EP #372] [ENG] - How to leverage AI with your PPC campaigns - Cameron Scot

EP #372] [ENG] - How to leverage AI with your PPC campaigns - Cameron Scot Unknown Speaker: Welcome to The Ecommerce Lab By Ecomcy. This is the place for everything related to Amazon private label and e-commerce. Learn exactly what you need to start or scale your business. Get insights from the top industry experts who will discuss the latest trends and best practices in the world of Amazon. From choosing products and sourcing from a supplier to setting up your Amazon account and marketing your business, you will hear it here. Let's get started. Here is your host, Vincenzo Toscano. Speaker 2: Hello, guys. Welcome to another piece of The Ecommerce Lab By Ecomcy, the place where everything related to Amazon FBA, private label and ecommerce. My name is Vincenzo Toscano, founder and CEO of Ecomcy. And today we bring another special guest. His name is Cameron Scott, and he's the founder and CEO of Merch Jar, which I would say is one of the top solutions out there when it comes to really helping you automate and maximize what you can do with your PPC, which we know is one of the things that most Amazon sellers are struggling with, especially in 2026. It's getting more expensive, more complex, so many new tools and features that Amazon is rolling on a weekly basis at this point. And if you're not on top of it, it can be very overwhelming. And there is some, what I want to bring Cameron today is because given that MerchShark, they specialize on everything that's to do with automation and re-leveraging AI. I just want to really hear from a founder behind this tech, what is happening, what Amazon sellers are seeing implementing this kind of tools, the pros, the cons. I'm just really understanding, you know, where Amazon PPC is heading when it comes to leveraging tech like this. So, Cameron, welcome to the show. How are you doing? Speaker 1: Hey, thanks for having me. Yeah, it's a lot. This is a big topic, so I don't know if we're going to get to all of it today. Yeah, but it's hard to not talk about today too, right? Like with AI. You guys are using AI, I'm assuming, at some level. Speaker 2: Yeah, yeah. Speaker 1: In your workflows. Speaker 2: Yeah, yeah. Speaker 1: Yeah, it's a really interesting time, especially just given the landscape. I feel like Everyone's at like a different point in their AI journey. Like there's a lot of agencies and just people in general that haven't really gotten into AI yet. Or maybe they're just starting with like the chat bots. And now we're getting into more like the more agentic stuff. You got the Claude codes, CoWorkNow, Codex, all these that are now actually like doing the things. And so we got people at different levels. I tell you what, these early adopters are, they're eating pretty good with what's possible. Yeah, where are you at in this like journey? Speaker 2: With like you guys and your company, I mean, for us, yeah, AI is being leveraged significantly across all pillars, not only PPC, but also, for example, listening optimization, which is copy images. We also do a lot of research, provalidation, ongoing competitor analysis, reporting. So it's pretty much embedded on every single thing that you might have to do as an Amazon brand on a daily basis. I think the goal, of course, is we also have to try to some extent to make sure that this is always A supervised by humans in this case or experts in house because I think the point that I feel a lot of people also getting wrong is that the field is only click and forget and they're letting this. You know, systems and this software running on their own. And that's what I think actually, you know, getting the most of them, because I feel most AI solutions and software solutions out there are meant to make your workload easier and be able to get to the outcome in a more, you know, efficient way overall. But I feel it's not still designed to just kind of set and forget, which is what most people are doing. That's why they're getting bored. Speaker 1: It's not a magic bullet, which I think a lot of people wished it was. Speaker 2: Like, right. Like, it's cool. Speaker 1: Like, we all want that easy button to just do it. I think what we're starting to see is that it's just it's more of an amplification of someone's talents and knowledge, efficiency. But the way it's going, it's not going anywhere. I think what What I see is in five years, there's going to be agencies utilizing AI and just Amazon brands in general. And then there's going to be ones that aren't in business. Like it's going to be one or the other with how much more efficient these companies can be. And it's not that it's replacing the people on the staff, it's just allowing them to be so much more efficient, which then you can be a cost efficient with your clients and so forth. Speaker 2: I know, I know. I think for us it's the same. We're seeing that it's not necessarily, even internally, we're not even reducing our teams from anything. It's actually the opposite in the sense that now with the team that we have, we're able to output, you know, 5, 10x what we used