July 22, 2025

071: Tactic Tuesdays: AI for Amazon: Hype vs. Reality

The line between AI hype and reality is blurring for Amazon sellers. What's actually working right now, and what still falls short of the promises? We're witnessing dramatic improvements in AI image generation tools over the past 6-12 months. ChatGPT and Gemini can now create compelling product images for certain categories and make sophisticated edits that previously required professional Photoshop skills. Need to change a background, adjust lighting, or swap in different models using your ...

The line between AI hype and reality is blurring for Amazon sellers. What's actually working right now, and what still falls short of the promises?

We're witnessing dramatic improvements in AI image generation tools over the past 6-12 months. ChatGPT and Gemini can now create compelling product images for certain categories and make sophisticated edits that previously required professional Photoshop skills. Need to change a background, adjust lighting, or swap in different models using your product? AI handles these tasks remarkably well, creating natural shadows and maintaining proper perspective automatically.

The most striking revelation? Amazon products with obviously AI-generated images are selling in massive volumes with excellent reviews. While photorealistic perfection might be ideal, consumers are demonstrating that "good enough" AI imagery doesn't hinder purchasing decisions for many product categories.

Where AI delivers immediate ROI is rapid A/B testing. You can now create multiple image variations in minutes rather than hours, testing which demographics, backgrounds, or visual elements drive the highest click-through and conversion rates. For brands willing to experiment, these incremental improvements compound into significant sales growth.

Custom GPTs represent another breakthrough, with Amazon sellers building specialized AI assistants for listing optimization, keyword research, and content generation. These tools require minimal technical expertise yet dramatically streamline workflows that previously consumed hours of time.

Not everything lives up to the hype yet. Fully automated Amazon advertising remains problematic – while AI excels at bid optimization, it often goes "nuts" with keyword harvesting without human oversight. The technology is advancing rapidly, though, with today's "impossible" demonstrations likely becoming standard practice within 3-6 months.

Ready to leverage AI for your Amazon business? Start with image testing, explore custom GPTs, and remember that for any business challenge you face, someone has probably already built an AI solution that can help. The future of Amazon selling is arriving faster than most realize – are you prepared?

🚀 Transform your brand on Amazon by building a powerful customer list with the After Purchase Funnel Blueprint course. Click here to get the full course for free.

➡️ Ready to go deeper into your Amazon FBA journey to accelerate your success? Get your hands on ALL of the Brand Fortress HQ resources, mentorship, and knowledge base by visiting us at BrandFortressHQ.com

⭐️ Want to help our show grow so we can continue bringing you the very best of guests and actionable content for your Amazon FBA business? We'd greatly appreciate if you took two minutes to give us a five star rating and review. Thank you!

00:00 - Introduction to AI for Amazon

03:16 - AI Image Generation Progress

10:49 - Photo Shoots vs. AI Imagery

15:16 - Repurposing Imagery with AI Tools

22:58 - The Value of A/B Testing with AI

28:26 - Customized GPTs for Amazon

32:16 - Separating Hype from Reality

36:00 - Final Advice for Amazon Sellers

WEBVTT

00:00:01.784 --> 00:00:03.907
Welcome everybody to another Tactics Tuesday.

00:00:03.907 --> 00:00:09.368
Today, what we're going to talk about is AI for Amazon and specifically talking about the hype versus reality.

00:00:09.368 --> 00:00:24.361
I think we bring some very unique perspectives to this, both from the brand perspective and then also from the agency perspective of how much is AI really moving the needle on Amazon, what does that look like and what things out?

00:00:24.361 --> 00:00:37.323
There are still very much hype when it comes to AI for Amazon sellers, and so with that, I'll actually turn it over to you, matt, because I know you spend quite a bit of time immersed in this world.

00:00:37.323 --> 00:00:42.014
As far as connecting AI and Amazon, what do you see on the field right now?

00:00:43.761 --> 00:00:53.210
In terms of AI for Amazon, I think we're talking about images specifically, and so I've seen a lot of really really.

00:00:53.210 --> 00:01:16.149
Now, chatgpt just updated their image generation tool, which for me, in my opinion like I didn't play around with mid-journey all that much as anywhere near as much as I know that Michael did but for me, the Chat, chat GPT image generation tool, when they upgraded it about a month ago, is night and day better than it was before, and I did a lot of extensive playing around with it.

00:01:16.149 --> 00:01:22.287
Now, for some reason, ai doesn't understand certain products, like pool cleaning tools, for example.

00:01:22.287 --> 00:01:28.900
It just does not understand what a pool brush is used for or how it's used, no matter how many times you ask it to change.

00:01:28.900 --> 00:01:34.813
And that's still the case even with the upgraded version of ChatGPT's image generation tool.

00:01:35.060 --> 00:01:37.147
However, there are other types of products.

00:01:37.147 --> 00:01:41.551
A girl in my mastermind sells flower seeds, for example.

00:01:41.551 --> 00:01:50.754
Flower seeds, for example and the images that she's been able to make with the chat GPT image generation tool, but also mid journey, are fantastic.

00:01:50.754 --> 00:01:53.424
I mean, better than what she was getting from her graphic designer before.

00:01:53.424 --> 00:02:03.173
So there's still a lot of really really good use from these image generation tools for certain products, but none that we work with directly.

00:02:03.173 --> 00:02:16.868
Unfortunately, it just doesn't understand what the products are being used for, what the use case is, and none of the products really fit well within just put this product in a different scenery, which is, I think the image generation tools are really really good at that.

00:02:16.868 --> 00:02:28.913
But yeah, so for the products that I work with directly, it's just it's not good enough yet to understand what the use case of the tools are, so I haven't been able to really use it in that regard, at least for the brands that I'm involved in.

