms of just the, the variability of outputs that you get for the same process. Right. So, you know, I might have one team that is amazing at writing briefs and always comes out with, you know, great content, the other one, you know, is inconsistent and being able to give them those tools, those best practices packaged up, you know, in a, in a cloud skill and just know that I'm always going to be able to get, you know, the right hooks that I want, the right demos that I want for the new products with much higher hit rate is incredibly valuable as you think about just empowering your organization and enabling you to then move on to doing other stuff.

Sean (28:28)

Consistency of communication and communicating internally. Good, good use cases. I actually wanted to talk about creative next because do you guys remember when Instagram started tagging things made with AI? Like they were like, hey, this is made with AI. And then the CEO of Instagram came out and he's like, actually it's going to be easy for us to tag stuff that is made with people because everything's made with AI now. So he's totally inverted that. And Craig, you said earlier it's like indistinguishable from created by a human. And that's a great 2026 prediction, is that video will be there. Like, you know, right now even the best video models, if you look really, really hard and there's somebody speaking, you can still basically figure out that it's AI. I think that totally changes by Q4 of this year. I think this is where the puck is definitely going, is that creative will almost entirely be AI. Taylor Holiday came on here and he said 20% of spend in most of his ad accounts are fully AI. And he called that a loss because he's like, look, you know, why isn't it 50%? It's going to be 50%.

Postscript Representative (29:30)

Longtime sponsor North Beam is launching incrementality later this quarter. This means that you can now have the trifecta of marketing measurement all in one platform. That is multi touch Attribution media mix modeling and incrementality holdouts all inside of North Beam. You can automate that lift testing end to end, unify results with your MTA and your. Mmm. This is a lot of letters, but if you know, you know. And you can start to cut what doesn't work and you can scale what works and you can do this all with confidence. This is why this is such an incredible ad to North Beam. North Beam's incrementality measures what results marketing is actually generating, not just what they're claiming credit for. As a CEO, that's like music to my ears. Sign up now and you can lock in 50% off unlimited tests for the year.

Sean (30:16)

YouTube Shorts is now reporting that 20 to 40% of all YouTube Shorts are basically just AI. So it's like AI video. But consumers are consuming this even if they don't want to or don't mean to. And I think all of us as brand marketers are going to have to create ads using AI. And if you haven't done so so far, it's like the next obvious thing, infinite iterations on video. So, Matt, are you guys bringing video and static ads that are made with AI into your ad account?

Matt (30:43)

Yeah, we are. I mean, we've been experimenting it now with, with our agency. You know, they're building their own tools to help, you know, we're seeing it more useful in copy and static ads. We're still struggling with video. I think to see, you know, to have it be at the quality that we want it to be at. But again, the assumption is it's going to get there. And so we're going to keep, you know, iterating on it, but we are definitely embedding it into our, into our content creation workflows. And completely agree that in 2026, you're going to see a significant increase in, you know, our own copy, but in our own content. But, you know, across the board, people using AI to produce their content.

Sean (31:18)

And Craig, what are you seeing right now? You must be excited about this.

Craig (31:23)

Yeah, I mean, look, it's a spectrum, right? When we talk about AI, there's. And creative, there's kind of the Coca Cola, like, big commercial, whatever. But then if we go back to this, like, analogy of these tools, just as Excel, right? Like, it is a tool as part of the creative process. You can use AI across all different variations. Excuse me, all elements of the creative process to speed things up. And when it becomes easier for creatives to create more content, what inevitably happens, I would imagine is better Segmentation, right? You're able to hit and have the right creative asset and messaging hit the right consumer because it's so much easier to create more and more output. I think that is an inevitability. I mean the labs have effectively said as much. I love creating AI versions of customers to pressure test messaging against, et cetera. And so that I think phase one is inevitably that I think phase two is almost like an agentic build where it's just studying performance and then creating net new, you know, assets based upon what's working and what's not. I don't think we're there now and I don't know that 26 is the year for that, but I think the segmentation element is where this is. But just to wrap a bow, AI for creative is not just Coca Cola, right? It's not just, oh, that's 100% of the output. It's where and how to plug this technology into the entire creative process.