to output. Speaker 1: So you bring on more clients. And that's what we're seeing too. I mean, traditionally, With agencies, it's been, you know, you need to grow, you need to bring on more clients, right? And to bring on more clients, you need to bring on more headcount. So it's like your labor's always scaling, right, with your revenue. Now that's just a different equation entirely because the same people that you have can just do a lot more. Yeah, exactly. It's incredible. I'd say I'm like on the newish side of like getting into the agentic stuff. Like I'm not a technical person. My co-founder is, which is fortunate because I get to like, Learn from, you know, these real nerdy guys that do all the crazy stuff. So I'm like newer on the cloud code and so forth side and it's just been mind-blowing like what you can build. And so for like Mercer like the short of it is like we make your data accessible, your Amazon advertising data accessible to these AI platforms and it's not to replace people kind of like what you said is to amplify what's possible like in the hands of You know, an expert Amazon ad strategist. It's I mean, it's it's mind-blowing what you can you can do. So that's kind of sort of like we're making this data accessible because we see the future like AI is it's going to be completely different in just a few years after five years who absolutely knows like you eyes are changing. It's already changed my workflows entirely just from since starting using it. Like I these tools are becoming more of like the interfaces. That you're using like I'm spending more time in Claude co-workers is my choice. It's a little bit easier than the terminal type stuff for me. Yeah, it's a great entry point. And I think it's if you're not, you know, any of your audience isn't hasn't dipped their toes in yet because it is like intimidating, right? Like as a non-tech person. Absolutely. You should be at least like dipping your toes in this because It's survival over the course of time. You're going to have to. You're going to have to keep up and it's only going to move faster. It's crazy how fast it's already moved. Another point you kind of touched on is the kind of safety factor of AI. I think this is something we do see. So like agencies or brands trying to implement AI and they are kind of just letting it do its thing and they don't have these guardrails around it. That's really important, especially as an agency. You're making decisions. For clients, I mean, can you tell a client that, you know, the AI did it? Like is that? Speaker 2: Yeah. I mean, it's no idea. Like, you know, for us, like we always, when we do anything with AI, we always say our clients, you know, it's like, well, of course, AI is a big kind of source of insights and data for everything that we do. But we always have the human factor having the final say if this is right or not, because As I say, like I think where people is getting AI wrong and that's why they're having the desired outcomes is they're letting it autopilot. So I think for you to avoid this, at least at this stage of where AI is, you need to have the expertise of that consultant that really understand what the outcome should be high level and then cross check it with internal expertise or what they have seen in the space working and just using real judgment that at this point AI cannot really bring to the table. It's more analytical, you know. Speaker 1: Yeah, absolutely. I mean, there's a few points there. I mean, it's one thing I see is that amplification like it not only like it does. Someone that's brand new to PPC or anything like it can help kind of elevate their skill set, right? But it does the same to like an expert too to where I can do some kind of statistical analysis and it's been 20 years since I took a stats class. I don't remember how to do that but it can help me walk through those things and that's kind of the approach we've taken too with we've kind of seen the same just across the industry where It's kind of you got these tools that are these like if this and that type tools that everybody's used like real basic. That's like last decades. Speaker 2: And they still call it AI. That's the thing. Speaker 1: Well, yeah, that is it. Unknown Speaker: Yeah. Speaker 1: I mean, AI, it's really just algorithms all the way down if you really think about it. But the other side of it is like what we're seeing is like other tools are just like bolting AI on. And we're seeing that with like, oh, AI is going to manage your bids. And it's always like some kind of black box. You don't really have control and against that magic bond. Great when it works. But, you know, what happens when it's not? And as an agency, it's just not it's not defensible. You can't tell your clients like, I know it's doing it. It's like, trust me, you'll be fine. It's just learning, you know, give it a few months or whatever else. You can't do that. Like they're coming to you. For the expertise and I like if you want to boil it down Companies hire agencies for your expertise like your business is scaling that expertise across multiple clients So you like you got to