Sept. 30, 2025
075: Tactics Tuesday: Using AI to Create Your Own Amazon CEO
Ever wished you could clone yourself to handle all the demands of running your Amazon business? That's exactly what our team has accomplished using custom GPTs to create a "Digital Amazon CEO" that thinks like us, responds like us, and dramatically accelerates our workflow. In this tactics-packed episode, we pull back the curtain on how we've built an AI infrastructure that's transforming our brands and client businesses. Matt shares how he went from using other people's custom GPTs to creat...
Ever wished you could clone yourself to handle all the demands of running your Amazon business? That's exactly what our team has accomplished using custom GPTs to create a "Digital Amazon CEO" that thinks like us, responds like us, and dramatically accelerates our workflow.
In this tactics-packed episode, we pull back the curtain on how we've built an AI infrastructure that's transforming our brands and client businesses. Matt shares how he went from using other people's custom GPTs to creating his own in just 30-40 minutes, resulting in specialized AI tools that slash work that once took hours down to mere minutes. You'll discover how an "AI Command Center" with departmental GPTs for marketing, sales, and specific campaigns creates a powerful ecosystem where AIs collaborate to solve problems more effectively than any single model could.
The game-changer? These custom GPTs remember everything about your brand, eliminating the need to provide extensive context with each new task. Imagine crafting comprehensive Amazon listing image briefs in five minutes or producing an entire email sequence in three minutes—all while maintaining consistent brand voice and strategy. We explore why specialized AIs hallucinate less, produce more accurate results, and why using AI aggregator services gives you the perfect tool for every job without breaking the bank.
Whether you're AI-curious or already incorporating these tools, we provide actionable steps to start leveraging this technology immediately. Begin with existing custom GPTs in the marketplace, learn to create your own, and remember you can always ask AI itself for guidance on using it more effectively. The future of Amazon brand management isn't just using AI—it's creating an AI system that thinks like you do, only faster.
🚀 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 as an Amazon CEO
05:20 - Creating Custom GPTs for Brand Management
10:44 - The AI Command Center Concept
17:15 - Why Multiple Specialized GPTs Work Better
22:47 - AI Aggregator Tools and Model Selection
30:48 - Action Items for Leveraging AI in Business
WEBVTT
00:00:00.321 --> 00:00:02.970
Welcome to another episode of the Brand Fortress HQ podcast.
00:00:02.970 --> 00:00:14.433
On this Tactics Tuesday, we're going to talk about how to use AI to create your own Amazon CEO and just other ways to actually leverage AI in order to move your brand forward on Amazon.
00:00:14.433 --> 00:00:31.885
And so with this, I'm actually going to turn it over to Mike, you and Matt, because you guys are kind of the big brains on this and I've got a couple of experiences and thoughts towards the end, but you guys have really been kind of pushing the boundaries with AI and how you leverage it to be essentially kind of that Amazon CEO.
00:00:31.885 --> 00:00:34.112
So love to hear your guys' experience.
00:00:34.799 --> 00:00:38.271
Yeah, I'd say go ahead and take the reins, Matt, and then I'll pick up after you.
00:00:41.023 --> 00:00:47.231
Yeah, so for about the last I'm going to say eight months or so I've been using other people's custom GPTs.
00:00:47.231 --> 00:00:48.694
So there's a GPT marketplace.
00:00:48.694 --> 00:01:00.023
For those of you that aren't familiar, I think it's only on the paid plan so only the $20 plus plan that you can search for GPTs, Similar to like Facebook groups.
00:01:00.023 --> 00:01:09.667
There's a custom GPT that was trained specifically on pretty much any topic that you can think of and, like I said, I was using other people's for quite some time.
00:01:09.667 --> 00:01:14.447
There's one that I used a lot for marketing a quick clarification on that map.
00:01:14.548 --> 00:01:21.367
So you I don't know if you need to be on the paid plan in order to search for them, but you don't need to be on the paid plan in order to use custom GPTs.
00:01:21.367 --> 00:01:22.750
Got it, Okay?
00:01:23.331 --> 00:01:23.951
Oh, good to know.
00:01:23.971 --> 00:01:28.584
Okay, that's people out there that are like hey, I ran across this custom GPT.
00:01:28.584 --> 00:01:35.347
First of all, just be you know, know where it came from, cause obviously somebody has programmed that and they might've done it.
00:01:35.347 --> 00:01:39.939
Well, they might have not all kinds of different things that could happen with that Just like a piece of software.
00:01:39.939 --> 00:01:42.203
But in order to use those custom GPTs, you can.
00:01:42.203 --> 00:01:44.929
You can use them on the free version of ChatGPT.
00:01:45.290 --> 00:01:45.510
Got it.
00:01:45.510 --> 00:01:55.438
That's a very good clarification and that's actually a really good segue too, because I got a lot of use out of the GPTs that I was using from other people.
00:01:55.438 --> 00:02:05.808
But I went through a training that taught you how to create them yourself, which, now that I know how to create them myself, create them myself.
00:02:05.808 --> 00:02:10.068
But I actually I use chat GPT to help me create and we'll kind of talk about that as we go on.
00:02:10.068 --> 00:02:13.384
But it's not as heavy of a lift as it sounds to like.
00:02:13.384 --> 00:02:19.866
It sounds overwhelming to create a custom GPT and it is like 12, 12 to 14 different training files.
00:02:19.866 --> 00:02:25.189
But, like I said, I leveraged chat GPT to help me walk through that map and also walk through the creation of them.
00:02:25.189 --> 00:02:45.110
But anyways, I went through a training that was called AI Command Center, and basically the idea of the AI Command Center was using a collection of custom GPTs, and what I learned and didn't know at the time is that custom GPTs are powerful, but bringing multiple custom GPTs into the same conversation is a thousand times more powerful.
00:02:45.110 --> 00:02:53.258
So that's what this AI command center concept was essentially is creating a basically a digital CEO.
