Boost Your Amazon Sales with Better Optimization & Launch Tactics with Colin Raja – Episode 47 of the Agency Operators Podcast

Boost Your Amazon Sales with Better Optimization & Launch Tactics with Colin Raja – Episode 47 of the Agency Operators Podcast Speaker 2: Welcome, everybody, to the Agency Operators Podcast. My name is Pasha. Today, I'm joined by Colin Raja. I'm really excited to have you on, man. You're actually the first seller that I've had on the podcast. I mean, I've probably spoken to people who are first service providers and also sellers, but you are seller first, pivoting into service. So I'm really happy to have you on, man. Thanks, man. Speaker 1: I appreciate it. I'm excited to be part of it. I'm super excited about this call. Speaker 2: So we actually met in person the first time at the MDS event last year in Seattle. Yes. Yeah, that was a fun event. Everybody was drunk and playing ping pong. And then there was just people gathering around and then I saw you, I think, on the next day somewhere, maybe on the floor at the actual convention center. Speaker 1: Yeah. Speaker 2: So that's where we got to talking. MDS is a place where there's a kind of a mix of different service providers and sellers and it's interesting to get the seven-figure perspective. Seven-figure sellers, they're at a point where they can argue with the service providers because they think they know better. Maybe sometimes they do. The proof is in the pudding and if you've grown a brand to that point then you know what to do. Speaker 1: The funny thing is the difference between being a seller and service provider or like you're doing the service or something. When you're being a seller you see only one set of niche actually. You collect one set of data. We launched a crazy product like we launched a hundred products. So we have like so much of data at this point to figure it out. Anything happened, we can able to figure it out. We can able to come up with a solution super easy. So that's the major difference between a seller and somebody who is doing active launches, active data management. So it's a huge difference at that point. Speaker 2: Right. Yeah. And doing launches is very different than managing an existing six-figure, seven-figure, eight-figure account. Because when you're launching, especially today, 2025, launching a product is harder than ever. We all remember the days when you had this Ranking solutions and search find buy stuff and you know many chat Facebook like. Oh yeah. It was easy. You find a person who knows how to do this or you find a software or a tool or whatever and you can launch your brand. But now Amazon obviously cracked down on it. They want there to be an even playing field. So what are some of the strategies that you implement with some of these launches that you're helping out with? Speaker 1: Sure. So basically the basic difference what people are doing wrong in the launches, they just go with the hacks. Accent the basic format of what they're doing actually. That's fundamentally wrong, what I find. So each product works uniquely, each keyword works uniquely. So you need to find a way how we can rank this keyword in a right way. There's no right or wrong way, first of all. How we can be able to follow the data and take a pivot on like a ranking strategy I'm talking about. So some keywords like just doing external traffic, you can able to get ranked. Some keywords just doing a PPC, you get ranked. Some keywords are like you need to do external traffic, PPC, plus a potential external sales as well needs to have. So there is a hierarchy of Okay, if this is not working, let's go try this. If this is not working, we're going to try this. If this is not working, we can try this one. By defining based out of data, actually. That's where more interesting things happen. We can be able to find out, like, oh, we're not going with the hack, but we are creating the hack at that point. Unknown Speaker: You understand what I'm saying? Speaker 1: So every launch, what we do, we basically test at least, like, 15 parameters I'm talking about. So each and every parameter is just like a different answer. So we lose so much, but also we win a lot actually. So it's basically like, okay, how we can use the technology and data, basically push yourself in a better way. That's what we implemented. The last three years, our whole launch strategy was changed. Speaker 2: Okay, and so every subcategory you're saying is totally different. You really have to just read the market first to see what are the levers that can be pulled for that particular product, right? Speaker 1: Exactly right. That's where the data comes in, right? If you launch it through subcategory, right, any subcategory, it behaves, baby category, it behaves differently. Health and health subcategory behave differently. And we need to find out, like, okay, if it behaves differently, what are the different attributes, what we need to follow? That's the one thing. The second thing is that basically, most of the bigger brands, they have no idea what an active organization is. They have no idea what is a browser. They have no idea what is product line. They have no idea anything about item type. And GL, DOW, they don't know anything about it. So worst case, what they're doing is like, most common mistake, what I find is basically is like, once the keyword ranking drop, they