provide that service and and I don't see AI competing with that necessarily. It's just gonna be you know They're gonna agencies are going to make more of it. So, you know having this in some skilled hands is always gonna be much better More efficient and skilled than someone that's not. And we've taken that approach with that. It's like the data side is one of them. I mean, if you're working with Amazon APIs, they're a mess. First of all, you got so many different reports. You got to join all the data and clean it and do all this stuff. You got to structure it before it's even really like usable. And it's a lot of data. I mean, we process billions of rows a day of data. So that's like one piece and that's one part that was I mean, it's a real engineering challenge just to get the Amazon data like in a usable format. So that was like one piece of it and then our other side to like make this usable. What we're seeing with AI is a lot different than what I expected. And I think for a lot of people a year ago, where it felt like AI was gonna be almost like a commodity, like everyone's gonna have access to like the best models, like the cost is gonna keep going down. And we're at this like inflection point. It's like, we're at the switch part of the date and switch where like everybody's using it. It's great. And now we're seeing like cost go up, usage limits change, all these things. So it's like we're seeing it's like no AI is like actually gonna be like kind of expensive. Speaker 2: Yeah, I always say that to my people. It's like people think that, oh, yes. And I see that, by the way, I also think that's a huge mistake for agencies and also brand owners. They're going the whole stream of firing all their teams, rely everything on Claude and everything. It's like, do you know that most likely in the coming year, we're already seeing, as you mentioned, tokens are going up, limits are getting increased. In some instances, I've seen people spending more on tokens on Claude. To the point that it's not even financially feasible in certain case scenarios. And we are still early. Imagine when they actually deploy the real prices that they have in mind. Speaker 1: Yeah, like we're not even, like all these companies are blowing money. Like they haven't gotten to the point where they're looking to recoup it. This is like, you know, the first hits free is where we're at right now. Speaker 2: Just to get all your data and make you so, it's like a drug to some extent. It's so dependent. Speaker 1: So it's, it's crazy to think like a year ago, it's like, oh yeah, there's going to be these massive context windows. You just dump all your data and it's going to do it all. And it's going to be super cheap. Nah, like it's, it's not going to be the case. So it's like, and this is, it's funny cause like we've been, we've built these tools like specifically around AI where most of them, bolted, bolted it onto just like their existing structure. Like we've spent the last 18 months rebuilding from scratch for this new data landscape. So we're really like building. A platform that works well with AI is essentially what it comes down to, and not every platform does that. So start with the database. We bring in all your data, but it's also a specialized database where it's not only able to be read really quickly, but we can also write to it really quickly, which is really hard in databases if you're not the technical nerds. But the right portion is the action side of it. So a lot of tools can read the data and build dashboards quickly, but they're not actually performing actions on it. And that's where ours is a little bit different. Speaker 2: First of all, if they try to do with all their clients at once, they try to limit how many changes they can do on all of that, you know? Speaker 1: Yeah, exactly. And then the data side of it, you can't dump that on your AI agent. Like, you know, all this data, for one, it's gonna be expensive and it's just gonna cause confusion. It's gonna hallucinate. You're not gonna get good data. And the other side that I think a lot of people don't think about, like when you're just like letting The AI decides something, it's not deterministic. So it's going to be, you could have the same inputs, the exact same data, and a different day, it's a different response. And it's another thing, like, when you need to have control over these situations, like, that's, that's a hard pill as well. It's like, I don't know what it's going to do, technically, like, or even like, you think it's down, then the model changes. And we see that now, like, we're seeing with Claude, like, The same opus is not today is not the same as it was two weeks ago. We're getting different response from the same input. So that's another challenge. So we ended up building in between your AI and this data warehouse that we have of all your data structured that's easily accessible or like ready to be used. We have this layer in between this query logic layer. So it actually allows The AI to access your data, but get only the information it needs. So it's able to create its own queries and get just the slice of