00:02:53.258 --> 00:03:01.802
They called it the organizational GPT, but to me it makes more sense to call it the digital CEO, but essentially it knows everything that the actual CEO.
00:03:01.802 --> 00:03:04.507
I created one for ProTuff and it knows everything.
00:03:04.507 --> 00:03:31.480
And actually I used Mike actually extracted everything that his chat GPT knew about ProTuff, extracted it out into one big giant report that I then used top of a deep research report that I created and kind of synthesized all of that together to create the training files for the ProTuff digital CEO, which knows as much as Mike does about ProTuff and actually has the power of chat GPT to go out and do more stuff and learn more stuff.
00:03:31.480 --> 00:03:34.307
So that's kind of where it starts is this digital CEO.
00:03:34.307 --> 00:03:37.222
But then there's like departmental CEOs, if you kind of.
00:03:37.222 --> 00:03:42.724
If you continue on with this kind of organizational structure of them, the digital CEO knows everything the CEO knows.
00:03:42.724 --> 00:03:49.788
Then you have your departmental GPTs, like your marketing and sales and data analysis and things like that.
00:03:49.889 --> 00:03:53.245
So marketing was the one that I started out because that's what we needed it for.
00:03:53.245 --> 00:04:04.907
I created a chief marketing GPT that was trained on how to market specifically a physical product on Amazon, tiktok, walmart, all the different platforms.
00:04:04.907 --> 00:04:13.661
So again 12 different training documents on exactly how, specifically on Amazon and TikTok and PPC and DSP and things like that.
00:04:13.661 --> 00:04:16.490
So then under that you have your specialty GPTs.
00:04:16.490 --> 00:04:20.362
I have, for example, a TikTok affiliate success manager one.
00:04:20.802 --> 00:04:44.232
We have a creator connections campaign on Amazon that we launched and we have about 20 or 30 creators that we work with on a regular basis and our admin uses this collection of custom GPTs to talk to each other, to respond to these creators in a very systematized way, in a very similar voice, which is our voice, and it's been fed so many different things to be trained on.
00:04:44.232 --> 00:04:59.961
So essentially, it's kind of the idea is that you're using this collection of custom GPTs in a way that they're trained on not only the brand and everything that comes with that all of the SKUs but then also how to market it, but then how to have conversations with affiliates with it.
00:05:00.423 --> 00:05:29.839
So, matt, quick question for those that aren't super familiar with no-transcript.
00:05:30.663 --> 00:05:43.084
the brands that I worked directly with is that every time I wanted to start a conversation, I had to go back and start with a whole bunch of context, whether it be because I work with five, six, eight different brands at a time.
00:05:43.084 --> 00:05:50.463
So I have to start with okay, this is the brand that I'm talking about, this is their products, this all all the stuff that basically a CEO would know.
00:05:50.463 --> 00:06:01.125
And I had to start every conversation with that context and then give it the context of what I wanted to do and then give it the context of all the information that I was giving it to help me, help, help it, help me.
00:06:01.125 --> 00:06:03.107
So that's a big long process.
00:06:03.107 --> 00:06:09.196
So, essentially, what a custom GPT allows you to do is pre-train it on a very specific thing.
00:06:09.399 --> 00:06:23.810
So breaking that up similar to at a company, like when you get to be a big enough company to require hiring more than five people, naturally those people fall into different departments customer service or sales or something along those lines.
00:06:23.810 --> 00:06:29.540
So there's that separation because that person specializes in that one particular thing for that brand.
00:06:29.540 --> 00:06:41.483
The same thing is true with chat GPT is when you give it a very specific task and can pre-train it on it, you remove that whole process of having to give it background in the first place and you can just go straight to town.
00:06:41.483 --> 00:06:43.629
So now for me creating an email.
00:06:43.629 --> 00:06:56.588
I use three custom GPTs in that process, but I can literally pump out a three email, a copy for three different emails in a series, in three minutes now, as opposed to the day that it took before.
00:06:57.610 --> 00:07:11.103
And it's also, I think, part of what John might've also been asking is you know, if you're talking custom GPTs, why not create just one giant custom GPT that knows everything, instead of all of these separate ones that you're bringing into the conversation?
00:07:11.103 --> 00:07:18.728
And I think the two things that I would say that are the most important reasons that you might want to do.
00:07:18.728 --> 00:07:22.033
That is one, everything that I had.
00:07:23.862 --> 00:07:32.620
Oh, I'm sure Mike was gonna make a very good point, but just as a reminder that this podcast is live, he just cut out, so I'm sure he's gonna say something really important.
00:07:32.620 --> 00:07:39.807
Yeah, that said, so I again, and he's probably gonna say something oh, here it comes back, so we'll let him finish.
00:07:39.807 --> 00:07:41.377
So you just cut out, mike.
00:07:41.377 --> 00:07:44.528
You were saying something very important to have a job.
00:07:44.528 --> 00:07:48.800
It was very suspenseful, so anybody who was listening, I'm sure I was on the edge of my seat.
00:07:49.002 --> 00:07:50.980
It's a needle so what you have to say about it.
00:07:50.980 --> 00:07:52.848
I think the government tried to shut me down, I don't know.
00:07:52.848 --> 00:07:59.966
We'll have to edit that out.
00:07:59.966 --> 00:08:03.732
And AI, at least right now.
00:08:03.732 --> 00:08:10.288
You know we still get situations where you know it hallucinates, it still gives incorrect information sometimes.
00:08:10.288 --> 00:08:17.973
You know, and I think that the more broad its information base, the more likely it is to do that.
00:08:17.973 --> 00:08:28.596
And so in a sense, that's why when you're just talking with the normal chat GPT and you're not giving it a whole lot of context, you're not telling it okay, pretend you are this type sort of person or whatever.
00:08:28.596 --> 00:08:32.590
It tends to hallucinate more in those situations.
00:08:32.590 --> 00:08:39.049
It tends to give poor responses more frequently when you do that, because it's working on a much more generalized knowledge base.
