In this episode of the Ecomm Breakthrough Podcast, host Josh Hadley shares seven practical ways his e-commerce business uses AI to optimize operations and scale growth. Drawing from his experience building an eight-figure brand across Amazon, TikTok Shop, and Shopify, Josh covers strategies including building custom GPTs, automating TikTok Shop listing optimization, streamlining hiring processes, leveraging Alexa data, analyzing meeting transcripts, scaling ad creative production, and cloning leadership decision-making into AI-powered SOPs. Josh emphasizes treating AI like a new team member requiring proper training, offering actionable, real-world insights over hype.

Bullet Points:

Timestamps:

00:02:00 Weekly Custom GPT Creation
The speaker’s 35-person team is required to create or enhance a custom GPT weekly to automate their specific tasks.

00:04:05 AI Agent for TikTok Shop Optimization
An AI agent integrated with the TikTok Shop API continuously tests and optimizes product titles, descriptions, and main images weekly.

00:08:32 Automating Hiring Case Study Scoring
AI is used to automatically score applicant case studies based on a predefined rubric, saving hours of manual review time.

00:11:29 Custom GPTs Integrated with Alexa
Creating custom GPTs that analyze customer questions on Amazon Alexa to optimize product listings and improve Alexa recommendation rankings.

00:12:11 Analyzing Company Meeting Recordings
AI analyzes transcripts from all company meetings to identify business constraints, track team progress, and provide a leadership pulse.

00:13:56 Scaling Ad Creative Production
Using AI video generation tools to quickly produce a high volume of ad creative for Meta and TikTok campaigns.

00:14:48Cloning Leadership Judgment and Decision-Making
Using AI to document processes and decision-making frameworks from leaders, creating an internal knowledge base to empower team members.

Links and Mentions:

AI Tools:
“ChatGPT“: “00:02:00”
“Claude AI“: “00:02:00”
“Fireflies AI Notetaker“: “00:11:25”
“Veo3“: “00:14:27”
“Notion“: “00:16:24”

E-commerce Platforms:
“TikTok Shop“: “00:04:05”

Videos and Resources:
“30 60 90 Day Onboarding Framework“: “00:07:52”
“Episode on Cloning Yourself Utilizing AI“: “00:15:24”

Transcript:

Josh Hadley 00:00:00  Today, I’m going to be walking through seven different ways that we are implementing AI into our e-commerce business and practical steps that you can take to implement it in your business as well. Welcome to the Ecomm Breakthrough Podcast, I’m Josh Hadley. I’ve scaled my own ecommerce brand from 0 to 8 figures, and I’m actively building towards nine figures in sales. This podcast is where I document that journey and share the systems, the strategies, and the lessons learned in real time so that you can learn what actually matters and scale your own business. My name is Josh Hadley. First and foremost, I’m a man of faith. I’m a husband to a beautiful wife and also the father of four children. I’ve been selling in the e-commerce space for over a decade now, doing multi-million in revenue on Amazon, TikTok, shop and Shopify. And I am also the host of the number one business strategy podcast for ecommerce, and that is E-com breakthrough. Today, I want to dive into the practical use cases of how we’re implementing AI into our business.

Josh Hadley 00:00:58  Today. I hear a lot of noise going on in a lot of the e-commerce groups. There’s a lot of like doom and gloom of, oh, you’re getting left behind if you’re not actually implementing AI in your business today, if you don’t have an agent managing your PPC campaigns, you’re late to the party, etc., etc. there’s a lot of fear. And then what ultimately happens is there’s a lot of entrepreneurs that because there’s so much fear and anxiety around it and feel like they’re already behind. They just stay stuck and they’re just kind of like frozen because nobody’s providing actionable content regarding like, here are the actual practical use cases of AI. Yes, there are some incredible features with Claude and integrating it to your email system, right. And being able to monitor your emails for you. Yes, there are some incredible ways to use ChatGPT and the new images that it’s able to produce, right? Like, there’s a lot of good things that are happening that way, but a lot of times the practical use cases where actually maximizes value in the business gets left to the side, or nobody’s actually addressing them.