data, the segment of data that it needs. So it's context efficient and it's only going to be, you know, have this to work with not billions of rows. And then it's to your point about the safety. We do the action, so changing bids and so forth. The AI can do these things. It can write this logic and create these automations, but it can only do the things that our logic layer allows it to do. It doesn't have direct access to your Amazon. I think that's really important. You need to have this middle layer that you can have guardrails on. You can see exactly what it's going to do, and that's where I think the real differentiator from anyone else I've seen is seen with doing the black boxes. Ours is all open. You can see exactly what it's doing. And it's deterministic. So like we this logic layer that sits in between for it to query. It's going to run the same every time. So your bids are going to change with the exact same logic every time. So AI's not managing your ads. It's helping you create the logic. Speaker 2: The behavior. Speaker 1: To manage your ads. Help you like create those automations and do things that like would really be hard for a human to do. Because it's really good at like the programming stuff. We've seen that with AI's like turns out it's like it's really good at programming. So we let it speak to your data in a language it understands. And you can just talk to it in a natural language and do things that are absolutely incredible, far beyond what these, you know, if a cost is above 30% lower, it's 5%. Like it's whatever you want it to do. And that's the other side of it. Like every agency is different, right? Like your workflows are probably pretty unique to you. The same as every other agency, and that's where I say, man, these tools are really going to shine. Everyone's going to have like their own custom workflows. This is going to be the brain of it all. It connects to, you know, your Amazon data with Mershar, and then you can connect it to whatever other applications you use in your stack. Speaker 2: Yeah, cool, cool. So now with all this kind of introduction, of course, the importance and the landscape and everything, I think something I would like to bring to the conversation is maybe some quick use case scenarios where AI can really be impactful on your day to day operations. And I'm sure you're seeing this with a lot of your clients. Essentially what I'm trying to ask is if somebody listening and watching this episode right now, they haven't used any AI at all when it comes to PPC, what do you think is the first thing that you can mention that AI can help them now do better? Maybe it's bids, maybe it's budget, maybe it's finding keywords. What is the initial layer they should start exploring with AI? Speaker 1: Yeah, I mean, it's going to depend on your experience level for sure. Amazon PPC and kind of where you're at in that AI journey as well. So if you're like just if you're like just getting started with AI in general, you know, I'm probably saying like start with some personal questions in the chat bot, you know, kind of like get the hang of it, but like there is a lot to learn. So if you're there like You got to get you got to try to get caught up like it's it's so important to learn like how AI works at least at a basic level. So if you don't understand like how we've been talking about like context and things like that, like those are the things you have to really know, like knowing how to prompt and if you're using like cloud code and co work, setting up your instructions and file systems. I mean, that was a learning process for me. And honestly, it's like, just ask it like, it'll like all these things, mostly from The AIs themselves and you there's all kinds of YouTube channels. I don't have anybody specific. There's a bajillion different content creators for this sort of stuff, but it definitely like get creative when it comes to PPC specifically. I mean, I have to use it within the context of Mercedart because I just don't think anyone's making your Amazon data exposed to AI in a way that really works. I mean, you can download like your spreadsheets and stuff and like dump it into your chatbot if you want to do that. Ask us some questions and that can provide useful Answers, but it's not going to like automate your account. And this is like usually one of the like first things everyone's trying to do, right? Because you need, you got to manage bids. You got to manage budgets. None of it's hard, right? Like this isn't hard. We do the same thing like every week for the most part. There's a little bit nuance there, but it's just, it's almost like busy work. It's more like tedious than anything. So like getting those out of the way is probably going to be like the first step. Our AI helps with analyzing the account and the data. And I mean, it's really going to be whatever your biggest pain points are when it comes to like an agency perspective. Usually they're going to be using some kind of tools to already be doing this. So in our case, it more opens up like the analysis side, be able to look at multi-time frame windows or be able to do