00:02:30.060 --> 00:02:37.775
Well, to kind of summarize that, at least from what I'm seeing, is, first of all, the tools have gotten a lot better.

00:02:37.775 --> 00:02:42.320
So, if you haven't used them in the last six to 12 months, they've gotten dramatically better.

00:02:42.320 --> 00:03:06.366
When we talk about, specifically, ai generation, and what I would say out of that is, is that kind of like what you were talking about, matt, which is, if you look at doing things like changing out backgrounds and kind of things in the backdrop, if you have, you know, a photo of your product and you want to change out kind of that background, it does work well for that.

00:03:06.366 --> 00:03:15.171
Um, however, with that said, if you're looking at people or you're looking at complicated products, those types of things, um, there's still some significant challenges there.

00:03:15.171 --> 00:03:52.473
Um, that said, I do think that there's a great opportunity for, you know, even if you already kind of have your set images, you know, um, you know, we were just working with a client the other day where we were looking at, uh, a father's day promotion, and so, you know, that's a perfect example where I think, um, you know, there's an opportunity for chat, gpt, where the the main photos and we've got some um resources already built there that then we can use something you know, chachi Batty or another AI image generator in order to create those kind of, you know, promo specific images that can make a lot of sense.

00:03:53.901 --> 00:03:56.614
But I do think that we're still too early for the days of.

00:03:56.614 --> 00:04:00.948
I know I've already gotten this question from a few clients of like, hey, do I need a photo shoot anymore?

00:04:00.948 --> 00:04:03.768
Can't I just, you know, whip it up in AI?

00:04:03.768 --> 00:04:06.562
And it's like, well, we're not quite there for most products yet.

00:04:06.562 --> 00:04:10.652
You're gonna have to do some sort of, you know, photo or video shoot.

00:04:10.652 --> 00:04:19.351
Now the nice thing is is I think that you can extend those resources, once you have them, from a decent photo photo shoot to do a lot more with it than what you could do in the past.

00:04:19.351 --> 00:04:29.259
So think about it more from a you know, getting a lot more bang for your buck as opposed to completely replacing any sort of photo or video shoot.

00:04:29.259 --> 00:04:31.689
So that's what I'm seeing from it, especially on the image side.

00:04:31.689 --> 00:04:36.291
You know, mike, I'd love to hear you know what you're seeing from the brand owner perspective.

00:04:37.661 --> 00:04:38.745
Yeah, I think a few things.

00:04:38.745 --> 00:04:45.473
One thing that I would say is I do think that we are moving into a space like you just said.

00:04:45.473 --> 00:04:50.245
You're not going to be able to skip the photo shoot or the video shoot, right?

00:04:50.245 --> 00:04:55.846
I think that for some products, that might not be true.

00:04:55.846 --> 00:05:02.430
I think there are products that you probably could skip the photo shoot.

00:05:02.430 --> 00:05:12.149
You might not be able to as easily skip the video shoot, but, by the same token, you could use UGC for that.

00:05:12.149 --> 00:05:18.973
You could use influencers for video stuff potentially, as opposed to necessarily a video shoot.

00:05:18.973 --> 00:05:21.362
Now, it depends on what kind of marketing you're trying to do.

00:05:21.362 --> 00:05:22.887
It depends on where the videos are going.

00:05:22.887 --> 00:05:33.711
Maybe you still need the video shoot, but when I say that you might be able to skip the photo shoot, I think that we've gotten some really good renders done of our products.

00:05:35.434 --> 00:05:39.408
Now, again, as Matt said, our products are not really AI friendly.

00:05:39.408 --> 00:05:42.322
I can't just give it a pool net and say this is our pool net.

00:05:42.322 --> 00:05:43.122
You know, put it into this.

00:05:43.122 --> 00:05:45.687
You know it doesn't do that, it doesn't know what to do with it.

00:05:45.687 --> 00:05:56.802
But, that being said, there are a lot of products that AI can do that with, can really take, you know, especially you know like supplements.

00:05:56.802 --> 00:06:00.531
You know like a supplement container or a drink container or a you know.

00:06:00.531 --> 00:06:08.045
You know kitchen gadgets or things like that, where it kind of understands the scene, it kind of understands how people use it.

00:06:08.045 --> 00:06:36.370
You know it can place it in the kitchen with the mom and the kid it's getting really good at that kind of thing and I think, especially for lifestyle images that are more about creating an emotion than they are about showing a specific feature of a product or something like that, I think you definitely can do a lot of that with AI right now for some products, and if you do some renderings which are way faster and way cheaper than a photo shoot, a lot of those renderings can then be turned into AI imagery.

00:06:36.370 --> 00:06:38.567
So I think you might be able to skip the photo shoot for some things.

00:06:38.567 --> 00:06:54.072
We definitely can't, Um, but as you said, what AI is doing for us is is providing a much easier way for us to repurpose the digital content that we already have.

00:06:54.072 --> 00:07:05.855
So if you're a brand that already has some relatively decent imagery but you're looking to update it, I don't think you necessarily need a new photo or video shoot.

00:07:06.437 --> 00:07:20.166
You may be able to use AI to revamp those images, to give them better lighting, to make them look, you know, change the positioning or something to change out the person Like Gemini.

00:07:20.166 --> 00:07:27.536
Actually, their newest version for image generation and editing is actually quite good.

00:07:27.536 --> 00:07:31.406
You can feed Gemini an image.

00:07:31.406 --> 00:07:37.146
Now I think you might have to be on a specific paid plan to utilize this feature.

00:07:37.146 --> 00:07:40.480
I'm not positive, I'm on it so it works for me, but for some people.