Sean (32:40)

Here's, here's a good practical use case for everybody. We have a collegiate program. So we have every school and you can get your, you can get a ridge wallet with the school logo on it, the mascot, the team, whatever those campaigns. An individual school level might only spend $50 a day. It's like, look like Alabama has a lot of fans. They don't have that many fans, right? And you know, put that to every school across the whole country. There's a limit on how much we can invest in actually getting that creative, right? Like, am I going to get a jersey for every team? Am I going to get like a fan wearing the jersey? Talking about how you can buy the wallet. It's like, even if I want to just do that digitally with a designer, it takes a really long time. What we do is we have B roll, we have photos of everyone because we have product pages for everyone. And then we can use AI to put in a fan in a jersey, right? So now I actually have a more compelling Alabama ad or Harvard ad or whatever, right. And I can do that for 250 schools now. And it could be two days of work instead of literally, you know, a three week long thing trying to get all of these individual UGC pieces shot and sent back to us. So it's like it just starts unlocking alpha at the edges of your ad account where like you weren't able to do it in the past, know that's where like a practical tip where I think it is useful right now. It is blending some amount of AI. So AI UGC with Actual B roll. So you're not hitting somebody with a 60 second AI only ad, because I just don't think it's there yet. But you're putting in, you know, three to five clips of AI that are all three seconds each.

Matt (34:07)

It's obviously in that same vein. SEO is obvious as well in terms of just creating a lot more blogs, a lot more. Particularly, you know, we've got a thousand SKUs in our catalog. You know, there's a lot of SKUs that just don't get attention from, you know, whether it's, you know, creating content or, you know, blog content or some of those static images. And now you're getting to get to those, that long tail of your product assortment that you just wouldn't touch before.

Craig (34:30)

Absolutely. I think that's it where you as the leader just step in and say, oh, hey, this buy is only a thousand units or 10,000 units. It's not 100,000. That's not as important as some of these other things. Boom. That goes to the AI production cycle like with some of these big companies, right? Like, you know, people don't even have visibility to the size of the buy or they don't even know if what they're working on is going to be part of a massive campaign or not. And so I think that's where leadership comes in. And just from an efficiency standpoint, you're able to segment it. I mean, Sean, that's a perfect example.

Sean (34:58)

Do you guys think this will be an advantage to incumbents because they have more capital, they're more durable, they have, you know, larger infrastructure already of distribution? Or do you think this is just another tool for like the rebels, the small shopify merchant to, to, to get bigger, faster, better, like who actually wins the most right now?

Matt (35:17)

I don't know. I would say right now it feels like the rebels, it's an advantage to, because they're just, they're nimbler, they're, they're quicker to adapt it and to experiment. I think in a lot of ways maybe it's lower risk for them to experiment. So right now I do think it's an advantage, like I'm, I am looking to, you know, what's happening in dtc, what the vibe marketers are putting together to, to kind of see what's at the edge from a content creation capability. But you know, I could see as it becomes more, you know, frankly, better, you know, becoming an advantage for the incumbents, you know, because what I, what I think about is what is this going to do to like the long term margin structure of our businesses, right? Does it, does every business become, you know, five, five points more profitable or does that extra five points that it frees up because your content budget is smaller get redeployed back into more ads, you know, higher customer acquisition costs and, and if you don't figure this out, then you're at a disadvantage. And you know, I think everyone will. And so I think it's more going to be the latter where this is just going to free up dollars and that are going to get reinvested in more growth and it's going to, you know, just increase the level of competition even that much more.

Craig (36:26)

You know, Sean, to your point about capital, like we're talking about something that costs $20 per user per month, right? So it doesn't matter that you're sitting on billions in cash when things are that cheap, right? And I go back to this quote, you know, 50% of the Fortune 500 from the year 2000 are gone. They just don't exist anymore. Right. They either went out of business or whatever. Right. And so I think we are talking, I feel strongly about a transformation like the scale of the industrial revolution, like this is the real deal. And so to me, right, when you're at a big company right now, there's so much uncertainty and so much of your time is spent trying to justify being able to use the tool rather than actually leaning in and leveraging it. So no, I don't think that this is, I think this is for the rebels. Right? And I would go so far as to say, like as an employee, if you are hindered in any way from your ability to master and leverage this transformational tech, like that's not awesome for you and your career and you should probably go somewhere where it is more celebrated and encouraged. I don't know what the net outcome is for some of these larger companies and disruptives from incumbents. We're obviously talking about retail. If I was a venture backed software company like Monday or even figma, which just IPO'd, I'd be very, very scared of some of these up and coming tools that don't have the headcount or institutional stuff that we've got as a retail brand, I'm not sure. But no, this is for the rebels. And by the way, I mean it sincerely, I think that people are doing themselves a massive disservice if they don't learn how to leverage this technology and if they are encumbered in any way from doing it, that's problematic from A career development standpoint.