00:08:39.801 --> 00:08:50.067
When you do a custom GPT, having it specialize in something very specific and giving it instructions that tells it, this is the way I want you to think.
00:08:50.067 --> 00:08:52.352
This is the kind of person I want you to be like.
00:08:52.352 --> 00:08:55.346
This is the experience that I want you to pretend that you have.
00:08:55.346 --> 00:09:19.201
It kind of extracts, first of all, it takes the information you give it in the knowledge base, which is very specific, but then also this is my theory is I think it also extracts from its own generalized knowledge base the parts of it that are important for what you want it to do, and it kind of sets aside a lot of the rest of that knowledge base, and so it's much less likely to make mistakes.
00:09:19.201 --> 00:09:19.984
So that's thing one.
00:09:19.984 --> 00:09:26.581
Thing two, though, I think is related to something that we used to do in the early days when Chet GPT first came out.
00:09:26.581 --> 00:09:29.184
This is something that I was doing, and even back then it was powerful.
00:09:29.184 --> 00:09:31.586
Now it's 10 times more powerful.
00:09:32.126 --> 00:09:41.894
But when you have GPTs or AIs talk to themselves, but you give them different personalities or different experience bases.
00:09:41.894 --> 00:09:53.000
So I used to take and instead of having a conversation myself with Chad GBT, I would tell Chad GBT I want to develop this new product in such and such a category.
00:09:53.000 --> 00:10:05.460
So I want you to have a simulated conversation between a materials expert, an engineering expert and a product launch expert or product development expert, or something like that.
00:10:05.460 --> 00:10:14.347
So these three people are going to have a conversation about what products we might launch, and what happens in that conversation then is the.
00:10:14.347 --> 00:10:18.586
So you tell it, this person has such and such experience in, you know, product development.
00:10:18.625 --> 00:10:19.268
Yada, yada, yada.
00:10:19.268 --> 00:10:45.328
You tell it, you know, you give it a little background on each of those three people and then you say go and it just continues this conversation all the way down the page between these three very distinct individuals who would come at this problem from a very different perspective and with a very different knowledge base, and what happens is you get much more creative and much more useful responses at the end of it.
00:10:45.328 --> 00:10:46.804
So the same thing is true here.
00:10:46.804 --> 00:11:03.336
By creating multiple very specialized chat GPTs that you then integrate into the conversation that are, in a sense, kind of talking to each other, you get a much more well-rounded response to whatever it is that you're trying to do, and I just think in the end you're getting a much better result.
00:11:03.697 --> 00:11:15.015
Yeah, yeah, and I think that's such a good point of you know being so specific with the different ChatGPTs and telling it like giving it context, because really otherwise it has.
00:11:15.015 --> 00:11:18.232
I mean, it's almost like having a conversation with an entry-level employee.
00:11:18.232 --> 00:11:25.846
Like if you just out of the blue say, hey, I need you to come up with a new product for this brand, Like their brain's going to explode.
00:11:25.846 --> 00:11:31.370
There's a million different directions that are going to go in and you're going to get a whole bunch of bad ideas before you finally get a good one.
00:11:31.370 --> 00:11:39.691
But if you said, hey, I need a product that does this, is made of these materials, does this like now they can actually do something with that to research what does that look like?
00:11:39.691 --> 00:11:43.448
And, even better, with chat GPT, you can actually say okay to what you were saying.
00:11:43.490 --> 00:11:46.875
Mike, what would an expert look at in this situation?
00:11:47.524 --> 00:12:17.047
And I think where I've found this the most valuable recently is and, Matt, you've been working on it a little bit with some of our clients where you know, developing a GPT that looks at, for example, like how do you build out, or quickly building out briefs for a product, listing image stacks, and what I found really interesting out of that was is that it pulled in some ideas that like weren't really, you know, on our radar or other options that I thought was really useful.
00:12:17.187 --> 00:12:57.110
But then I also noticed that there were other things that I've seen as best practices around, like, okay, make sure that your key benefit or unique selling proposition, like is in one of the first images, so that way people understand, like, why would I buy from your brand, you know, in this product over you know some other brands, products and kind of those things that I think are a little more advanced don't necessarily come up you know pre-programmed in the box for ChatGPT, and so you just really get the best of both worlds if you have somebody that has that expertise and then also using ChatGPT to kind of combine those two things together, and I think that's where really the magic happens.
00:12:57.110 --> 00:13:02.328
So it's not about like, hey, ChatGPT is just going to take over you know all of Amazon and all of marketing.
00:13:02.328 --> 00:13:08.326
It's really okay Identifying what is the 70% of what ChatGPT that gives you.
00:13:08.326 --> 00:13:14.852
That is really good, and then what is the 30% that it's missing because it's not general knowledge.
00:13:16.174 --> 00:13:18.537
That example that you gave that custom GPT.
00:13:18.537 --> 00:13:22.921
I think I call it Amazon Image Optimizer GPT, I think is the name of it.
00:13:22.921 --> 00:13:31.471
It's a really, really good example of how I feel like I have a superpower now because I know how to create a custom GPT.
00:13:31.471 --> 00:13:41.195
Or in the aggregator tool that we use, it's called a project and actually custom GPT has projects now that you can actually train very similarly to a custom GPT.
00:13:41.195 --> 00:13:43.971
So, anyways, there's a lot of different ways that you can do that.
00:13:43.971 --> 00:13:58.092
But, like when I, we had probably about three or four different clients that were going through an Amazon image refresh or they were launching a new product or a product line that they needed new images for, so we had a need for that.
00:13:58.092 --> 00:14:04.614
So, because I had that skill of being able to build a custom GPT, it took me about 40 minutes and that's optimized.
00:14:04.614 --> 00:14:10.133
Now I've gotten that down from a couple of hours to about 30 to 40 minutes to create a custom GPT.
00:14:10.634 --> 00:14:16.917
I use chat GPT to go out and find everything that all best practice and aggregate all of that data.