00:07:40.480 --> 00:07:52.403
But you can feed it an image and change out a very specific component of that image and it doesn't change anything else, it just changes that and it actually does quite a good job at that.

00:07:52.403 --> 00:08:07.209
So, and it's not just background replacement, although you can do that, it's also I've taken images and put them in there of a particular person using our product and I basically said, hey, make them older or make it a woman or make you whatever.

00:08:07.209 --> 00:08:09.173
It actually does quite a good job at that.

00:08:15.319 --> 00:08:18.468
I think for some products you might even be able to change the positioning or how they're holding the item or how they're using the item.

00:08:18.468 --> 00:08:20.213
You know, change the background and the person.

00:08:20.213 --> 00:08:35.043
So I do think we are moving into a space right now where, even today, you can definitely use ai to repurpose a lot of imagery that you already have and make some of those edits without having to be a photoshop expert.

00:08:35.043 --> 00:08:36.609
You know to be able to do that and make it look right.

00:08:36.609 --> 00:09:00.644
Because one of the nice things about ai this is the thing that I think this is probably the biggest thing that makes ai useful for image editing is because a lot of times if you try to change out a person or change the positioning or add an item to an image or something, the biggest thing that's difficult if you're not an expert in Photoshop is getting the lighting and the shadow right.

00:09:00.644 --> 00:09:13.604
It's really different, you know difficult getting the perspective right on that object and having it be able to fit into the image properly so that the eye doesn't catch it and be like, yeah, that something's not right there.

00:09:13.604 --> 00:09:17.493
Ai has actually become really good at doing that.

00:09:17.493 --> 00:09:27.221
You know where it can insert something into an image and give it the proper lighting, shadowing and perspective so that you can't really tell it's been inserted.

00:09:27.221 --> 00:09:31.312
It looks like it was there all the time, and so that I think is really useful.

00:09:31.740 --> 00:09:38.524
The other thing that I want to add is just a practical component, and that is if you look at Amazon.

00:09:38.524 --> 00:09:58.048
There are a lot of products that are selling massive volume, have really good reviews and are doing exceptionally well on the platform, and if you look at their images, it is obvious that they are renders or AI generated.

00:09:58.048 --> 00:10:05.390
You know like you can tell, but people are still buying those products and still giving them good reviews and they're still getting good sales volumes.

00:10:05.390 --> 00:10:11.894
So don't get super wrapped up in making sure that your image is perfect.

00:10:11.894 --> 00:10:17.533
I do think you know it'd be nice, you know like, if you can create an image where somebody can't tell, I think that's better.

00:10:17.533 --> 00:10:30.073
I don't, however, think that it's absolutely necessary, especially in certain product categories, that it's absolutely necessary, especially in certain product categories, if most of your competitors are, you know, are using images, that it's obvious, it's AI.

00:10:30.073 --> 00:10:37.783
If you could just do a better version that maybe it's still somewhat, you know, noticeable, but it's not as obvious, that might be enough in your category.

00:10:39.166 --> 00:10:46.029
Yeah, well, and I think you know, if we set aside the words AI, you know we could have said the same before AI came along.

00:10:46.029 --> 00:11:01.605
You could say the same thing about, you know, I think we've all seen them over the years of products that do really well, even though their images are ugly or what we would consider ugly, and I mean we've had that happen to where you know we've had tested.

00:11:01.605 --> 00:11:07.577
You know what we, what is a prettier image versus an uglier one?

00:11:07.577 --> 00:11:16.913
And sometimes you know the not so perfect image wins when we start looking at clicks and conversion rates and really driving sales and things that matter.

00:11:16.913 --> 00:11:19.059
So, you know, I think that's such a great point, mike.

00:11:19.059 --> 00:11:32.301
The other thing that I would add, you know, to what you said that I think is really important is you know how you can adjust those images, because what I feel like maybe doesn't get enough attention in this conversation is you know we talk about going from hey, do I need a photo shoot?

00:11:32.301 --> 00:11:33.423
Do I not need a photo shoot?

00:11:33.423 --> 00:11:40.392
Those types of things.

00:11:40.392 --> 00:11:43.038
I think the big win really in the next year is going to come from being able to rapidly A-B test.

00:11:43.038 --> 00:11:56.893
You know, main images, secondary image, stacks, video, all these different things to really, you know, dial in the click through rate and conversion rate on your listings with a heck of a lot less resources than you needed before.

00:11:56.974 --> 00:12:03.778
Because, if you like you said, if you can change out that image to where you're like, hey, I have the same image and I'm able to make this person look older.

00:12:03.778 --> 00:12:06.013
Or, you know, make it a woman instead of a man.

00:12:06.013 --> 00:12:06.975
You know all those kinds of changes.

00:12:06.975 --> 00:12:07.216
Now you can.

00:12:07.216 --> 00:12:08.099
You know, make it a woman instead of a man.

00:12:08.099 --> 00:12:10.245
You know all those kinds of changes.

00:12:10.245 --> 00:12:10.966
Now you can.

00:12:10.966 --> 00:12:27.811
You know, really A B, test things a lot faster and with a lot less resource, because you don't have to do a whole new photo shoot or have a graphic designer spend a couple of hours, you know, in Photoshop cutting out that person and adding another person in and all those types of things.

00:12:27.811 --> 00:12:38.457
You can put it into, you know, gemini or chat, GPT or something like that, and have something back in you know 30 seconds and then be able to go test it.

00:12:38.457 --> 00:12:42.792
So I think there's a lot of value that comes from that testing aspect as well.

00:12:43.452 --> 00:12:47.499
Yeah, my Photoshop skills are miserable.