statistical analysis that you don't remember from college classes. So if you want to do those sorts of things, but the one point I, you touched on this, I want to circle back to it with where you saw mistakes of, Companies letting their employees go maybe a little prematurely when that happens. So like with a lot of agencies and this kind of happened, like goes back to you scale clients with headcount. A lot of the business knowledge lives inside those operators heads. Like you're hiring them because they're experts, but they're like performing this. You know every time themselves once they leave the business that like knowledge is gone like it doesn't exist anymore. So one area that the way we've structured MerchChart to work is that it helps codify that expertise into business SOPs really that can scale across all your accounts. So when you know you come up with some like Bit algorithm that's proprietary to you. That's now, you know, an actual tangible piece of logic that can be duplicated across all of your, you know, client accounts as an example. Yeah, you're like codifying and writing down this like, you know, you're at your staff's knowledge. Speaker 2: Yeah, actually on that note, like we're doing something Similar in the sense that we're trying to, all the learnings that we get from all our brands, try to create our own knowledge base of how to certain things. And that's just from the wins and the losses of all the brands that we have on our portfolio, which is significant now. And then essentially the brand, you can call it now, is setting the learnings from the foundation, all the things that we have done in the past with all our clients and all our team members. Speaker 1: So yeah, absolutely. I think there's an area AI can really help and there's so many different ways to use it. But I use it similarly in my business a little bit different but similar like with your clients and brands like you're able to use AI to like distill that information from your calls, like the call transcripts, your messaging with them, like distill that into Brand preferences or requirements like, hey, this brand always does this or just like additional context. That's how I want to like really frame is like this context about each client that AIs can use this context. So like in my own case, I'm going to set this up in my We're going to look at project folders with something like Claude coworker code, and it's going to use this. Now, AI has this context on each client to personalize recommendations or automations specifically to them, whether it is, you know, this is their target a cost. This is the kind of stuff they do. This is their brand list of keywords we never want to negate. Just those kind of instances can be hard coded into this context that gets used every single time. You know, you don't need to really worry about teaching every single, Hey, here's our list of clients, you know, make sure to memorize every single one. Like this is like built into the knowledge base. And that's kind of the other layer that's always going to exist. Like, as I see this AI landscape unfolding, like there's the data, there's this choreo layer that I think is going to be Requirement, just the way AI is going with the context windows and amount of data to process. There needs to be something there that distills that amount of data that it's being given to just what's necessary and useful. And then around the AI is going to be that knowledge layer. For us, it's like we developed our own AI copilot and it's really just Different files that give it instructions that teach it or, you know, tell it to, you know, work from the lens of an ad strategist, right? And it teaches it to be a knowledge, like here's how to work merch jar, but it's customizable. Like it's all like open source. Essentially, you can customize it so you can have these. SOPs, maybe that's not the right word, brand preferences, all those things just codified and then as you bring on a new employee or whatever it is, account manager. Speaker 2: They already have access to it. DNA. Yeah, the DNA of the brand, which is I think that's where in the past we were not able to achieve with AI. Before it was all KPI driven, A, cost conversion, profitability. But what about the actual DNA of maybe we're an aggressive brand when it comes to certain advertising strategies or maybe there's certain products that we intentionally want to lose money because we know down the line we make money with another product. Or the agenda goes beyond Amazon. We have retail. So that's where I think AI essentially now, I see it more flexible compared to one, two years ago. It's now adding that extra context, that extra kind of System that you just mentioned. Absolutely. Speaker 1: It's so much easier with these like agentic type of flows, because that's the biggest difference I see. I've been using AI since, you know, the ChatGPT 3.5 days. And the biggest difference I see is that like with that was like you could do some amazing things, give it context with files and so forth. But anything from it, you still have to like do something with it. Personally, like I still