00:14:16.917 --> 00:14:28.105
I think it, like for that specific GPT, it used 37 different websites for the data that it pulled in for the psychology of each image in the stack on Amazon.
00:14:28.105 --> 00:14:30.816
And it came back and I built these.
00:14:30.816 --> 00:14:36.551
I think there's like 10 or 11 different training docs for that specific GPT and we created it.
00:14:36.551 --> 00:14:38.274
It went through the process.
00:14:38.274 --> 00:14:40.379
So, for a client, I brought in their digital CEO.
00:14:40.379 --> 00:14:41.768
I told her what we wanted.
00:14:41.768 --> 00:14:56.024
Then we brought in the image optimizer, we told it all the USPs, we gave it their website, we gave it their brand intelligence guide that we use to train the digital CEO, and then we had it go out and say, okay, so, based on all those best practices, ceo.
00:14:56.024 --> 00:14:59.719
And then we had it go out and say, okay, so, based on all those best practices, give us what you would suggest for each stack in or each image in the stack.
00:14:59.719 --> 00:15:09.307
And what we learned and John brought this example up is that some of the jobs that it told us for a particular image weren't what we know to be best practice.
00:15:09.307 --> 00:15:16.072
So we had that conversation and then we went back to that training document and we updated it based on what we know to be best practice.
00:15:16.131 --> 00:15:22.208
For the second image in the stack, specifically, the two main usps called out in one image.
00:15:22.208 --> 00:15:23.272
So we changed it.
00:15:23.272 --> 00:15:28.251
We updated the knowledge base after we ran it through for a couple of clients and now it knows that for the next one.
00:15:28.251 --> 00:15:30.397
So that again, that what I told you.
00:15:30.397 --> 00:15:34.461
It would take me three minutes from start to finish to do an email campaign.
00:15:34.461 --> 00:15:41.724
It also would take me about five minutes now to do a campaign or a image stack brief for our designer.
00:15:41.724 --> 00:15:45.553
Where before that would take me multiple hours, now I can do it in five minutes.
00:15:45.553 --> 00:15:52.298
So that's the power of using these collection of custom GPTs and being able to create ones to fit a specific purpose.
00:15:52.784 --> 00:16:20.220
I also think that one of the things that causes people, the most problem with trying to use AI is that I think oftentimes the idea that we have on our head in terms of what we need ChatGPT to help us with, I think oftentimes we're in the middle of the conversation and we need to back up and kind of re-engineer where we should have started in the conversation.
00:16:20.220 --> 00:16:31.405
What I mean by that is we like when we have an idea about something you know we'll go in and we just kind of type in a prompt and we're looking for ChatGPT to make magic Right.
00:16:31.405 --> 00:16:41.051
But the thing that that makes GPT ork or I mean like we keep talking about chat GPT but honestly, claude, is exceptionally good.
00:16:41.051 --> 00:16:43.798
You know, sonnet 4, opus 4.1 are amazing.
00:16:43.798 --> 00:16:47.754
So it's not like you have to use chat GPT, but any of these AI models.
00:16:47.794 --> 00:17:05.688
The thing that makes them most amazing in my mind is oftentimes the things that it tells me that I should be thinking about, that I'm not thinking about, and so what I'll often do is my first prompt really is basically expressing to ChatGPT.
00:17:05.688 --> 00:17:09.699
This is where I am in the process what did I forget?
00:17:09.699 --> 00:17:11.443
Where should I go back to?
00:17:11.443 --> 00:17:13.188
What questions should I be asking that?
00:17:13.188 --> 00:17:14.570
Maybe I haven't already asked?
00:17:14.570 --> 00:17:30.092
What information should I be trying to find that maybe I haven't already looked for, you know, like letting it ask me questions about you know, this process before I delve into the actual conversation and trying to move forward, I try to move backward first.
00:17:30.092 --> 00:17:50.688
You know, it is shocking to me the number of times that I enter into a conversation with GPT and recognize I am I'm not even in the ballpark of where I actually should be in that conversation and the direction that I should be moving, because there was these other pieces that didn't even occur to me that I should be thinking about and I'm like, oh wow, you know like.
00:17:50.848 --> 00:17:52.131
OK, let's start there.
00:17:52.131 --> 00:17:53.614
Then you know, let's, let's move from there.
00:17:53.614 --> 00:17:55.317
So I think that's thing one.
00:17:55.317 --> 00:18:02.489
Thing two is don't look at it as like a genie in the lamp.
00:18:02.489 --> 00:18:05.317
You know where you rub the lamp and you get your magic answer the first try, or even in the first three tries.
00:18:05.317 --> 00:18:07.728
Let's say it's you need.
00:18:07.728 --> 00:18:19.634
You need to keep going with that conversation and keep iterating on that conversation until you get where you want to go, and sometimes that takes some time, and so I think that's a big value piece there.
00:18:20.605 --> 00:18:33.984
One of the things that I would point out with that is that that's one of the value pieces of that custom GPT is that a lot of times, that process of going through that prompt is giving it all that background.
00:18:33.984 --> 00:18:41.547
So having that custom GPT to start with that already has all that background, saves a massive amount of time, and so it's really a good idea.
00:18:41.547 --> 00:18:45.335
Other platforms have similar functions to a custom GPT.
00:18:45.335 --> 00:18:50.944
So if you're not using, you know, chat GPT, figure out what that function is and the other platform that you're using Do they.
00:18:50.965 --> 00:18:52.548
Because I haven't seen this with.
00:18:52.548 --> 00:18:56.998
I've messed around a little bit with Gemini and that type of stuff and I haven't seen it yet.
00:18:57.585 --> 00:18:58.708
Gemini had something.
00:18:58.708 --> 00:19:01.375
I was in there the other day and I could swear I don't remember what gems.
00:19:01.556 --> 00:19:02.598
Are they called gems?
00:19:02.598 --> 00:19:06.507
I think, yeah, I think they're called gems.