Sean (37:58)

So I'm of two minds, right? If this is the Internet, the Internet was very good for the rebels. The mobile phone was very good for the incumbent. Now I've launched two brands this year, right? So Ridge, awesome multi hundred million dollar business inside of Ridge. We spoke with a brand and we just launched. We're working on launching another one right now. Making the website in lovable three clicks, three prompts. I literally just said I'm making a supplement brand. This is groons. Make me a website like this. Here's my design pack. And literally I have a working functional website in three prompts. Now I can refine it, I can make it better. But like, like 50 prompts, it's like that's not, that's not that much work to come up with 50 props to make a working website. So that is really, really good for the rebels, right? Same thing in my ad account, bro, I will have 500 ads in an afternoon that are all high quality. I have a design pack, I have my talking points. I could spin up different Personas and it's like that's one person grinding for an afternoon. And I have what would take a thousand people at Walmart to make 500 really good ads right now. That is all for the rebels. Why incumbents might win is because of discoverability. If this eats the Internet and ChatGPT gets all of its product recommendations from Amazon. Well, I don't have an Amazon account for my new brand yet. I don't have an Amazon listing for my new brand yet. And that is like the deciding factor is if, if it's the gatekeeper of the Internet, then people who currently have Amazon listings are going to benefit the most. So it's, it's, so it's somewhere in between. I think the rebels are going to win. It's like, look, if I was 20, I would just have a product and grind all day to make the world's greatest AI brand. And there's a, there's a company called Minerals. I'm sure you guys have not heard of them, but they'll do a hundred million dollars this year. AI only creative like it is. And check them out, they have a crazy ad account where they just, they have a very strong point of view. They only use AI to make the ads. They have never shot a person. So that, that's, that's a, a rebel kicking. Okay, what is overhyped? Let's talk. We're trying to give people practical tips. What do they need to avoid? Right where is like, hey, it's not there, it's not going to help you. It's a waste of time. Is anything off limits? We'll go to Matt and then we'll go to Craig. What's up, Matt?

Matt (40:12)

You know, I think right now for me, what's been overhyped is just the ease of adapting these tools within larger organizations. I mean, the bottleneck for us right now is integrating these tools into our systems of record or, you know, our key operating systems and being able to actually embed them in the workflow versus being something that someone's doing off to the side as a kind of on their personal level. So when I think about, you know, an Oracle or a Windchill, which is our plm, every time we bring in an AI partner, they drastically underestimate the amount of work required to get these tools integrated into, you know, a corporation systems. And, and so to me, just the speed with which you can implement these tools within your organization is, is overhyped right now. It takes a lot more engineering or deployed engineering work than it's ever sold to us as.

Sean (41:10)

And this is, this is maybe with the bathwater type stuff where it might actually be better to throw out all of your systems and just build them again from the, from the ground up with AI. And that, that's the threat to legacy

Matt (41:20)

SaaS that, well, that goes back to your advantage as, you know, the, the rebel. Right? Like, you guys have more modern tech stacks that you're, and, you know, way less technical debt that you're dealing with than, than we do. Right? Like, it's, it's, it slows us down tremendously that we have to figure out how to get, you know, 30 years of Oracle data into these tools. And it's not clean and it's, it's, you know, there's so much customization. Everyone's Oracle instance is custom and it's, you know, so every time our partners come in, they have to learn a new instance of Oracle, even if they said they've done it before. So you guys definitely have an advantage in that regard.

Sean (41:53)

I always tell people just, oh, just move to Shopify. Get off all this stuff. And usually that's the right case. I have one friend who has a brand. It was subscription first. He was on a custom build out of something that wasn't Shopify. He moved to Shopify. And because Shopify is so much easier to cancel the subscriptions, he lost half his revenue in one day. So like just. Yeah, there is, there is some challenges when you actually start moving over legacy Systems. But Craig, what else is overhyped in the world of AI?