have to like deal with whatever it came up with, like, oh, these, you know, this client preferences, I still gotta like get it to where it needs to. And now with these new agents, you know, the co-works and so forth, where they actually have access to your local file system, can connect to all these different apps, like Mercedar, like Notion, like Google Drive, like all these other APIs and MCP connections. It can do the things. That's the biggest difference to me is like, not only can it like, Distill this brand information from whatever calls and you know, whatever context there is into this useful Information you can use ongoing as a business for this client, but then it's going to put it in the appropriate place that it can then be utilized. I mean, how often does it happen like especially a couple years ago where you're on calls or whatever it is so much information just falls through the cracks that never gets captured. It's never been easier to capture some of that and especially from an agency side. You can really utilize that with AI for I mean, all sorts of aspects on PC, PPC especially, but there's also, you know, every aspect of the agency, depending on what sort of services you offer from a PPC side, it's, there's a, everybody runs their ads differently. Every brand is going to have different preferences and campaign structures and so forth. So to be able to just have that. You know, codified in your knowledge layer around the AI. I mean, it's so efficient, saves so much time. It really like it protects like the knowledge you built up from like You know, years of servicing clients like your expertise and you're giving that expertise as context to AI and that's where the real amplification ends up happening and it still makes it, you know, it's not right. You can't replace that with AI because it doesn't have that expertise like built in like you still have like that secret sauce as an agency and AI is just amplifying all that and A lot like giving it to even someone that's like maybe like a little newer to Amazon as it's going to elevate how they're going to be able to work with your accounts just by having access to that data. Speaker 2: Love it. Love it. Awesome. So I think Cameron to start concluding and bring this to closure like is there any maybe advice you would like to point of A mistake that you see people doing a lot when it comes to AI that you think could be good to flag or maybe an inset of what is coming, like especially for that we've been talking about what has been happening, the changes, what is coming. So as your final remarks, could be one thing, could be two things. Speaker 1: I mean, the best advice is like, like this is gonna be broad, I guess, but like, keep learning. It's changing so fast. Like, it's honestly hard to predict what it's going to look like. I mean, a year ago, I would have, I wouldn't have been able to predict this. I don't know that, you know, we're getting to Where they're going to, you know, Terminator and, you know, AGI and stuff. I think we're a few years off maybe from that, but it's just keep learning. You got to keep up with this stuff because I mean, the one thing I can say is like the ones that are using it now are already seeing huge gains in efficiencies. And then five years times, like those are the only businesses that are going to exist. Like it's, it's that start. And I think that if you haven't started using it yet, Like you have no idea like how powerful and like the scale. I mean, it's orders of magnitude more efficient. And the more you know about it, like the more power you can unleash. And that's really like AI in the right hands is it's incredible. Speaker 2: Yeah, super, super powerful. Awesome, Cameron. So tell us how people can find you and also explain what you guys do at MerchChart. Speaker 1: Yeah, merchchart.com will get you to wherever you want to go. We have content on there. We're on YouTube. But merchchart.com, if you want to see how we use AI in your account, you can sign up for a free trial. Use it. No credit card required. Or you can set up a demo with me if you want to see it personally on your own account. Speaker 2: Awesome. Love it, man. So it was a pleasure. Thank you so much for this conversation. Yeah, this was fun. Speaker 1: Thanks for having me on. Speaker 2: And I'm sure a year from now, we'll have to do a recap. Speaker 1: Maybe we should do a prediction one next, and then we'll do a review to see how bad we messed up. Speaker 2: Yeah, we should. We should. Thank you, man. It was a pleasure. Speaker 1: Yeah, it was a pleasure. I had fun. Speaker 2: Thank you. Speaker 1: Yeah, bye. Unknown Speaker: Thanks for listening to The Ecommerce Lab By Ecomcy. Be sure to subscribe so you don't miss an episode. While you are at it, we would appreciate it if you could leave an honest rating and review on Apple Podcasts, Spotify, or wherever you listen. That will make it easier for others to find out about the show and benefit from it. Want more? Visit our website at www.ecomcy.com where you can get your first consultation for free or find us on Instagram, Facebook and LinkedIn at Ecomcy.