increase the PPC. Right, I'm going to give away like something, right? I'm going to do the Excel traffic and see that if you can able to come back. But the point is, if your backend was not optimized properly, you probably ended up like not able to rank any point of time. So we've done it for a couple of brands. They're like giving me like 500,000 units a week of Excel traffic and everything. They're not moving anywhere. And we were like, okay, let's take a look at the back and take a look at the super messed up. And we were like, okay. They immediately went back where they were at that point. So, yeah, that's the two interesting factors what we change in our launch. Not making a decision based on our intuition, not based on our emotional things. We purely act with data and data gives us all the answers and based on data we proceed with it. Speaker 2: And so you're saying that the backend is extremely important. Speaker 1: Extremely. I don't know how to emphasize this. I've been talking about this for a year, two years right now. How important is like, basically, I give this answer for, it's basically like, okay, you want to drive car, your road needs to be clean. If the road was like a really bump subject, like a holes and a whole bunch of things, you're probably not able to drive fast. That's what the back-and-forth decision looks like. If your back-end was not optimized properly, you are not able to drive, you are not able to drive faster. But the question is, some people come and say, hey, I'm in the wrong category. My sales are pretty amazing. Should I need to change? Or should I, what's going on here? So this is where it comes to like an old algorithm and a new algorithm right now. When the product was doing pretty good, there is a parameter called, they give a higher weightage for sales. So when the sales are much bigger, right, at that point, you don't have the other parameters are like, for example, product app, enterprise, it's like 0.5. Product app is like a 0.5. Reviews are like a 0.2. Ratings are like a 0.1. So all these things come together, they give the value. Based on that value, they push in you as a company. But if your sales is like crushed compared to your competitor, you're probably gonna win it actually. That's why you have like, you're not crushed yet, but this is what is gonna happen in case. If your sales went down, your product was suppressed, your product was like out of stock, at that point, you will see the problem, the new algorithm picks up, then you will see a lot of problems. That's what happened. In the 2020, people went out of stock. To bring it back because it's not nothing to do with the honeymoon. It's nothing to do with anything. It's just like an Amazon bot automatically update a bad data based on the foreign agents. So what are you doing? That's the first thing what we need to check at that point. Speaker 2: Mm-hmm. Yeah, you got to be driving in the right lane. And then there's like all of these other fields in your back end. Additional columns and things like that. And there's certain valid values that go into those fields too, right? Speaker 1: Yeah, it's called BTG Guide. I'm sorry, I'm a little bit sick. So basically, it's called BTG Guide. When you go to the browser, you will be able to see, there are like multiple attributes there, right? There's acquired attributes, optional attributes, and preferred attributes. So the refinements is like one of the attributes you can define based on the browser, right? So if you don't fill the refinements properly, you're probably losing 15% of sales at that point. Well, because when you have so much options, people start choosing the filtration process at this point. You're not optimized for the filtration. You end up losing it. And what we started doing is like, even though we are very freshly testing Roopers and everything, it works much better because we make sure that the backend data was filled up properly. And then our image optimization was like built up properly to make sure so anytime that group has asked questions, we automatically pick stuff for them. Speaker 2: I love it. And so before all of this, like working with all of these launches, you were a seller yourself for a long, long time. What were some of the things that really worked for you? Speaker 1: Yo, we are crazy, man. We are crazy. So, I'll give you one example. So, we got a first issue with our back-end data, right? Chairperson team apologized, and also they gave me like a, sent a whole crowd-source ticket, nothing available online. So, what we did is that we basically take a look at all the competitive products, right? And whoever messed up the back-end data in 2014, We started launching that product in the first day with the proper backend optimized. We are like getting like 50 to 60 sales on the first day itself. So because it's like we adapt to the new algorithm, not adapt to the old algorithm actually. Right now, it's like we are like a dynamic re-ranking and EQR model. At this point before, we are the relevancy model. So it is not like a It's not like they are not like replacing it, they update it on top of it, put up a priority on the latest one. But still, the old one still exists, but that got into like the lowest priority at that point. We found that kind of strategy and then it's like a pretty much work and always we try to innovate something new, even a