Josh Hadley 00:02:00  So that’s what I wanted to do today, is actually provide you with practical use cases that if you’re an e-commerce brand, you can go replicate these exact same frameworks and implement AI in your own brand. So here we go. I’m going to start with this. First and foremost, my team of 35 people. Here is one thing that I have charged them with to do every single week. And that is number one. They are going to create custom GPT in ChatGPT every single week. That is their number one thing, and they have to report back every single week. There are two different ways that my team members have to report back how they’ve used a custom GPT over the past week. So number one, they can either create a brand new custom GPT to solve a very specific problem that they’re working on in the business. Or number two, they need to tell me what are some of the enhancements that you made to an existing custom GPT, and how are you utilizing that today? Now, the reason why this is so important and again, is this one of the shiny bells and whistles of I’m not using cloud code, I’m not using an open claw or anything like that.

Josh Hadley 00:03:06  Those are the the fancy things that everybody likes to be talking about right now with AI. But it’s like I’m trying to get to the basics with my team of 35, just trying to get really good repetition of like, hey, here’s the power of AI, and I want you guys to start automating many of the things that you used to take a few hours every single week. Doing AI can now do it in a matter of minutes. That’s the mindset shift that I’m trying to get all 35 of my team members to experience. So number one, that is the use case for AI in our business is I’m trying to get leverage of not just me having to think of like, hey, here’s all of the SOPs or here’s all of the processes we follow. I’m going to go start building and custom coding all of these agents to go take care of these tasks. No, I’m going to weaponize my 35 team members to go and build it themselves at the same time, because I know that for me, as a leader, I get more leverage from my team than I do from me being the wicked smart guy that knows how to hack his way through the jungle.

Josh Hadley 00:04:05  All right, so number two, the second way that we are utilizing AI in our business right now is we have now set up there is an agent that is set up through cloud code that is tied into the API with TikTok. And in the back end it is optimizing the title and the description and the main images on a weekly basis. So the reason why I’m starting here on TikTok shop, rather than on Amazon is because Amazon is number one, quite noisy. Number two, Amazon is our biggest sales channel. So if I start making changes over there, there’s more damage that can be done. If it doesn’t turn out the way that I ultimately want it to turn out. And so what’s happening is on a weekly basis, we’re going to change the title. And then this agent is going to analyze the data from TikTok shop, because it’s all being fed in from the API. If our click through rate or the conversion rate has increased since making the change to that title. You know, it would kind of count it as a win.

Josh Hadley 00:05:06  And so every week it is testing new title, new title, new title. And then we’re going to go through a series of tweaks where it’s going to test new main image, new main image, new main image, etc. and we’re going to do that on repeat until we have statistical significance that the changes that we are making are actually driving performance. And again, we’re only testing one lever at a time. You can’t test, you know, the description and the title and the main image and say, well what happened to my click through rate or what happened to my conversion rate? Because at that point you’ve changed three different levers. Now you have no idea which lever actually impacted the conversion rate. So it’s never going to be a clear cut case, because obviously there could be a time where a product goes viral with a video that week. And obviously at that point the conversion rate would spike. But I’m just talking like when it is kind of status quo. That’s what we’ve trained the agent to look for is just kind of like average status quo week over week.

Josh Hadley 00:05:58  This is the number of views that we’re getting in sessions right now. And so what we’re doing right now is this is like our guinea pig test to say, if we can make this work over here on TikTok, then we’re going to be able to really scale this up on Amazon in a really powerful and meaningful way. But here’s the key thing. We’re not just testing this over like a two week time period and be like, yep, that works. We’re going to be testing this over a period of six months. I have one dedicated team member that is responsible for overseeing the success of this process, and so he’s constantly looking at the data that came in and retraining that agent over and over and over again. So that’s the second way that we’re utilizing AI in the business. But I want to add one additional caveat to this. And this is something that I think way too many people overlook with AI, and that is this. They think that AI is this magical unknown, you know, being or something like that, that just knows everything.

Josh Hadley 00:06:53  And it’s going to do everything flawlessly. Well, quite the contrary. The way I actually view AI is nothing more than the way I view a new hire to the team. So ask yourself this question when you first hire somebody new to your team, does that team member know everything about your business, and are they able to pull all of the reports, analyze them correctly, draw the right business conclusions correctly every single time right out of the gate? The answer is no. And so likewise, AI is in the exact same boat. Now the advantage with AI is like you don’t have to worry about somebody’s dog dying or somebody being sick, or any of the things that you have to deal with when you’re dealing with human beings. But what you do get from AI is something that will train itself faster and faster and faster over and over again. So just like the amount of time that I invest in kind of getting my new hires up to speed, which, by the way, if you’re wondering, like Josh, what is your onboarding process? I just recorded a couple of weeks back about our 30 6090 day onboarding framework.