00:12:47.499 --> 00:12:59.270
My stick figures are unrecognizable, but what the new version of ChatGPT's image generation tool has allowed me to do is iterate, like you said, a whole lot faster, john.

00:12:59.270 --> 00:13:03.241
For example, we have a client that sells hangers, and there's different types of hangers.

00:13:03.241 --> 00:13:04.594
One of them has a bar with clips on the bottom of it.

00:13:04.594 --> 00:13:06.458
I needed another image, a duplicate image of the hanger that didn't have those clips.

00:13:06.458 --> 00:13:11.254
So instead of it, I needed another image, a duplicate image, of the hanger that didn't have those clips.

00:13:11.293 --> 00:13:18.380
So, instead of going back and forth with my graphic designer and waiting a couple of hours for the revision, I just uploaded the picture inside of chat GPT.

00:13:18.380 --> 00:13:19.432
I asked it.

00:13:19.432 --> 00:13:20.399
No, this was just a test.

00:13:20.399 --> 00:13:21.427
I didn't know if this was going to work.

00:13:21.427 --> 00:13:23.272
As a matter of fact, I didn't think it was going to work.

00:13:23.272 --> 00:13:27.957
I just asked it to remove the roll bar with the clips, and it did a 90%.

00:13:28.097 --> 00:13:30.000
It did a way better job than I would have been able to do.

00:13:30.000 --> 00:13:44.062
It cut out the corners of it, and I think that was just something that, if I would have was better at prompting for images, I think I could have resolved that, but 10 seconds later I had an image that was good enough for me to at least split test where I was using it for.

00:13:44.062 --> 00:13:56.774
So and that's probably saved me not four hours of work but four hours of communication back and forth with a graphic designer, and I was able to get something good enough for testing in a span of about 15 or 20 seconds.

00:13:56.774 --> 00:14:04.150
And this is coming from someone who doesn't know the first thing about Photoshop and actually gets very frustrated because I can't get things to look how I want it to.

00:14:04.150 --> 00:14:16.552
So for me, it's made me a thousand times faster in just iterating and getting something that I can split test to see if it converts better than what we currently have, so way, way faster than the how I used to do it.

00:14:17.333 --> 00:14:30.322
Well, it's also important, I think to to to recognize that because of that, if you have anybody on your team who is even marginally good with Canva or GIMP or.

00:14:30.383 --> 00:14:51.702
Photoshop or something like that, like just basic skill, then that person becomes an individual who can do a lot more for you in that space, now that they have AI to do most of the heavy lifting and they can just do a little bit of editing after the fact to repair whatever little artifacts there might be, or something.

00:14:51.702 --> 00:14:55.480
That's way easier than them having to do all the work themselves.

00:14:55.480 --> 00:14:58.758
And so, again, like you said, you don't need that.

00:14:58.758 --> 00:15:02.749
You know super high-priced editor, you just need somebody who needs some basic skills.

00:15:02.749 --> 00:15:07.134
Either you learn some of those basic skills If you're a one-person operation.

00:15:07.815 --> 00:15:11.052
Take a few days to learn how to use Canva.

00:15:11.052 --> 00:15:13.277
Well, canva is pretty powerful.

00:15:13.277 --> 00:15:19.418
It's not Adobe Photoshop, but it's pretty powerful and it's got some really sweet tools in there that I'm not even sure that Adobe has.

00:15:19.418 --> 00:15:29.344
And so if you just took a few days and got to know how to use Canva, that, combined with some of the other AI tools out there, make you a super ninja.

00:15:29.344 --> 00:15:31.592
You know, in terms of being able to create.

00:15:31.592 --> 00:15:34.041
You know graphic content and make adjustments.

00:15:34.041 --> 00:15:34.783
You know here and there.

00:15:36.830 --> 00:16:12.153
Yeah, and I think you know, just kind of wrap up the image, ai, image conversation, because I think this brought something to my mind, based on what you were talking about, matt, which was, you know, we had a client that accidentally, you know, for one of their family of products accidentally deleted the listing, and then they also deleted kind of the backup files that they had for those images, and so we basically had to recreate those images by looking at one of their other product families that was similar, and, I mean, the only reason we were really able to do that was because of AI.

00:16:12.777 --> 00:16:58.466
And so, you know, from that use case, I think there's also an opportunity to look at your competitors, and you know, probably, I think you know, circling back to what you said, mike, that I just want to double click on, which is look at what your competitors are doing and if there's something that you're like, wow, I wish we had an image like that or we could do something like that, and then one step better, like you can do that so quickly with AI to take that image and say, okay, now add our product into it, add this one additional feature, benefit to it, and now you have, you know, an image that's even better than what your competitors have with, you know, less than two minutes worth of work, so I think there's a huge opportunity there for you know kind of adjusting images using a lot of the AI tools that are available now.

00:16:59.692 --> 00:17:00.235
One of the things.

00:17:00.235 --> 00:17:01.302
Well, go ahead, matt.

00:17:01.322 --> 00:17:03.192
Look like you're going to say something I was just going to say.

00:17:03.232 --> 00:17:26.842
Yeah, so I mean for me, like we do still have a graphic designer, there are still things that we use that graphic designer for, and, john, you mentioned the father's day image, as a matter of fact, for that client today, I knew what I wanted in my mind's eye the band, the badge to look like on the packaging, and so I just went into chat GPT, knowing that it wasn't going to be the final product, just that I was going to give something to kind of spark the imagination of our graphic designer.

00:17:34.150 --> 00:17:34.993
It took me three seconds Now, it didn't?

00:17:34.993 --> 00:17:36.681
It kind of it changed up a lot of the elements on the box to make it.

