043: Tactic Tuesdays: Amazon AI Optimization

Transcript

WEBVTT

00:00:00.020 --> 00:00:02.229
Welcome everyone to the Brand Fortress ATU podcast.

00:00:02.229 --> 00:00:19.228
Today we have another Tactics Tuesday episode where we're gonna be discussing something that's been talked about kind of on the edges over the last six months or so and we're starting to see how Amazon's implementing a little bit more, and that is their AI constructs, both as sellers and then also as consumers.

00:00:19.228 --> 00:00:28.640
So with that said, mike, I'm actually gonna let you kick it off and kind of your experiences so far with Amazon Comprehend and the other AI inputs that you're seeing from Amazon.

00:00:29.422 --> 00:00:29.623
Yeah.

00:00:29.623 --> 00:00:51.985
So a couple of things rebuilding our listings to a degree to speak to the AI as opposed to just speaking to the consumer or to Amazon's typical ranking algorithm.

00:00:51.985 --> 00:01:05.096
Let's say, of the past, the AI is making much more of a play, I think over the coming months, and certainly over the next year or so I would say it's going to be pretty critical, so preparing for that is important.

00:01:05.096 --> 00:01:18.128
So we started taking our listings and putting the text of the listing through Amazon's Comprehend service software whatever you want to call it.

00:01:18.128 --> 00:01:28.543
It's built into AWS and so essentially, we'll take the bullets from our listing and we'll paste it into a text box within Amazon Comprehend.

00:01:28.543 --> 00:01:47.131
We hit submit and essentially what Amazon Comprehend does is that it breaks it down to basically show you how the AI sees the text that you've entered in terms of you know, semantically, how does it break out the individual phrases within the text that you've entered?

00:01:47.131 --> 00:02:17.010
Does it give a positive or a negative semantic relevance to the phrases that are in there, whether that beIA seeing the key phrases that you're trying to target properly, or is it breaking them up improperly, and how can you rewrite the text in order to have it see those key phrases the way that you want it to see them?

00:02:17.010 --> 00:02:27.241
And then, secondarily to that, does it have high confidence that it knows what those key phrases are and how relevant you are to them?

00:02:27.241 --> 00:02:32.050
And then, thirdly, does it view them in a positive or a negative light?

00:02:32.050 --> 00:02:40.783
And, I'll be honest, I don't fully understand exactly the nuances of you know situations where you might want it to be positive or negative.

00:02:40.783 --> 00:02:56.280
I can actually see scenarios where, depending on what the keyword was, maybe you might want it to have a negative semantic relevance versus positive, and maybe you'd want it to be strong versus weak, so there's some things that you probably want to research.

00:02:56.800 --> 00:03:00.860
The one thing I would say is, though, we've taken two listings and run it through that process.

00:03:00.860 --> 00:03:04.688
We put our title in, then we put our bullets in, we put our description in.

00:03:04.688 --> 00:03:06.753
You know, we kind of walk through that process.

00:03:06.753 --> 00:03:11.043
We made those edits to those listings over a month ago.

00:03:11.043 --> 00:03:18.354
We have seen zero change in ranking, in sales, in anything you know.

00:03:18.354 --> 00:03:22.526
That would suggest that those changes made a difference for us.

00:03:23.347 --> 00:03:29.921
However, I do believe that it will produce benefits down the road.

00:03:29.921 --> 00:03:53.593
At the least, I think one of the benefits that will end up coming into play is Amazon has this new AI tool for consumers called Rufus, that will be playing a stronger role, I think, going forward, in presenting the right products to a consumer based on their search or their description of what they're looking for.

00:03:53.593 --> 00:04:14.084
So it's going to be much less of a type something in, as it is probably going to be speaking to their phone this is what I'm looking for for this purpose, whatever and Rufus is going to then find the product that fits, that fits the category of what they said, and combines that with who they are, because Amazon knows that customer.

00:04:14.084 --> 00:04:17.642
They know what they bought, they know what they like, they know what they watch, they know all of that.

00:04:17.642 --> 00:04:22.882
So Rufus is going to, over time, put that all together, and so you're preparing for that.

00:04:23.423 --> 00:04:25.567
Okay, so a couple of questions around this.