Craig (42:24)

I would agree with Matt wholeheartedly. I think that the ease of adoption, the change management of this, I think HR is going to be such an important point or part, I would argue even more important than a CIO in this. We are talking about transforming workflows and many elements of the nature of work. And so I think the ease of rolling this out, Matt went deep into some of the technical stuff, but I just mean some of that, the cultural elements, I think that's overhyped. I think, you know, I mentioned in my role, you know, we, we worked with a lot of venture firms and I would just argue that, respectfully, a lot of these, like VC backed AI startups are overhyped. I spoke with one buddy, oh my gosh, he, he raised at a $100 million valuation. He's still figuring out what his product is like, literally, like, you know, didn't give up that much of the company. 100 million bucks. You think that's legitimate? And we were just brainstorming on like what he needs to go build. So I think that there's a lot of hype in that space too. And I do, I just want to build on one of the rebel comments too. I don't know if now's the time, but there's something I want to add there. Go ahead. I think when we, we talk about, and I'm curious for both of you on this, right, Like Matt, with you and your, your CFO lines, I think speed, speed, speed, speed, right? In big companies, there's all these different reviews and matrix processes, etc. You know, Sean, you just outlined how in three clicks you built effectively a prototype of a, you know, of a website. And for you as the executive, you are now able to go dictate to your team, just go build me this, right? I don't need to go through all these different reviews or all these different processes. I've told you what I want, now go do it. Because I have this new skill set that I can bring this to life. Like, that's just work now. Like, how do you justify the return on that? Right? Like Bryce alluded to earlier, you know, he framed AI as like, what is the ROI of electricity in your office? What is the ROI of a laptop? Like, this is just core to how work gets done. And so when we talk about speed and justifying some of these things, how do you approach that? Because it's challenging.

Matt (44:21)

Sean, you said you triggered something for me too, but I was super excited when AI started taking off around this notion that legacy SaaS would get replaced by new tools or that you'd just be able to, you know, spin up your own solution to replace that. So far that has been kind of, from my perspective, overhyped. You know, we've got more than $25 million of SaaS spent and I would love to, you know, right size that I haven't been able to, you know, you know, get rid of a significant portion of my, of my legacy SaaS. Yet it's all been kind of incremental spend for me so far.

Sean (44:57)

All these things feel so close, but it's like when, when you start talking about SoC2 reporting, which you're going to require or whatever, it's like I don't think the lovable speed like spun up ERP is going to have that right. And you know, Craig, on your point, talking about like people like creatives were mad when AI came out, like, and they're still mad. If you go on Instagram or if you go on any design forum, people like are mad where people are making stuff with AI. But like it is just democratizing creativity, right? Like it is so good that I can 3 click a website then show it to my designer who's going to make it better because they're professional, they've studied it. But like it just now we can speak the same language. Like, just like having AI translate my phone calls from Polish to English so my reps can take care of them. Just like internally we can all be speaking the same thing when we're making briefs. It just lets me speak design language to my designers and I think that is obviously a positive thing. I think we'll make more ads. Sean here to tell you about Saris analytics and Saris Pulse. Ridge is profitable every single day and we've taken that super seriously since we built this business. We track contribution margins by day, we look at the SKUs we sell every single day and we have to do this manually up until Sarace Linux came out. We take all of our SKU level data, we build it into the data warehouse. Everything that goes into making a true P and L I get on a day to day basis. Sara's Pulse gives you clarity. So your CLO and your CFO and your CMO start speaking the same language. Contribution margin shifts teams away from hoping profits survive the season to manage them in real time. Book a walkthrough with the Saris Pulse team today. Click the link in the description and thank you Sarah for bringing you this Show

Craig (46:42)

a couple things to build on that. Right? Of course people are unclear or threatened or in certain ways, but I think we all need to get comfortable with it. I often say that it's like analogous to online dating where like I'm 37 a decade ago, like when my buddies met their wives or whatever on, on online early on I was like, oh, that's weird, right? Like, and now it's normal, right? It's like, of course you met your friend on Hinge. You wouldn't meet them at the bar. Like that's what's weird now. And I think, think there's this like Overton window where you wouldn't challenge an analyst on how they came up with their forecast. And like, oh, you used Excel, that's not appropriate. It's like, no, of course. It's just core to how work gets done now. And I, I think that's the, the, the evolution that is inevitable, particularly as more AI native folks rise the, the corporate ranks. But Sean, you mentioned communicating in the same language. Now I would go one step further if I could where like I can now communicate Sean, in your language. One of my favorite use cases, just like, like create an AI version of your manager, right? I did this all the time and I can say, okay, you know, this is what I plan to say to Sean based on how, you know, how he likes to process information. Is this appropriate? And like, no, he wants this, this and this at the top. And so now because of this tool I'm able to communicate in the way you like to, to, to communicate with or a similar example. And then I'll be done like, hey, I know our executive team likes their executive summary formatted it this way. These things are just algorithms and copycats. So I'm going to just take the last six months worth of executive summaries or whatever and then I'll be like, hey, this is what good looks like. Now I'm going to tell you what I'm trying to say. Just spit it out in the way that is optimized for what our executive summaries look like, right? I think this idea of communication but tailored communication in the style that people see fit is like a no brainer, super easy, you know, impactful use case.