00:17:36.681 --> 00:17:39.029
It wasn't usable, but it allowed me to give, instead of having to type out on a Slack message, what I wanted to see.

00:17:39.029 --> 00:17:45.111
It allowed me to really, really quickly use chat GPT to give him an idea of what it was that I was looking for.

00:17:45.111 --> 00:17:57.358
So again, even though I was using the graphic designer to give me the finished product, it allowed me to show him what it was I was seeing in my mind's eye a whole lot faster than me explaining it to him.

00:17:57.378 --> 00:18:09.678
So the other thing that I wanted to say is you know, like we're talking a lot about, you know, using AI, with Amazon images, but we're talking a lot about generation of images and editing of images, which is important.

00:18:09.678 --> 00:18:10.913
I mean, like there's a lot of that.

00:18:10.913 --> 00:18:14.030
That goes on and I think it's valuable and just real.

00:18:14.030 --> 00:18:22.318
Briefly, before I move on to what I actually want to say, I do want to you know, just double click on that idea, john, that you were talking about in terms of sellers.

00:18:22.318 --> 00:18:25.962
And it might not be so much because they don't know.

00:18:25.962 --> 00:18:39.950
It might just be because in the past it's been so difficult to do and so expensive to do, because you'd have to pay a graphic designer and you know, go through all this hassle that it just hasn't happened.

00:18:39.950 --> 00:18:53.057
And that is the difference in click through on an image that has a guy versus a lady, or a younger guy versus an older guy, or an image that has a dog in it versus an image that doesn't have a dog in it.

00:18:53.057 --> 00:18:57.534
Like there are so many different small changes that you could make to an image.

00:18:57.534 --> 00:19:05.656
That for the avatar that you're after could instantly improve your click-through rate by 50% or double it.

00:19:05.656 --> 00:19:10.676
That are very simple changes that could be made with AI in you know a couple of minutes.

00:19:10.676 --> 00:19:49.241
That's the sort of area where I think there's a lot of power in this in terms of those iterations is thinking about and this ties into the next thing that I want to say, which is the other value piece with AI at this point is it now can understand images, and so if you give it an image and you tell the AI what the purpose of the image is and a little bit of background on who your customer avatar is, it can analyze that image and give you some ideas about how you might improve on it.

00:19:49.241 --> 00:20:00.902
And then you can take those ideas, figure out which of those are the easiest ones to implement with AI, make those changes and then do the split tests based on that information that AI is giving you.

00:20:01.250 --> 00:20:17.490
We're also moving into a space where, if you know what you're doing and you can create a custom GPT or maybe some sort of an N8N workflow that does this, where you know there's we've got tools like PicFu and IntelliV and things like that out there where you split test images beforehand.

00:20:17.490 --> 00:20:21.057
You can now start the process.

00:20:21.057 --> 00:20:37.436
There are ways to do this where you can split test two images with AI and basically tell it who's the avatar that you're aiming at and you know what are you trying to accomplish with the image, what's the benefit that you're trying to solidify, what's the emotion you're trying to generate.

00:20:37.436 --> 00:20:47.492
Whatever, Give it the two images and have it do its own poll and you would be amazed I think people listening would be amazed at.

00:20:47.492 --> 00:20:56.102
There are some models out there People are generating these models that almost all the time, the AI is picking the true winner.

00:20:56.102 --> 00:21:03.151
Like, if you run the split test on Amazon, the results from that split test match what the AI is telling you is the better image.

00:21:06.517 --> 00:21:10.243
Yeah, I think we're just kind of touching the surface on that.

00:21:10.243 --> 00:21:16.211
We've talked a lot about image generation, which I think is great, because I think that's where we've seen Well one.

00:21:16.211 --> 00:21:21.702
I think there's a lot of value in it for brands that are looking at increasing clicks and conversions and ultimately, sales.

00:21:21.702 --> 00:21:44.491
I think the other thing I want to kind of touch on as well, especially, mike, as you brought up some of the custom GPTs and that type of thing that people have put out there again, you know, I would I say this with a word of caution which is those GPTs are only as good as the people that have built them and they are, you know, static, meaning they're only as good as the time that they were, they were built.

00:21:44.491 --> 00:21:49.583
If we're talking about custom GPTs I know Mike you're working on, you know, say the name of it again.

00:21:50.829 --> 00:21:51.271
N8N.

00:21:51.271 --> 00:21:54.101
So, nancy, the number eight, nancy, n8n.

00:21:55.089 --> 00:22:08.205
Yeah, n8n, which is much more closer to you know, generative AI, where it learns as you go, and having, you know, access to what are essentially AI agents, I think is kind of the terminology that's being used for those right now.

00:22:08.205 --> 00:22:11.589
Agents, I think is kind of the terminology that's being used for those right now.

00:22:11.589 --> 00:22:17.162
But even with, you know, take a custom GPT, you know, I think, something that we have found really helpful and looking at.

00:22:17.162 --> 00:22:37.497
You know even ads and you know strategies of like, hey, we're looking at how do we boost sales for this brand and asking the AI a bunch of questions and, frankly, you know, for someone that has been doing this for five years, probably 80% of that, the ideas that it gives you, you're either like, oh, we're already doing that or that's not a good idea for this brand because of X.

00:22:38.258 --> 00:22:44.316
That said, there's another 10 to 20% that, like you know, just jog something in your memory.

00:22:44.316 --> 00:22:48.232
You're like, oh, yeah, we can run that type of campaign.

00:22:48.232 --> 00:22:56.482
Or you know this type of coupon or whatever happens to be that tool in the toolbox that helps just kind of prime the pump and help you.

00:22:56.482 --> 00:23:01.762
You know, think through ideas on how to boost sales, how to increase conversions, different.