small element of it in your listing. What we did is basically like find out like all the reviews, download all the reviews, put it up in like a find out like what are the common words they use, what are the What are the things they like? What are the things they don't like? You can also look for like a I wish or something that keyword, people gives a clue for you what this product they want it to be, right? So those are things we take it before the product development phase. So what we did is basically that's the one thing, pivoted our business in a different way. What happened is we're not launching any kind of like a go-to product. We're launching product with like a unique change, unique data points, what it's like people are missing. So we started taking that value and implementing it. So our whole goal is like, before the whole goal was like, okay, can I make 30, 40 sales out of one product? Can I launch 1,000 products at that point? That's it, right? We don't have to be top of the category. We don't have to go on a crazy stuff, actually, at that point. But that was the previous goal. Right now, since we have so much data, we see so much gap, man. The gap was crystal clear, how we can be able to grow from one Why are your competitors winning? And why are you losing? What the backend data looks like for a competitor? What's your backend data looks like for you? Unknown Speaker: Why are the front-end winning? Speaker 1: Why are you not winning? And what is the keyword level? Why are they winning a lot of sales? And why are you not winning? If you have these four brackets, it's a data at that point. So you have everything in your hand to make the pattern for your listing. So first thing first, what we do? Unknown Speaker: Okay, let's go product opinion. Let's throw up a platform of product opinion. Put it up something like find out why somebody is here. So we take a look at this and hire a designer, come up with like a crazy thing, just really good and we put it up there and we get the proofing. So our CTR went from 1.5 to 5.6 dollars because of that one change in what we're doing, right? Why are we not doing it? Speaker 1: So when you do that, then the backend data, it's already obvious it's fixed. You're going to fill it up. On the other side is keyword. You know how many keywords your copywriter ranks and how many keywords you rank. So what is the gap? You say, hey, okay, my copywriter ranks for 40 products, 40 keywords. I rank for only like three keywords. Okay, all I need to do is like, I need to rank for this 40 keywords. So it's a very crystal clear roadmap what we have when we take a look at that level of data. Speaker 2: Amazing. And so how does that scale? So you're saying, okay, if we can get 50 sales per product, well, let's do it to a thousand products. Are you actually seeing that when you start launching new products? Is everyone a hit like that? Or even if you're applying the same formulas, it's not going to work for every single SKU. Speaker 1: But again, of course, we can be able to launch a hundred products, a thousand products for sure. There's another level is like, do you have a budget for it? Unknown Speaker: Do you have like an audit, right? Speaker 1: And also the fear factor and a whole bunch of things involved in place. Even though you're not making emotional decisions, but when you're paying money, that's where we get like a butterfly feeling in your stomach like, oh my God, this is going on. Am I going to be good or whatever? It's a real fact, actually. That's what it is. Speaker 2: You know, there's never a guarantee. Yeah. Speaker 1: I don't even want to pay any of my supplier. Like I get panic. So I tell my account that they can pay this money and like verify and pay. That's it. I don't want to know actually completely. So it's like, okay, I'm not putting my emotion. Why? Because when I put my emotion in place, I stop there and then I freeze based on the emotion. That's a good learning what I have done. Small thing is like, the emotions are given by God, which is like a precious thing, very precious thing, right? But most of us suffer out of our emotion, not feeling it actually. So what I put my principle towards in the business and also personal life, do not suffer out of emotion, just feel it, observe it, and let it go actually. Speaker 2: Yeah, we're suffering from our fear, which is coming from our mind. Unknown Speaker: Yeah, exactly, completely. Speaker 1: You're a second rumor, actually, completely. You're literally like throwing a whole bunch of things on you, saying, you know, it's not going to work, it's not going to work. Are you scared? Like, you're not going to be scared that you're going to be broke again? Okay, yeah. It's going to happen. Speaker 2: But that's the beauty of it. Like you said, it's data. So if you have, all right, in the beginning, you're taking a risk, no doubt. Like, you don't know for sure. But even still, you're taking an educated risk because you have data. Then you prove the hypothesis is true with these launches. You're doing all of the things. You're applying the knowledge that you have in order to do it in the best way with the most chance of success. And boom, you sell through that first batch of inventory. You get through, let's say, 50 of it, 50% of it. And then you say, all right, Now I have data to prove that this next investment. Speaker 1: If it's a similar product, I still, we're going to do it. So I give you one example of what we did. We did one of the baby product recently, three months, four months ago. We launched for breakeven costs because it's a super cheap market, but they are coming up with a premium segment. So we were like a little bit above the breakeven cost. So we say, okay, we'll slowly increase it, right? We launched a white version of it and we got like close to 40-50 sales. So we ranked for most of the, we ranked for 50% of the keyword on the top completely, but there are people who are making 5,000, 10,000 sales. 