00:04:25.567 --> 00:04:31.346
The first one is so for Comprehend you mentioned, so this is AWS, which is owned by Amazon.

00:04:31.346 --> 00:04:47.826
For people that may not be familiar, is this a publicly available tool that somebody can go out and find this AWS, comprehend, and then are they just uploading the link to their listing, or how does that work in order to get out of that?

00:04:47.826 --> 00:04:48.168
Yeah, so with.

00:04:48.187 --> 00:04:49.310
Comprehend, it's totally free.

00:04:49.310 --> 00:05:02.012
So if you have an AWS account and an AWS account is also free, there are services within AWS that you would have to pay for, but you can set up a root account on AWS and it doesn't cost anything.

00:05:02.012 --> 00:05:04.206
Amazon Comprehend is completely free.

00:05:04.206 --> 00:05:05.249
It's directly within there.

00:05:05.249 --> 00:05:16.845
So if you were to set up an AWS account, I think if you just typed into the search bar in fact, I think this is what I originally did Just type Amazon Comprehend or even just Comprehend into the search bar, It'll bring up that tool.

00:05:16.845 --> 00:05:18.488
And so you open it up.

00:05:18.488 --> 00:05:20.312
There's a text box at the top.

00:05:20.312 --> 00:05:29.788
You paste the text into that box and hit submit, and then Amazon tells you what it thinks about that text, what the AI understands from the text that you've entered.

00:05:29.788 --> 00:05:31.781
So you don't give it a URL.

00:05:31.781 --> 00:05:44.391
You would take snippets, you know pieces of text from your listing or whatever you want it to evaluate, and just paste it in the text box and hit submit and it will spit back what it wants, what it thinks it knows about it.

00:05:44.391 --> 00:05:51.990
And then what I was doing then was simply rewriting the text so that it would.

00:05:51.990 --> 00:06:00.146
Here's an example One of the key phrases that we target is pool poles for cleaning.

00:06:01.086 --> 00:06:07.276
And the reason that key phrase exists is because there are pool poles that have nothing to do with cleaning.

00:06:07.276 --> 00:06:11.451
It's to set up the structure of an above ground pool.

00:06:11.451 --> 00:06:15.911
You have poles that go in, and so pool poles could mean that.

00:06:15.911 --> 00:06:23.649
And so the reason that people use that pool poles for cleaning phrase is to specify no, I want for cleaning, so we target that key phrase.

00:06:23.649 --> 00:06:25.569
No, I want for cleaning, so we target that key phrase.

00:06:25.589 --> 00:06:47.137
Depending on how I type that in and whether I put commas around it or whether I put dashes around it, or you know the way that the sentence is phrased Amazon may break that up and we target pool pools and, yes, cleaning is relevant, but I want the whole phrase.

00:06:47.137 --> 00:06:53.661
So I might have to change how I word that sentence so that Amazon doesn't break it up that way.

00:06:53.661 --> 00:07:09.975
So it recognizes that phrase as being one holistic phrase, and that might involve putting commas around it, brackets around it, dashes around it, or it might mean you have to change exactly how you word the phrase, the entire sentence, to change.

00:07:09.975 --> 00:07:13.244
You know whether it recognizes that phrase in its entirety or in pieces.

00:07:14.447 --> 00:07:25.531
So the idea behind running it through Comprehend is that, as a community, what we think and Amazon's put out a white paper on this.

00:07:25.531 --> 00:07:29.302
That Vanessa the episode that we did with Vanessa Hung.

00:07:29.302 --> 00:07:30.889
She talked quite a bit about this.

00:07:30.889 --> 00:07:34.141
I mean it was fairly new, so she went a lot more in depth.

00:07:34.141 --> 00:07:45.812
But the idea behind this is that the language model that Comprehend uses is going to be similar, if not the same, as the language model that Rufus uses.

00:07:45.812 --> 00:07:54.447
So Comprehend is giving us that data so we can kind of get more of a glimpse as to what's happening behind the scenes with Rufus.

00:07:54.447 --> 00:07:55.189
Is that the idea?