00:23:01.762 --> 00:23:10.614
You know ad types if you're running your own PPC, but I think that's another area where there's a lot of value in having those conversations with.

00:23:10.614 --> 00:23:22.686
You know chat, gpt or some of these custom GPTs to help brand owners get some ideas and answers pretty quickly on the problems that they're facing, with some good answers.

00:23:24.290 --> 00:23:26.779
Don't even get me started with custom GPTs.

00:23:26.779 --> 00:23:41.637
I have become addicted to building them now and actually I've probably built two dozen of them in the past week and a half, two weeks and one of them actually talking about Amazon and optimization.

00:23:41.637 --> 00:23:43.340
I built one that I.

00:23:43.340 --> 00:23:46.633
First of all, I learned how to optimize a listing for Rufus.

00:23:46.633 --> 00:23:57.238
I went through a couple of deep research prompts and learned that, but then also use that deep research to create a custom GPT to help me rewrite listing copy for Rufus.

00:23:57.238 --> 00:24:05.121
You know, it probably took me about an hour or so to learn what I wanted to learn and use that learning to create the custom GPT.

00:24:05.121 --> 00:24:15.551
But now I've already done it for, you know, three or four client listings just input of the existing title gave it a little bit of information about the product and the avatar and it rewrites titles for me.

00:24:15.592 --> 00:24:17.576
So I mean custom GPTs.

00:24:17.576 --> 00:24:40.063
I have been using them for about six to eight months to make my life a lot easier, but they're so easy to create that you know, like I said, every day I think of another use case for another custom GPT and, to be honest, most of them are all involved around optimization on Amazon and I mean it's made my workflow so much easier using these chat GPTs.

00:24:40.063 --> 00:24:53.153
And now, after a couple of conversations with Michael, we're going to turn some of those custom GPTs into, add some automation to it using N8N, so they're going to be steroid on steroids here very, very soon.

00:24:53.153 --> 00:24:57.090
So there's so many custom GPTs already out there, but also they're not that hard to create.

00:24:57.090 --> 00:25:04.693
So if you have an idea for one or have a use case for one, learning how to create them are super easy and they're so so, so useful.

00:25:05.835 --> 00:25:40.200
The other thing that I think is important on the AI front and with regards to custom GPTs and a number of other aspects, is that AI I feel like one of the things that people tend not to recognize as a value component, let's say, is using AI to tell you what you don't know and what you need to know to move forward with something.

00:25:40.200 --> 00:25:44.917
So, in other words, reverse engineering a problem, let's say so.

00:25:44.917 --> 00:25:47.943
Let's take the situation that Matt's talking about.

00:25:47.943 --> 00:25:57.734
Let's say you're an individual who, number one, doesn't know a whole lot about how to optimize an Amazon listing.

00:25:57.734 --> 00:26:02.714
Number two, also, doesn't know a whole lot about how to build a custom GPT.

00:26:02.714 --> 00:26:15.824
You could step into ChatGPT and say I'm an Amazon seller and I want to update my listing and improve the click-through and conversion rates on my listing.

00:26:15.824 --> 00:26:16.566
What you know, how do I do it?

00:26:16.566 --> 00:26:18.675
What are the things I need to know that I don't know?

00:26:18.675 --> 00:26:19.637
What are you know like?

00:26:19.637 --> 00:26:26.215
Let the AI train you on what needs to happen in order for me to be able to fix that.

00:26:26.215 --> 00:26:27.258
What are the things I need to know?

00:26:27.258 --> 00:26:28.460
What are the things I would need to change?

00:26:28.460 --> 00:26:40.701
It will tell you those things and then you can back up again and say, okay, how do I learn that, or how do I do that, or how do I, whatever?

00:26:40.701 --> 00:26:46.739
So you just keep stepping back as far as you need to step until you reach the point where you're at the place where your knowledge foundation is enough to move forward.

00:26:46.739 --> 00:26:51.545
From there you just keep backing up until it reaches the point where you have the knowledge base you need.

00:26:51.545 --> 00:26:53.827
So you know again.

00:26:53.867 --> 00:26:55.430
Same thing with the custom GPT.

00:26:55.430 --> 00:27:07.769
Let's say you now, okay, the AI has told you these are the things that generally, you'd want to pay attention to in order to update a listing For your customer avatar and your product category.

00:27:07.769 --> 00:27:10.117
These are the things that you, things that you would want to do.

00:27:10.117 --> 00:27:11.480
Okay, how do I do that?

00:27:11.480 --> 00:27:12.521
All right, this is how I do that.

00:27:12.521 --> 00:27:14.925
Then, all right, now I want to build.

00:27:15.066 --> 00:27:22.189
Now you tell ChatGPT all right, I want to build a custom GPT that will help me do that.

00:27:22.189 --> 00:27:24.260
What would be the prompt?

00:27:24.260 --> 00:27:34.480
What are the instructions that I should put into that custom GPT so that it will do what you just told me are the right things to do in order to update a listing?

00:27:34.480 --> 00:27:36.164
It'll write a prompt for you.

00:27:36.164 --> 00:27:40.464
Then you can ask it how would we improve on that prompt and it'll write a new prompt.

00:27:40.464 --> 00:27:41.596
It'll improve on itself.

00:27:42.116 --> 00:27:45.767
What sort of documents should I upload for the knowledge base for this custom GPT?

00:27:45.767 --> 00:27:47.640
Where would I find that information?

00:27:47.640 --> 00:27:49.404
How would I put the documents together?

00:27:49.404 --> 00:27:56.380
Ami will walk you through the entire process of building that custom GPT.

00:27:56.380 --> 00:27:57.201
Where would I find that information?