00:19:07.028 --> 00:19:11.251
Now I didn't play around with it that much because I don't work in there that much.
00:19:11.251 --> 00:19:31.847
Most of my time and this is probably a good segue, most of my time these days is spent using an aggregation service Matt has used in the past.
00:19:31.847 --> 00:19:33.291
I kind of latched on to Abacus, which I like quite well.
00:19:33.291 --> 00:19:49.656
I've never used Magi, maybe it's even better, but Abacus is what we use as a company and to clarify, you know, an aggregator service essentially what it does is it provides you with access to all of the different AI models at any given moment.
00:19:49.656 --> 00:20:08.487
So anytime I start a chat, I can do one of two things I can either allow Abacus to decide which chat model to utilize in response to my prompt, and basically what it's doing is it's using past data to establish what we have determined over time.
00:20:08.487 --> 00:20:13.146
Is this type of question is best answered by this model.
00:20:13.146 --> 00:20:47.596
So maybe it chooses Chet, gpt 5.0, or it chooses, you know, quad, anthropic, you know maybe it uses Opus 4.1 or whatever, and you get charged a certain number of credits based on you know the amount of processing and the type of model that you use, because some models are more expensive than others, at least by API, but you have access to everything, and that's not just, you know, grok and Claude and ChatGPT and Gemini, it's also Midjourney and Flux and all of these different types of AI models that are either image generators or video generators.
00:20:47.596 --> 00:20:55.116
Or, you know, all of these various different models that you could use deep seek and all of these more exotic things that a lot of people don't touch.
00:20:55.116 --> 00:21:05.894
All of them are accessible, and so you can either let Abacus choose or you can specifically select which model you would like to use for this particular prompt.
00:21:06.575 --> 00:21:12.488
For me, whenever it's a prompt that I'm like I really need some serious thinking power in this.
00:21:12.488 --> 00:21:17.445
There are certain models that I go to that I feel are better with logic and reasoning.
00:21:17.445 --> 00:21:19.411
I'll often use Clause Sonnet 4.
00:21:19.411 --> 00:21:25.355
If I really need something good, quite frankly, I'll often use Clause Opus 4.1, which uses a lot of credits.
00:21:25.355 --> 00:21:27.982
But I get to the answer.
00:21:27.982 --> 00:21:29.661
I was actually having a conversation with one of my employees earlier today, so one of the.
00:21:29.661 --> 00:21:29.619
But I get to the answer.
00:21:29.619 --> 00:21:40.507
I was actually having a conversation with one of my employees earlier today, so one of the things that I think oftentimes when people are using these aggregators because it'll show you how many credits you used and you purchase a certain number of credits.
00:21:40.527 --> 00:21:52.748
You know, based on your plan, you get a certain number of credits every month, and so a lot of times what people will do is they're like oh, like, I know that GPT 4.1 would use like five credits for this prompt, and if I use Opus it's going to be like 20.
00:21:52.748 --> 00:21:54.771
So I'm going to use 4.1.
00:21:54.771 --> 00:22:10.727
But the problem is you're asking it a question that, reasonably speaking, is a little bit over its head, and so what happens is it takes four or five or six prompts to actually get to where you need to be in that conversation, to where you get a real answer.
00:22:10.727 --> 00:22:19.329
Well, now you've already used those 20 credits, but you've also now wasted the time of going through multiple prompts to get there.
00:22:19.329 --> 00:22:39.796
So you're better off sometimes using that more advanced model that might use four or five, six times the credits, but you're going to get the right answer right out of the gate based on that first prompt that you put in, rather than spending another 10 minutes, you know, reiterating the prompt and changing this and going and doing this.
00:22:40.404 --> 00:22:59.633
So it's one of the reasons I really like aggregators, because I'm not always convinced that ChatGPT is the best model for what it is that I'm trying to do, and so it allows me to choose the model that I want and for image generation, like, if you use ChatGPT for image generation, it's not bad, it can do some pretty decent images.
00:22:59.712 --> 00:23:08.198
But if you get more specific, you know, and I want to use Midjourney or I want to use Flux or I want to use whatever I can do all of that right inside of Abacus.
00:23:08.198 --> 00:23:13.161
I don't have to go out and go to MidJourney, I don't have to go out and go to Flux or whatever.
00:23:13.161 --> 00:23:23.378
It's all in that same system and it's actually cheaper because I don't have to pay for MidJourney and ChatGPT and Grok and Claude and all of these.
00:23:23.378 --> 00:23:25.692
I don't have to have a service on any of them.
00:23:25.692 --> 00:23:34.173
I pay for Abacus and Abacus uses their API to connect into all of these other services and I don't have to pay for any of them.
00:23:34.173 --> 00:23:49.028
So if you want access to a lot of those other models, especially at varied types of models you know, for video and images and things like that, using an aggregator could be a way to do that in a way that doesn't break the bank, you know, for your business but gives you access to a lot of the tools that you're looking for.
00:23:49.789 --> 00:23:53.374
I mean, it saved me money and it gave me access to more things.
00:23:53.374 --> 00:23:55.316
So I'm curious.
00:23:55.415 --> 00:23:59.840
You know kind of taking a little bit of a sidestep here.
00:23:59.840 --> 00:24:10.181
One of the other things, because we talked a lot about custom GPTs but there are other tools now out there that essentially act like custom GPTs, but across multiple platforms.
00:24:10.181 --> 00:24:11.243
That I know.
00:24:11.243 --> 00:24:17.096
That's something that we're well I'm starting to look into with my agency and looking at leveraging that.
00:24:17.096 --> 00:24:20.230
So that way, custom GPTs are great.
00:24:20.230 --> 00:24:31.386
But again back to what you were talking about, mike of like okay, well, if we can get a better result out of you know Gemini versus ChatGPT, like we just care what you know model gives us the best result.
00:24:31.386 --> 00:24:36.698
I'm not married to you know ChatGPT versus Grok or you know whatever.