00:07:55.771 --> 00:08:11.783
Yeah, the assumption would be that if Amazon Comprehend recognizes the text that you've entered in the way that you would like it to, that you think you would want the AI to understand it so that it's likely to present your product in the best light and to the right customers.

00:08:11.783 --> 00:08:24.129
The assumption is, if Amazon Comprehend sees it this way, rufus will see it this way and so get in the hands of the right, or it will be put in front of the right customers as the product that they want.

00:08:24.129 --> 00:08:25.853
That's the idea, the assumption, okay.

00:08:26.595 --> 00:08:58.684
Yeah, and so just I want to make sure that listeners understand that comprehend is available through AWS, and so what that allows us to do is kind of test, like you said, like the language around what we're doing, whereas Rufus is actually the customer facing on amazoncom buying products, portion of it that, as a customer, I might put in gift for six-year-old birthday party, and then it would come up with some different ideas or something like that, something that's more robust than what we currently have through search, where it's very keyword focused.

00:08:59.386 --> 00:09:00.450
And I would actually recommend.

00:09:00.450 --> 00:09:13.647
We haven't done this, but I would actually recommend and I'm just thinking about this now as we're talking about it we really should start testing Rufus as sellers, to see how it works and what it does and what it spits back.

00:09:13.647 --> 00:09:18.668
It's going to change over time, but we really should be playing around with that, too, as if we're a consumer.

00:09:18.707 --> 00:09:28.062
Yeah, so, matt, I'm curious from your perspective for the brands that you work with and then what you're working on with your own brands.

00:09:28.062 --> 00:09:37.490
Have you changed anything based on the info that's come out over the last six months on AI and, if so, what's your approach right now for this?

00:09:38.360 --> 00:09:51.081
So the only I have not played with Comprehend at all, and I just saw Rufus for the first time in my app yesterday and I wasn't in a place like I was super surprised to see it because it was the first time that I had, but I didn't.

00:09:51.081 --> 00:09:52.684
I wasn't in a place that I could play with it.

00:09:52.684 --> 00:10:02.591
Where I've experienced Amazon's AI is in the listing builder tool and in my opinion it leaves a lot to be desired.

00:10:02.591 --> 00:10:18.602
I've never used it for, I've never actually allowed it to, but like so for me, when I launched my most recent product, I just I put it in the, I chose the category that I was launching in and I let it write the listing to make sure that it kind of just got.

00:10:18.602 --> 00:10:24.591
I just wanted to know what Amazon was understanding about that product and category, but it just it wasn't great.

00:10:24.751 --> 00:10:25.701
The listing wasn't great.

00:10:25.701 --> 00:10:32.929
I think that at some point it will be a lot better than it is, but that's really the only experience that I've had with it and, to be honest, it wasn't a great experience.

00:10:32.929 --> 00:10:35.427
And I think that other sellers also experienced.

00:10:35.427 --> 00:10:41.144
I mean I heard some sellers where the AI changed their title without them wanting it to and they couldn't change it back.

00:10:41.144 --> 00:10:53.090
So from my experience limited experience in Amazon's kind of playing around with this AI it just wasn't great in terms of what the listing builder gave me compared to what I used in the end.

00:10:54.053 --> 00:10:57.129
Yeah, I think there's a couple of things I want to double click on there that you mentioned, matt.

00:10:57.129 --> 00:10:58.600
The first is talking about your listing.

00:10:58.600 --> 00:11:14.101
So we've definitely seen that with brands that we work with where Amazon has changed listings, although frankly, not dramatically, but in added things that we didn't want there, et cetera, et cetera, kind of minor things with the AI and what.

00:11:14.101 --> 00:11:27.120
I'm pretty sure it's been a little while since we looked at this, but there is a setting that is on by default that you can turn off, so that way Amazon does not adjust your listings based on AI recommendations.

00:11:27.120 --> 00:11:57.232
So, you know, I would recommend people do probably a little Google search or a little bit of maybe a chat GPT question on that in order to find that information whatever kind of suits your fancy, but definitely something to look into Because I think, as things go on, obviously Amazon is pushing more and more towards AI having a bigger influence over listings and how their logarithm works and a lot of these things in a shift that we haven't seen in probably four or five years.

00:11:57.232 --> 00:11:59.256
So it's coming.