Josh Hadley 00:07:52  Like go check out that video. It walks you through in immense detail every step we take for a new hire in their first 90 days. And that’s why we get team members scaled up faster in the first 90 days. Then most corporate companies get a team member scaled up in their first year. So the same principle holds true with AI agents. You need to train them and invest meaningful amounts of time right out of the gate to make sure that you’re putting in the proper guardrails. They’re actually executing the tasks the way you would want them to, and they’re ultimately drawing the business conclusions that you would have been making at the same token, but hopefully you get them trained well enough that they can execute without you. So we’re running this for six months. We’re babysitting this process for six months before we would ever feel confident to say, yep, let’s go replicate this over here on Amazon. Let’s talk about the third way that we are implementing AI into our e-commerce business, and that is through our hiring funnels and our case studies.

Josh Hadley 00:08:46  So many of you are familiar with my hiring funnels. And my approach to any type of hiring is I want to get at least a thousand applicants for every single position. Then I’m going to have many of those applicants complete a case study. And this is going to be a case study where they actually have to analyze a specific problem going on in the business, and then they have to record a five minute loom video, put together a presentation or a one page document with like, hey, how would you solve this problem in the business? Okay. Now in the past, I would have to go through and review every single one of these videos. And ultimately I was looking for communication and I had like a scoring rubric that was like, hey, if they show their video at the bottom of the loom video like I instructed them, plus two points and duo is their English like perfect? Is it flawless? Could I put them on two x speed and still be able to get everything that they’re they’re saying really well okay then that’s a five.

Josh Hadley 00:09:42  That’s five points. So on and so forth. So I had a bunch of like the scoring rubrics that I would give. Well guess what I’ve done through cloud code. You can now set up like, hey, here’s my scoring rubric, here’s what I’m looking for for this role. I’ve already drafted the role profile. Here’s the case study. Here are the answers from my perspective about this case study. And here’s the way I would like to score these candidates. And guess what. Now what happens is as soon as somebody submits a case study that will get pulled in by Claude, and then it will analyze the loom video, okay. And then it’s going to basically score every single case study for me, and it’s going to assign a score and put it on a spreadsheet. And so now my responsibility is once a week I will come take a look at that spreadsheet, and it will show me every single one of the people, the applicant that has taken the case study and completed it in what their score is.

Josh Hadley 00:10:34  And so now for me, instead of having to analyze these, which would take hours at a time, I can now go in there and say, oh, here are the top five. Let me go watch their videos. I watch their videos, see who I want to bring on for an interview. So this has saved me at least five hours a week in work. And to be honest with you, it’s a lot more accurate than I was, because let’s let’s face it, if I’ve sat there for one hour by the and I’ve just been watching case study after case study after case study, by the time I’m reviewing that last person, like my brain’s already fried, I’m already kind of tired with this process, I’m ready to move on to something else. And so I’m either going to judge that I’m either going to score that case, study overly harshly, or I’m going to be more liberal just because I’m like, yeah, let’s just be done with this thing. Five you know? But in reality, maybe I shouldn’t have been that way.

Josh Hadley 00:11:25  So the nice part is AI doesn’t get tired and they’ll continue to repeat what you want them to repeat for you. So that is the third way that we are utilizing AI in the business. Let’s talk about number four. Number four we are creating custom GPT that are integrated with Alexa okay. So Amazon Alexa kind of replaced Amazon Rufus. And what’s happening is like Alexa has all of these questions that you can ask about a certain product. And so we’re feeding we’re basically scraping that on an ongoing basis to understand what are the questions being asked. And then there is a custom GPT that will help you update your product, copy your title, your listings based on the questions that people are asking. Therefore, you get ranked and served more quickly into the Alexa recommendation. So that’s the fourth way that we’re using it. Number five, all of our company meetings are being recorded. Okay. So we’re going to be using fireflies. That’s what we use to record all of the company meetings. And what happens is this has a direct kind of API integration into Claude.