00:27:57.201 --> 00:27:58.064
How would I put the documents together?

00:27:58.064 --> 00:28:00.170
Ai will walk you through the entire process of building that custom GPT if you don't know how to do it.

00:28:00.170 --> 00:28:00.972
So don't get intimidated by the process.

00:28:00.972 --> 00:28:03.618
You will learn a lot just by asking AI how do I do this?

00:28:03.618 --> 00:28:05.903
What do I need to know in order to do this?

00:28:05.903 --> 00:28:11.983
And just let it walk you back until you reach the point where you're like, okay, I know enough to move forward from here, and then just do it.

00:28:13.686 --> 00:28:18.363
So I think we spent quite a bit of time talking about and I feel like all three of us feel like that.

00:28:18.363 --> 00:28:28.846
You know, ai is definitely a big value add for brand owners that are, you know, building on Amazon and beyond.

00:28:29.306 --> 00:28:35.673
I do want to touch a little bit on, you know, what we feel like is the hype out there where AI gets overhyped.

00:28:35.673 --> 00:28:44.936
And I'll start with the first area that I feel like AI gets most overhyped is probably in the, the PPC or the ads realm.

00:28:44.936 --> 00:28:54.387
I still feel like there's this pipe dream out there where people like, well, I'm just going to put in my brand and my product and AI is going to magically do all the work and adjust all the bids and that type of stuff.

00:28:54.387 --> 00:28:59.680
Now, there's a lot of good machine learning and there's a lot of things that AI can do around.

00:28:59.680 --> 00:29:19.330
You know, analyzing data and doing a lot of that work for you comes down to actually, you know, steering campaigns and that type of stuff I've still seen, still feel like there's a lot of value of having the human in a loop rather than just kind of handing it off to AI when it comes to Amazon ads, at least as of right now.

00:29:21.155 --> 00:29:25.006
Yeah, I think AI can do bid optimization.

00:29:25.006 --> 00:29:41.588
Essentially, bid optimization is a math equation that computers are good at, but where I think I've used I've used a couple of tools for Amazon PPC management and I think for me where I would second guess using AI is to like keyword harvesting, for example.

00:29:41.588 --> 00:29:51.144
You know it does keyword harvesting, just based on rules, and if a keyword converted in a research campaign, it'll automatically create a manual campaign for that.

00:29:51.144 --> 00:29:53.340
And when I did that it just went nuts.

00:29:53.340 --> 00:29:56.657
It went absolutely nuts and added a whole bunch of irrelevant keywords.

00:29:56.657 --> 00:30:00.236
So I think having a human in the loop in terms of advertising is super important.

00:30:00.236 --> 00:30:09.137
I don't know if that's going to go away anytime soon, but bid optimization I think bid optimization is something that you can use some sort of a tool for and be pretty confident.

00:30:09.137 --> 00:30:13.840
But, yeah, human in the loop when it comes to advertising for sure, at least in the short term.

00:30:15.182 --> 00:30:31.688
I think the other one is well, in a general sense, I would say anytime you see, you know, a new hype video come out or things like that, you're seeing best case scenario of what that AI can do.

00:30:31.688 --> 00:30:39.359
And I can tell you from experience virtually every single video that I've seen where somebody was like look at this really cool thing.

00:30:39.359 --> 00:30:53.826
You know, whatever I tested it and if I was to pull back on the expectation by about 75, you know like maybe 70% of what they said was real, then yeah, no, it can do that.

00:30:53.826 --> 00:30:59.269
And if 70% of that is useful to you, then great, like that's perfect.

00:30:59.269 --> 00:31:08.183
But recognize that in almost every case the things that you're seeing people doing with AI, it's a very specific use case.

00:31:08.183 --> 00:31:15.445
It is not generalized well to other use cases, other products, other, you know like it's this thing.

00:31:15.445 --> 00:31:17.855
You know it was able to do that really, really well.

00:31:17.855 --> 00:31:30.162
Now, in three to six months, it's highly likely that whatever it was able to do in that very specific instance, it will be able to generalize and do better in a whole bunch of other instances.

00:31:30.162 --> 00:31:31.564
So I don't think it's.

00:31:31.564 --> 00:31:32.003
You know.

00:31:32.104 --> 00:31:47.758
I guess the one caution I would issue is don't ignore the hype, just recognize that whatever the hype is today is probably reality in six months.

00:31:47.758 --> 00:31:48.138
So be ready for it.

00:31:48.138 --> 00:31:49.740
Make sure you understand what that was that was able to do that.

00:31:49.740 --> 00:32:00.278
If you're looking at that thing and you say that's something that would really be valuable in our business, and take a step back and make sure that's true, that it's not just shiny object like oh wow, that's really cool.

00:32:00.278 --> 00:32:02.905
Just because it's cool doesn't mean it will be valuable in your business.

00:32:02.905 --> 00:32:04.777
So make that evaluation first.

00:32:04.777 --> 00:32:23.103
But if you believe it would be, if you think that it would pull some lever in your business that right now you're not able to effectively pull, then keep that in the back of your mind, write it on a notepad, put it in a sauna to come back and check it, you know, in three months or something, because there's a good chance in three to six months you will be able to apply that to whatever it is that you want to.

00:32:23.494 --> 00:32:29.381
Things are moving that fast, but today it's probably not going to generalize to what you want.

00:32:29.381 --> 00:32:33.243
So that's one of the things, and I think video is one of those.

00:32:33.243 --> 00:32:37.008
Right Like you can create some pretty impressive eight-second videos.

00:32:37.008 --> 00:32:41.311
Very few of us is an eight-second video going to be very useful to us.