Sean (48:25)

And we're just trying to get information like transmitted more efficiently, right? It's like the bit rate of data transfer. It's like people can only talk so fast, but if you're speaking the same language, it's just a little bit faster. So I think that's a great use Case, we're going to end the podcast talking about the future. Where is this going to go? Craig, I could tell you're incredibly bullish on a. It's the industrial revolution. I think it's more like the Internet, right. Where it's going to change how we work and live on most things. But this isn't going to be like, like people starving and then they stop starving because we invent farming or whatever. Right. But I'm still incredibly bullish. I think, I think you're, you're 101% on like 90%. Matt, where do you think this is going?

Matt (49:05)

I'm closer to Craig than you. I'm, I think, like I said, my assumption is that it will get good enough to do, you know, kind of any one of the functions we need it to do. And you know, I think one of the debates you hear a lot about is like, is it going to be, is the solution going to be like a general chat agent? Does, does chat GPT just continue adding features and capabilities where it just becomes a one size fits all or you know, general chatbot? Or do these, all these like product wrappers that are, you know, wrappers around LLMs, are they a viable long term solution? My experience is, is that these applications or these wrappers are actually, you know, viable and where I think a lot of the value is going to, to be created because you know, I think you need, there's so much domain specific expertise required to, to do any of these functions, whether it's, you know, creative or finance. And you know, we need, you know, we need someone that has that deep domain expertise wrapping it around the LLM and providing that context and creating the guardrails that are necessary within that function to make sure that, you know, doing the things it's supposed to be doing that is properly being evaluated. That's, that is a very significant value add that I don't think someone like an OpenAI is going to be able to do. So I do think it's going to be product specific applications for each function that are, you know, the best, you know, the best technical people combined with deep domain expertise that can handle things like those integrations and all the very specific content to copywriting or things like that to make it really perform.

Sean (50:41)

Yeah. So how are you guys changing your personal lives with the belief of where the world's going? And here's an example right now, legacy SaaS isn't dead, but I think all of us are sitting around being like, I don't think Adobe is going to be worth $200 billion in 10 years or whatever. Right. That really feels like a radio station in the 30s. It's like, yeah, eventually TV and movies are going to replace this. But I mean, I could obviously be wrong. So, I mean, Craig, are you doing anything in your personal life to change this and then we'll end on Matt?

Craig (51:10)

I am. I want to just. I agree with everything that Matt just said. I think one of the biggest changes that I've made in my personal life that has actually been incredibly rewarding and helpful. About a year and a half ago when I started leveraging this tool, I was just like, oh, okay. Like I have to make the. This thing needs to know every single thing about me in order to be my everyday AI assistant, my Einstein in my pocket. And I need to figure out a way to feed it and make that happen. So I started journaling like a year and a half ago. Now. It's like a 200 page document and I can now like reflect on myself how I handle certain situations, why I handle certain situations. I bring in a ton of different books that I'm either reading or downloading and it's almost like anthropomorphizing them to guide me in my day to day. Right. Like I've now got Jim Collins coaching me on my unique specific situation, grounded in my background. So, so this idea of journaling, logging, I am recording every conversation that I have, you know, so that I've got this everyday assistant, particularly in Notebook lm, that is just helping me every, every step of the way. I think that's been one of the, the biggest transformations and I'm so much better at handling certain situations as, as a result, that's been one of the bigger unlocks for me in a personal life.

Sean (52:21)

This is why it's like, it's so, you know, omnipresent because we're talking about how it can help you with your work and your job and making creative and making copy. But also we just described it being a coach and a therapist and like, what, what, what can it do? So Matt, how are you making changes in your personal life?

Matt (52:36)

I mean, I'm, I'm really just trying to lead by example right now. If I'm going to ask the organization to embrace these tools, I need to demonstrate that I'm doing it. And so, you know, I'm in Claude code, you know, every day. And I literally had not ever opened a terminal, you know, before four months ago. And now I'm in there trying to code and do these things and then you know, showing and demonstrating to the organization. This is what I built this weekend, this is what I did. And so trying to do that and then trying to really understand the technical is something that I, I consider myself non technical really trying to get it like a decent basic understanding of the technical aspects and the architectures and things like that of, of these tools so that I can be the translator to the organization. Because I think that's where there's just still a massive gap when you think about like, like you know, a Main street company like us. You know, I, I see on Twitter, I listen to all the podcasts around the, the amazing techno things that are being done but the vast majority of people still have no idea what that is or how you know how to do that. And so I'm trying to be that kind of middleman between you know, the, I'll call it the west coast and, and kind of Main street operators like, like Thistle.