Josh Hadley 00:12:31  And so guess what. As soon as one meeting is done, it gets fed into our bigger kind of Google Drive, where it’s storing all of these videos and these recordings. It takes the transcript. And then Claude is going to analyze that transcript. And this is happening not only the meetings that I’m having, but this is the meetings that are happening throughout the entire organization, with or without me. And so guess what happens? Claude is getting all of this massive amount of context about what is going on in the business. So I can easily go to Claude and say like, hey, based off of the last month of videos and the last month of meetings that the team has had, what’s the real constraint? Where is my team stuck? What do I need to do as a leader to move the business forward? And this allows me to, number one, better understand the constraints of the business. Number two, understand like where are people getting stuck in the business and being able to make decisions that I can go coach them on for leadership development.

Josh Hadley 00:13:29  And ultimately, this provides a great pulse on the business that I never had before. And I shared this previously, but I start every week with a 15 minute kind of CEO update. I’m sharing all of the wins that occurred over the past week in that loom video. And then I am also reiterating what the constraints of the business is and the issues that we are facing as a business, and that we need everybody to help, like it’s all hands on deck to solve the number one constraint in the business. Well, having all of the context of all of the recordings and meetings happening in the business, getting those transcripts, extremely valuable context so that I can go direct the rest of the team. So the sixth practical use case for us in our business in utilizing AI is pushing out more ad creative on an ongoing basis with all of our content for meta ads, even some new TikTok ads that we want to be running, we are scaling very quickly and heavily with just coming out with more AI video content.

Josh Hadley 00:14:27  And so this is to a support TikTok growth and also the growth of Shopify from all of the ads that we’re putting out onto meta and where we’re directing that traffic. So that’s an easy use case. Everybody knows how powerful AI can be with video generation, but we’re using it like, Hey Jen, we’re utilizing Vo three. So many of these new softwares to be able to produce those videos at scale on an ongoing basis. The seventh and final practical use case of how we’re implementing AI in our business is to utilize AI to clone the judgment, skills and decision making from not only myself, but the top leaders in the organization. So if you go take a look at the episode I published just a couple of weeks ago, it was all about how to clone yourself utilizing AI. And basically the process is this you’re going to go do the thing, whatever you need to create a process around. Maybe it’s you want to be able to offload PPC management off of your plate. Maybe it’s R&D, you know, creating and finding new products for the business.

Josh Hadley 00:15:24  You want to get that off your plate. It doesn’t matter what the process is, whatever you want to get off your plate, you have to go do the thing over and over and over again. I recommend at least over the course of four weeks, record that process using loom. Narrate it out loud, and at that point you have four different videos. You upload the transcript into AI. AI is going to spit out some in SOP structure for you, and then it’s going to interrogate you with future questions that then allows you to like, create a more refined SOP. But it’s not an SOP. The SOP is there for like basic steps, but it’s helping replicate your judgment and decision making. That, at the end of the day, is why team members, when they come to you and they’re like, hey boss, what do you want to do here? It’s because you have not been able to, like, distill your judgment abilities upon that team member. And so that’s what I’m utilizing AI for, is to say, hey, here’s the SOP, and here is, here’s like all of the decision frameworks that goes along with R&D or PPC management.

Josh Hadley 00:16:24  And normally you don’t get that with the standard SOP. And then finally, last but not least, we’re integrating that entire SOP and decision making database into notion. And then notion has an API into Claude. So that and ChatGPT so that when my team members are executing let’s call it PPC management, they can go ask, chat and say, hey, I’m having an issue with this ad. I’m not getting enough impressions. What should I do? And then it’s firing up all of these processes, the decision frameworks that have already been extracted from my mind so that it’s as though I am there, present, and it’s providing that direction to the team member without that team member needing to send me an email and say, hey boss, here’s the problem. What do you want to do? That is the power of how AI can ten x the ability to make decisions and move faster in your business. So those are the seven practical ways that we’re implementing it into our business. I would love to hear your thoughts, so leave some comments down below.

Josh Hadley 00:17:21  Tell me what one you know impressed you the most. And maybe there’s some things that you’re doing that we’re not doing. And I would love to hear how you are implementing AI into your own e-commerce brand. With that, I look forward to sharing more details with you next week. But if you do know of an operator that needs to hear this message, please do me the favor of reaching out to them, sending them a copy of this episode, and letting them know how game changing it can be. So until next time, have a great day!