00:24:37.285 --> 00:24:42.645
Well, an abacus is actually nice in that regard and again, matt can speak to Magi and how it functions.
00:24:42.645 --> 00:24:48.615
But so within abacus, you create essentially a custom GPT, and abacus is called a project.
00:24:48.615 --> 00:24:51.319
Now, they're not exactly the same thing, but they're similar.
00:24:51.319 --> 00:25:00.415
So when you start a project in Abacus, you give it a knowledge base and you give it a set of custom instructions, which is pretty much what a custom GPT is.
00:25:00.415 --> 00:25:06.252
When you do a custom GPT, you give it a knowledge base and then you give it its custom instructions, and then you go from there, right, you're off to the races.
00:25:06.252 --> 00:25:15.347
So in that sense, they're identical.
00:25:15.347 --> 00:25:19.259
What's nice, though, about Abacus projects is is that then you have access to exactly what you're suggesting, john, which is one.
00:25:19.259 --> 00:25:27.671
You can open up a new project chat, and now it automatically has access to all of that knowledge base and the custom instructions that you gave it, which is kind of who are you, how do you think?
00:25:27.671 --> 00:25:34.707
What kind of you know like all that kind of thing that you gave it, which is kind of who are you, how do you think what kind of you know like all that kind of thing?
00:25:34.707 --> 00:25:36.814
But then when you start the conversation again, same as any other chat in Abacus.
00:25:36.834 --> 00:25:46.218
You can either let Abacus choose which model it should use to answer the questions and, by the way, I should clarify that model can change at any point in the conversation.
00:25:46.218 --> 00:25:52.188
So I can start with like like I start the conversation and let's say I just let Abacus choose.
00:25:52.188 --> 00:25:57.507
I ask it a question and it decides Gemini 2.5 is the right model, so it answers the question.
00:25:57.507 --> 00:25:59.792
Then I follow that with another.
00:25:59.792 --> 00:26:01.596
You know, question or reply.
00:26:02.845 --> 00:26:13.712
Abacus might decide okay, well, now in this point of the conversation, claude Sonnet 4 is better, or ChatGPT this is better, or you know whatever, and it will switch models at will.
00:26:13.712 --> 00:26:15.888
You can then also click regenerate.
00:26:15.888 --> 00:26:22.352
If it gives you an answer and you don't really like it and you think there's a different model that might work better, it'll give you suggestions of other models that you could have tried.
00:26:22.352 --> 00:26:31.028
You can regenerate that answer with a different model, but also same thing you can choose.
00:26:31.028 --> 00:26:32.913
So at any point in that conversation I can change it and say use QuadSonic 4.
00:26:32.913 --> 00:26:37.992
Or you know, now I'm at this point in the conversation I'm like oh, we need to use Flux right now because I want to do an image, or you know whatever.
00:26:37.992 --> 00:26:39.846
So you're in that project.
00:26:39.846 --> 00:26:48.838
It still has all the context, all of that information, just like a custom GPT, but now you can use any model you want within that chat.
00:26:49.819 --> 00:26:55.017
Yeah, and I think there's a couple of just important takeaways for listeners out of that what you just said, mike.
00:26:55.017 --> 00:27:01.925
So the first one is is that it means you have the right tool for the job you know in the sense of, or the best tool for the job you know.
00:27:01.925 --> 00:27:16.738
It's like you know, if you'll reach into your toolbox and you're like, okay, I need to, you know, pound in a nail, okay, and you're like, okay, I need to, you know, pound in a nail Okay, well, if the only thing you have is a wrench, like, yeah, you can make it work, but you know, a hammer would be a lot nicer.
00:27:16.738 --> 00:27:27.795
And so that's kind of, as we're starting to see these different tools, you know, kind of mature, they're good, good at different things, and having something like that aggregator, aggregator, abacus or whatever gives you access to all the tools in the toolbox.
00:27:28.435 --> 00:27:40.484
I think the other important thing, what we've found and again we primarily use ChatGPT, but just you know from what you bring up which is, it allows you to keep the results consistent.
00:27:40.484 --> 00:27:43.375
So you know, okay, which model gave us the best results?
00:27:43.375 --> 00:27:49.607
And so then if you're having your team use that consistently, you know that same project over and over again.
00:27:49.607 --> 00:28:07.634
You know like, hey, here are the parameters that give us the best outcome, given this project, that we may do multiple times a day, you know we may do, you know, use tons and tons of time, or do many, many times in order to get a really consistent result.
00:28:07.634 --> 00:28:20.429
That actually does have a very positive impact on the business by allowing you to speed up processes and deliver a lot faster than your competitors, which is super important, especially on Amazon.
00:28:20.990 --> 00:28:24.988
Well, and if you're a team, I mean also just the cost.
00:28:24.988 --> 00:28:29.739
As you said, different tools are good at different things.
00:28:29.739 --> 00:28:34.790
You could easily be subscribed to 20 different AI services.
00:28:34.790 --> 00:28:41.567
You know that do the various things within your business that you need done and you know this tool is better for this, or this tool is better for this, or this tool is better for this.
00:28:41.567 --> 00:28:53.848
Or you might be subscribed to five of them and you're like, look, this one is not the best at this thing, but it's good enough at this thing and I don't want to pay for the one that's the best for it, because then I got another subscription that I'm paying.
00:28:53.848 --> 00:29:03.330
So we're going to stick to these three or four or five tools, whereas with the aggregator, it gives you access to all of the tools under one subscription price.
00:29:03.330 --> 00:29:10.317
And if you're a team like if you got ChatGPT and you're a team like ours, where you got 10 people you know well.
00:29:10.356 --> 00:29:14.201
I got to pay for a seat for every single one of those 10 people, good Lord.
00:29:14.201 --> 00:29:29.135
The price adds up, you know, and if I need multiple tools, you know multiple AIs, that's an astronomical amount of money and most of the time, some of those tools you're paying for you only use occasionally, so having access to them through an aggregator is way less expensive.