Sean (53:44)

Last question before we end it. Who do we think is going to win the most? It's like or like or how do you position yourself to actually win in this? Right? Like you know, you sell a very practical good. I think even with you know, AI and robots and everything, everything, the world's going to change the next 25 years. People still have to vacuum, right? And Craig, you know, going back to Crocs, people are going to wear shoes but who do we think is going to be the biggest winners? Who's going to be the biggest losers? And you know, selfishly my thesis is oh, people are going to still buy stuff. This is a great time to be a consumer brand, right? Like I think it's a hard time to be in entertainment because I think everything could just be AI like stranger things of the ending. So it's like, yeah, might as well AI your own version of it. It's a horrible time to be in legacy SaaS because you don't know what's going to happen unless you're at like Nvidia building the next generation of hardware. It's really hard to understand where tech is going but I think people are still going to eat gummy vitamins. So maybe that's a good place to be building

Mike Beckham (54:43)

if you're scaling an E commerce brand today. Ads alone aren't enough. Aftercell focuses on the one moment that every brand already owns after checkout and turns the post purchase moment into more profit. Monetize every order with post purchase offers and thank you page experiences without disrupting checkout or hurting conversion. Enterprise grade tech used by Gap Ticketmaster, Macy's and Target. Now driving results for brands like True Classic, hexclad Ridge and Jones Road. I would know. This is the reason I ended up buying three pans from Hexclab instead of two. Aftercell has already generated over 1 billion in additional revenue for e commerce brands. Revenue that doesn't require more traffic or higher cac. So check out after sell and tell them that the operator sent you.

Sean (55:29)

Craig, how do you think you can position yourself to win in this, in this new economy, this new world?

Craig (55:35)

You know, I don't know who's going to win. I know Google is well positioned, you know, all that stuff but I know that the people that will win over the next three, four, five years are those who know how to use these tools to get better in their day to day to produce higher quality, you know, outputs more quickly. And when Matt talks about cloud code I'm just like, dude, take me with you on that journey, right? I'm non technical but I know I need to learn this stuff so I can tell you the people who will lose are those who don't know how to master this and those who win are the ones who are leveraging it non stop every day I feel, I

Matt (56:09)

mean, I don't think it changes kind of what makes companies successful today. I still think it's going to come down to, you know, the strategy of a company and, and whatever makes you successful today. Like if you're in floor care, you still have to bring the best product. You know, you need to bring new innovation to the market. You know, I think people lose track of that thinking about how to use AI within their businesses. They kind of are like a hammer in search of a nail. But it really should be the opposite way of like what is your strategy to win in your particular industry and how is AI a one piece or one tool in your toolkit for making, allowing you to, to execute that strategy or do it better. But you still need to have that differentiation, that strategy whether it's, you know, product or you know, whatever your business is. I don't think that necessarily changes everyone.

Sean (56:54)

Plug what you want to plug. So Craig, plug away. 20,000 people listening. How, how can they best get in touch with you? How can they have some more Craig in their life?

Craig (57:02)

I don't know that anybody wants more Craig in their life but if they want, just go to chat walrus.com we help normal people, we teach them how to work with AI, you know, to turn into, to sort of superheroes and so go to chatwallrus.com yeah, that's, that's that.

Sean (57:16)

Okay, Matt, your turn. Plug away, my man.

Matt (57:19)

Yeah, well, I definitely need the exclusive invite to Craig's weekly call here. You know, please support Bissell. We're, you know, family owned business, 150 years old. We make a lot of different cleaning products, not just vacuums. And so check us out and appreciate the support.

Sean (57:34)

And look, I highly recommend Craig's off the record conversations. I've spoken twice and he's had, I mean, people from OpenAI show up. He's had a bunch of great operators, he's had the hexcloud team, Mike Beckham, everybody there speaking and it's unrecorded and gets as raw as possible. So that's why he does it. Guys, thank you for being here. I think this is going to be a reoccurring segment. Just talking random AI stuff. I love that you guys both bring big company experience and are currently trying to help people use these tools every single day. Maybe I'll think of some more examples and I'll share them at the end. So thanks everybody for being here. Thank you for listening. Talk to you later. Goodbye.

Matt (58:10)

Yeah, thanks for having better.