00:32:41.311 --> 00:32:54.523
Now we might be able to piece together, you know, three or four or five eight-second clips with a couple of little clips that we already have of our own video and create a longer, you know video that maybe we could do something with.

00:32:54.964 --> 00:33:04.403
But also, again, you know, if you want the right person or if you want your product in it or you know things of that nature, video is not there and it won't be for a little while.

00:33:04.423 --> 00:33:24.066
So there are definitely some areas where the hype is, you know, unjustified in the moment, but you should be paying attention to it, and one of the things I would actually caution and suggest that people do is you need to create some sort of AI workflow agent, custom GPT task, whatever.

00:33:24.066 --> 00:33:27.438
There's a bazillion tools out there to do something like this.

00:33:27.438 --> 00:33:38.196
That is, keeping tabs on what is happening in AI out there that you might not know about because you don't have the time to be researching it 24 7.

00:33:38.196 --> 00:33:51.246
What are the new things that are happening that apply to you and making sure that you're getting a digest of those things and maybe a summarization of what it is and a link back to the reference so that you have opportunity to keep up on that thing.

00:33:51.246 --> 00:33:57.097
You know, whatever those things are, so you're ready for them when they you know, when they really are prime time.

00:33:58.541 --> 00:34:00.906
John, before you wrap up, I want to just real quick.

00:34:00.906 --> 00:34:07.909
Arif has been commenting and he had a question that Michael posed about split testing using AI.

00:34:07.909 --> 00:34:09.661
So I just want Arif to know.

00:34:09.661 --> 00:34:16.840
I'll have Michael go in and answer your question in the comments of this, so just want to make sure that we get that covered, because he's been asking a couple of questions in the chat.

00:34:17.021 --> 00:34:19.096
Perfect, yeah, yeah, and that's the advantage.

00:34:19.096 --> 00:34:31.132
So for anybody who's most people listen to this kind of after the fact that's the benefit of what we've started to do is at 12, 15 central time, around that time, post live on LinkedIn.

00:34:31.132 --> 00:34:42.161
So this is a good example of why to show up live and listen live on LinkedIn, because we will answer your questions live as a part of the podcast.

00:34:42.161 --> 00:34:42.903
So that's awesome.

00:34:42.903 --> 00:34:48.985
Thanks for bringing that in, matt, and I do think that this is a great place to kind of wrap for this episode.

00:34:48.985 --> 00:34:53.967
I think we've covered quite a bit of things that, again, I think are really helpful for brand owners.

00:34:54.315 --> 00:34:58.577
And, mike, I think you hit the nail on the head as far as, yes, a lot of things you're going to see out.

00:34:58.617 --> 00:35:12.586
There are best case scenario of how these AI tools are being used, but really they're a good indication of probably where things are going to be in 6 to 12 months, and if you can get at least 70% of that value out of it today, it's probably worth something spending some time on.

00:35:12.806 --> 00:35:24.481
And then what I would wrap up with for my pieces is that the one piece of advice that I'd have for listeners and brand owners are like hey, how do I get more involved in this AI and that type of stuff?

00:35:24.481 --> 00:35:54.460
Whatever problem you're facing in your business, take a little bit of time to look on YouTube Somebody's probably built an AI solution or a custom GPT or something like that for it and use an AI tool in order to solve that problem, and you'll probably find, a good portion of the time, that you can find a solution, or at least be pretty far down the path on a solution from AI tools or even a custom GPT that's completely free.

00:35:54.460 --> 00:35:56.244
That's already exists out there.

00:35:56.244 --> 00:36:08.579
So I would just, you know, really encourage brand owners to do a little bit of that research into how can I leverage an AI tool in order to solve whatever the problem is that you're facing in your brand right now.

00:36:08.579 --> 00:36:14.434
Solve whatever the problem is that you're facing in your brand right now.

00:36:18.085 --> 00:36:20.195
Any other last pieces of advice from you, matt or Mike.

00:36:20.195 --> 00:36:22.277
I'll just reiterate custom GPTs.

00:36:22.277 --> 00:36:30.784
If you're not familiar with them, if you haven't used them at all, go inside of your chat GPT right now and click on the little explore GPTs button, because there are literally thousands and thousands of them that exist.

00:36:30.784 --> 00:36:36.275
For any question, any specialty that you're looking for, there's already a custom GPT out there that exists.

00:36:36.275 --> 00:36:39.905
So if you haven't played around with them yet, definitely do that when you hear this.

00:36:40.775 --> 00:36:42.402
Yeah, and I would say, as you, test it.

00:36:42.402 --> 00:36:44.192
If you don't like it, test a different one.

00:36:44.192 --> 00:36:51.737
Oftentimes there's five or 10 different custom GPTs for what it is that you're trying to do, and if none of them are right, build your own.

00:36:51.737 --> 00:36:59.061
I mean, as Matt said, they're not as hard as you think, especially if you allow AI to help you reverse engineer the process.

00:36:59.061 --> 00:37:04.617
And then I would also just come back to make sure if you're not using AI for anything else.

00:37:04.617 --> 00:37:29.612
Use it for deep research it's really useful for that and use it at this point for image iteration and split testing, because, honestly, if you can take an image and in 30 seconds or a minute, create a new image that just changes out the person or the color of their shirt, or put a dog in the image or whatever, and you can test that, I guarantee you're going to find small incremental improvements in your CTR and CVR.

00:37:29.612 --> 00:37:34.840
And if you just keep doing that, that alone, if you did nothing else in your business, is going to improve your sales.

00:37:36.255 --> 00:37:38.945
Well, I think that's a great place to wrap for today's episode.

00:37:38.945 --> 00:37:44.844
Thank you everybody for listening and we look forward to seeing you next Tuesday on another Tactics Tuesday.