00:29:30.645 --> 00:29:33.680
Yeah, I think we've covered some good ground with this conversation.
00:29:33.680 --> 00:29:35.928
I know that there's a lot more that we could talk about.
00:29:35.928 --> 00:29:41.891
I do think this is probably a good place to wrap for today, so, with that in mind, maybe I'll start with you.
00:29:41.891 --> 00:29:56.477
Matt, one kind of action item you'd have for brand owners who are listening right now and are looking for more ways to leverage AI in their business what advice would you have or action item would you have for them?
00:29:57.505 --> 00:30:01.934
So I kind of use ChatGPT as my security blanket.
00:00:00.321 --> 00:00:02.970
00:00:02.970 --> 00:00:14.433
00:00:14.433 --> 00:00:31.885
00:00:31.885 --> 00:00:34.112
00:00:34.799 --> 00:00:38.271
00:00:41.023 --> 00:00:47.231
00:00:47.231 --> 00:00:48.694
00:00:48.694 --> 00:01:00.023
00:01:00.023 --> 00:01:09.667
00:01:09.667 --> 00:01:14.447
00:01:14.548 --> 00:01:21.367
00:01:21.367 --> 00:01:22.750
00:01:23.331 --> 00:01:23.951
00:01:23.971 --> 00:01:28.584
00:01:28.584 --> 00:01:35.347
00:01:35.347 --> 00:01:39.939
00:01:39.939 --> 00:01:42.203
00:01:42.203 --> 00:01:44.929
00:01:45.290 --> 00:01:45.510
00:01:45.510 --> 00:01:55.438
00:01:55.438 --> 00:02:05.808
00:02:05.808 --> 00:02:10.068
00:02:10.068 --> 00:02:13.384
00:02:13.384 --> 00:02:19.866
00:02:19.866 --> 00:02:25.189
00:02:25.189 --> 00:02:45.110
00:02:45.110 --> 00:02:53.258
00:02:53.258 --> 00:03:01.802
00:03:01.802 --> 00:03:04.507
00:03:04.507 --> 00:03:31.480
00:03:31.480 --> 00:03:34.307
00:03:34.307 --> 00:03:37.222
00:03:37.222 --> 00:03:42.724
00:03:42.724 --> 00:03:49.788
00:03:49.889 --> 00:03:53.245
00:03:53.245 --> 00:04:04.907
00:04:04.907 --> 00:04:13.661
00:04:13.661 --> 00:04:16.490
00:04:16.490 --> 00:04:20.362
00:04:20.802 --> 00:04:44.232
00:04:44.232 --> 00:04:59.961
00:05:00.423 --> 00:05:29.839
00:05:30.663 --> 00:05:43.084
00:05:43.084 --> 00:05:50.463
00:05:50.463 --> 00:06:01.125
00:06:01.125 --> 00:06:03.107
00:06:03.107 --> 00:06:09.196
00:06:09.399 --> 00:06:23.810
00:06:23.810 --> 00:06:29.540
00:06:29.540 --> 00:06:41.483
00:06:41.483 --> 00:06:43.629
00:06:43.629 --> 00:06:56.588
00:06:57.610 --> 00:07:11.103
00:07:11.103 --> 00:07:18.728
00:07:18.728 --> 00:07:22.033
00:07:23.862 --> 00:07:32.620
00:07:32.620 --> 00:07:39.807
00:07:39.807 --> 00:07:41.377
00:07:41.377 --> 00:07:44.528
00:07:44.528 --> 00:07:48.800
00:07:49.002 --> 00:07:50.980
00:07:50.980 --> 00:07:52.848
00:07:52.848 --> 00:07:59.966
00:07:59.966 --> 00:08:03.732
00:08:03.732 --> 00:08:10.288
00:08:10.288 --> 00:08:17.973
00:08:17.973 --> 00:08:28.596
00:08:28.596 --> 00:08:32.590
00:08:32.590 --> 00:08:39.049
00:08:39.801 --> 00:08:50.067
00:08:50.067 --> 00:08:52.352
00:08:52.352 --> 00:08:55.346
00:08:55.346 --> 00:09:19.201
00:09:19.201 --> 00:09:19.984
00:09:19.984 --> 00:09:26.581
00:09:26.581 --> 00:09:29.184
00:09:29.184 --> 00:09:31.586
00:09:32.126 --> 00:09:41.894
00:09:41.894 --> 00:09:53.000
00:09:53.000 --> 00:10:05.460
00:10:05.460 --> 00:10:14.347
00:10:14.347 --> 00:10:18.586
00:10:18.625 --> 00:10:19.268
00:10:19.268 --> 00:10:45.328
00:10:45.328 --> 00:10:46.804
00:10:46.804 --> 00:11:03.336
00:11:03.697 --> 00:11:15.015
00:11:15.015 --> 00:11:18.232
00:11:18.232 --> 00:11:25.846
00:11:25.846 --> 00:11:31.370
00:11:31.370 --> 00:11:39.691
00:11:39.691 --> 00:11:43.448
00:11:43.490 --> 00:11:46.875
00:11:47.524 --> 00:12:17.047
00:12:17.187 --> 00:12:57.110
00:12:57.110 --> 00:13:02.328
00:13:02.328 --> 00:13:08.326
00:13:08.326 --> 00:13:14.852
00:13:16.174 --> 00:13:18.537
00:13:18.537 --> 00:13:22.921
00:13:22.921 --> 00:13:31.471
00:13:31.471 --> 00:13:41.195
00:13:41.195 --> 00:13:43.971
00:13:43.971 --> 00:13:58.092
00:13:58.092 --> 00:14:04.614
00:14:04.614 --> 00:14:10.133
00:14:10.634 --> 00:14:16.917
00:14:16.917 --> 00:14:28.105