50 is not a lot, but it's still, it's like a very amazing, returns for $79 product actually. You know what I'm saying? Speaker 2: That's good. Speaker 1: So what we did is, once we did it, and then they were like, oh, can we launch the variation? And I was like, no, let's go separately. Let's go launch separately. So we did the same thing completely and launched it with the same set of keywords. Oh, that product, again, making like 20, 30 sales extra. Speaker 2: Great. Speaker 1: And we're launching another color right now. That color is going to make another 20 sales. So all of a sudden, we're hitting on the whole color base. We're going to get like 100 sales a day actually. You understand what I'm saying? So once you found the data, once you know that this data works, you just replicate as much as possible. Now I'm telling the brand, like, hey, can we come up with like 10 products like this? Can we design 10 more products like that? Can we just take over the whole niche? Like a million? Like 10 products. Why not? We can handle 10 products. Like a small version, bigger version, like under new designs. That's very possible. Speaker 2: What are you thinking about sourcing? Like it's only from China or maybe now, especially with all the changes happening politically? You know, I always know this thing is coming up, right? Speaker 1: So in 2017, what I did is we traveled close to seven or eight countries, right? And look for a partner, manufacturing partner. So our strategy is like what we launched the product before for my brand. First batch, first two batch, we go to China. It doesn't matter. But, on the third and fourth batch, we start working in India, Pakistan, Bangladesh, Taiwan, and all of the country, as long as the inventories, as long as the material they have, we'll be able to crack that code actually, easy. So, launch is much better when you go for a launch first, like launch first, like super fast. Using China market and you can optimize it. Launch doesn't have to be profitable. Launch is never gonna be profitable. Speaker 2: It's a very- It's just an experiment. Speaker 1: It's an expense. It's an expense for 45 days. Next 45 days you're gonna recoup the money. Three months you're gonna get the money back breaking the cost. Most of the time. Which business in the world give, let's say you put 20, 30,000 dollars. In three months you get all the money back actually. I mean, if people are not even ready to get that profitability, they're not in the right market. Speaker 2: Yeah. Yeah. If you want to be profitable from day one, it's impossible. Speaker 1: It's impossible. So you need to be like, I have a budget in place, make sure that what is the budget, check it out and everything, and make sure that you launch the product based on that. You don't have to launch three products. Launch one product first and build the momentum based on that. Speaker 2: And do you see AI changing the Amazon space at all? Speaker 1: There's so many things, man, but if you keep it simple, you don't have to too much worry about it, very honestly. For me, I'm using AI for all the, I clone myself in ChatGPT, making sure that all the knowledge of what I have, I dump it. I'm able to produce like all the data in terms of like algorithm aspect of it and everything completely. So, we just have to put the image in our ChatGPT or Claude AI. It automatically gets the whole listing and break it down into different groups I think. Like a keyword based on the groups, like a keyword groups and based on the semantic keyword mapping actually. So we divide both of these and see where is this, where are we getting advantage here? Whether we're getting advantage out of like a group keyword or we're getting advantage out of the semantic Seventy keywords actually mapping at that point. So based on that, we make a decision, okay, all right, this seems like, okay, the Libra and also the Spandex, right, are that those keywords are coming up because they're semantically mapped. So that means, okay, let's go behind that set of keywords. So you build a momentum at that point. Speaker 2: Love it. Colin, if somebody wants to reach out to you, they need some help with the launches, how do they find you? Speaker 1: For one, you can reach out to me through Facebook, Colin.Raja, there. The second one, it's like, I have the latest content, whatever, I post it on YouTube. You can be able to check it out. Those are the two places. It's like I'm wide open, sharing all the information completely on latest algorithm, latest algorithm changes and everything. So you can find more information there. You can reach out to me if you need help. Speaker 2: Awesome. It was nice chatting with you, Colin. Thank you, everybody. Speaker 1: 100%. Speaker 2: Check out Colin everybody. See you on the next one.