00:14:28.105 --> 00:14:30.816
00:14:30.816 --> 00:14:36.551
00:14:36.551 --> 00:14:38.274
00:14:38.274 --> 00:14:40.379
00:14:40.379 --> 00:14:41.768
00:14:41.768 --> 00:14:56.024
00:14:56.024 --> 00:14:59.719
00:14:59.719 --> 00:15:09.307
00:15:09.307 --> 00:15:16.072
00:15:16.131 --> 00:15:22.208
00:15:22.208 --> 00:15:23.272
00:15:23.272 --> 00:15:28.251
00:15:28.251 --> 00:15:30.397
00:15:30.397 --> 00:15:34.461
00:15:34.461 --> 00:15:41.724
00:15:41.724 --> 00:15:45.553
00:15:45.553 --> 00:15:52.298
00:15:52.784 --> 00:16:20.220
00:16:20.220 --> 00:16:31.405
00:16:31.405 --> 00:16:41.051
00:16:41.051 --> 00:16:43.798
00:16:43.798 --> 00:16:47.754
00:16:47.794 --> 00:17:05.688
00:17:05.688 --> 00:17:09.699
00:17:09.699 --> 00:17:11.443
00:17:11.443 --> 00:17:13.188
00:17:13.188 --> 00:17:14.570
00:17:14.570 --> 00:17:30.092
00:17:30.092 --> 00:17:50.688
00:17:50.848 --> 00:17:52.131
00:17:52.131 --> 00:17:53.614
00:17:53.614 --> 00:17:55.317
00:17:55.317 --> 00:18:02.489
00:18:02.489 --> 00:18:05.317
00:18:05.317 --> 00:18:07.728
00:18:07.728 --> 00:18:19.634
00:18:20.605 --> 00:18:33.984
00:18:33.984 --> 00:18:41.547
00:18:41.547 --> 00:18:45.335
00:18:45.335 --> 00:18:50.944
00:18:50.965 --> 00:18:52.548
00:18:52.548 --> 00:18:56.998
00:18:57.585 --> 00:18:58.708
00:18:58.708 --> 00:19:01.375
00:19:01.556 --> 00:19:02.598
00:19:02.598 --> 00:19:06.507
00:19:07.028 --> 00:19:11.251
00:19:11.251 --> 00:19:31.847
00:19:31.847 --> 00:19:33.291
00:19:33.291 --> 00:19:49.656
00:19:49.656 --> 00:20:08.487
00:20:08.487 --> 00:20:13.146
00:20:13.146 --> 00:20:47.596
00:20:47.596 --> 00:20:55.116
00:20:55.116 --> 00:21:05.894
00:21:06.575 --> 00:21:12.488
00:21:12.488 --> 00:21:17.445
00:21:17.445 --> 00:21:19.411
00:21:19.411 --> 00:21:25.355
00:21:25.355 --> 00:21:27.982
00:21:27.982 --> 00:21:29.661
00:21:29.661 --> 00:21:29.619
00:21:29.619 --> 00:21:40.507
00:21:40.527 --> 00:21:52.748
00:21:52.748 --> 00:21:54.771
00:21:54.771 --> 00:22:10.727
00:22:10.727 --> 00:22:19.329
00:22:19.329 --> 00:22:39.796
00:22:40.404 --> 00:22:59.633
00:22:59.712 --> 00:23:08.198
00:23:08.198 --> 00:23:13.161
00:23:13.161 --> 00:23:23.378
00:23:23.378 --> 00:23:25.692
00:23:25.692 --> 00:23:34.173
00:23:34.173 --> 00:23:49.028
00:23:49.789 --> 00:23:53.374
00:23:53.374 --> 00:23:55.316
00:23:55.415 --> 00:23:59.840
00:23:59.840 --> 00:24:10.181
00:24:10.181 --> 00:24:11.243
00:24:11.243 --> 00:24:17.096
00:24:17.096 --> 00:24:20.230
00:24:20.230 --> 00:24:31.386
00:24:31.386 --> 00:24:36.698
00:24:37.285 --> 00:24:42.645
00:24:42.645 --> 00:24:48.615
00:24:48.615 --> 00:24:51.319
00:24:51.319 --> 00:25:00.415
00:25:00.415 --> 00:25:06.252
00:25:06.252 --> 00:25:15.347
00:25:15.347 --> 00:25:19.259
00:25:19.259 --> 00:25:27.671
00:25:27.671 --> 00:25:34.707
00:25:34.707 --> 00:25:36.814
00:25:36.834 --> 00:25:46.218
00:25:46.218 --> 00:25:52.188
00:25:52.188 --> 00:25:57.507
00:25:57.507 --> 00:25:59.792
00:25:59.792 --> 00:26:01.596
00:26:02.845 --> 00:26:13.712
00:26:13.712 --> 00:26:15.888
00:26:15.888 --> 00:26:22.352
00:26:22.352 --> 00:26:31.028
00:26:31.028 --> 00:26:32.913
00:26:32.913 --> 00:26:37.992
00:26:37.992 --> 00:26:39.846
00:26:39.846 --> 00:26:48.838
00:26:49.819 --> 00:26:55.017
00:26:55.017 --> 00:27:01.925
00:27:01.925 --> 00:27:16.738
00:27:16.738 --> 00:27:27.795
00:27:28.435 --> 00:27:40.484
00:27:40.484 --> 00:27:43.375
00:27:43.375 --> 00:27:49.607
00:27:49.607 --> 00:28:07.634
00:28:07.634 --> 00:28:20.429
00:28:20.990 --> 00:28:24.988
00:28:24.988 --> 00:28:29.739
00:28:29.739 --> 00:28:34.790
00:28:34.790 --> 00:28:41.567
00:28:41.567 --> 00:28:53.848
00:28:53.848 --> 00:29:03.330
00:29:03.330 --> 00:29:10.317
00:29:10.356 --> 00:29:14.201
00:29:14.201 --> 00:29:29.135
00:29:30.645 --> 00:29:33.680
00:29:33.680 --> 00:29:35.928
00:29:35.928 --> 00:29:41.891
00:29:41.891 --> 00:29:56.477
00:29:57.505 --> 00:30:01.934