00:11:59.256 --> 00:12:06.686
It's probably not going to be super fast and, quite frankly, like a lot of things that Amazon rolls out, when they do roll it out to have more influence.

00:12:06.686 --> 00:12:12.932
It's probably going to break a lot of things and then we'll have to figure out on the fly what the impact of those things are.

00:12:12.932 --> 00:12:19.932
And I would just also double click on what you said around AI and listing building.

00:12:19.932 --> 00:12:23.850
So I remember being at Amazon Accelerate and I know that you were there as well.

00:12:23.890 --> 00:12:32.484
One of the sessions that I went into Amazon, the individual that was presenting on it from Amazon essentially showed a case study.

00:12:32.565 --> 00:12:41.126
That was kind of perfect case scenario where you would upload an image of your product and it would write the title and write all the bullets and all of it was fantastic and that type of stuff.

00:12:41.126 --> 00:12:51.030
The reality of it is is that it is still far from good of the information that you get from Amazon.

00:12:51.030 --> 00:12:54.942
When you upload it and have that AI piece, you know write your listing based on your title or based piece.

00:12:54.942 --> 00:12:58.919
You know write your listing based on your title or based on you know some keywords or even an image.

00:12:58.919 --> 00:13:12.777
However, what we started incorporating it for is having Amazon look at, especially when we're building a listing, upload that image, so that way we get an idea of what Amazon thinks that product is.

00:13:12.777 --> 00:13:17.447
So that way we know like, hey, do we need to change this main image?

00:13:17.447 --> 00:13:26.993
So that way, what we know the customers that we want to target matches what Amazon thinks that this product is and that part is valuable.

00:13:26.993 --> 00:13:34.756
So we're already starting to see that AI influence on how we manage things on the backend.

00:13:36.307 --> 00:13:38.394
And over time, that's obviously going to get better.

00:13:38.394 --> 00:13:43.716
You know Amazon is the AI is going to start to figure out.

00:13:43.716 --> 00:13:48.216
You know what those images actually are.

00:13:48.216 --> 00:13:50.312
I will agree with you 100%.

00:13:50.312 --> 00:14:08.839
Amazon completely does not understand our category Numerous products within our category it has no idea If I, if I upload an image, no matter what style the image is lifestyle, you know, white background, whatever Amazon literally cannot figure out what it is.

00:14:08.839 --> 00:14:14.047
Every time I upload it and I ask it to create a listing, it creates a listing that has nothing to do with pools.

00:14:14.047 --> 00:14:14.969
It's not.

00:14:14.969 --> 00:14:16.932
It's not even close in the category.

00:14:17.553 --> 00:14:28.874
But it is going to improve, and I do think that one of the things that's going to be important for most sellers is you at least have to be testing it and you have to know what stage Amazon is at.

00:14:28.874 --> 00:14:33.308
I mean, if you're not testing this, you're not going to know, you're going to be behind the eight ball.

00:14:33.308 --> 00:14:44.850
So I would recommend that you continue to upload images and allow Amazon to create a product listing and see how is the AI changing in terms of how well it understands your category and understands your products.

00:14:44.850 --> 00:15:06.692
And I would also say that some categories Amazon understands pretty well, and so for those categories where that's true, then using that strategy of uploading those images and seeing what Amazon thinks that image means is, I think, going to be a really effective strategy going forward to make sure that you know, make modifications to that image as best you can to make sure Amazon understands it the way you want it to.

00:15:06.812 --> 00:15:22.087
Yeah, I think the other thing, quite frankly, is pushing towards a best practice that you should have anyway that Vanessa mentioned, which is looking at In the episode that we did a couple of months ago with her, which was looking at your flat files.

00:15:22.708 --> 00:15:30.812
So there's a lot of different aspects of your listing that may not, that aren't visible first of all, especially from the customer perspective.

00:15:30.812 --> 00:15:51.513
But even if you're looking on the dashboard, that you're not going to file, that you're not going to see unless you download that flat file and see what your product listing looks like and what you know different not categories, but different sections are filled out and what data is in there, because all those are little breadcrumbs for Amazon's AI.

00:15:51.513 --> 00:15:54.558
So that's going to become more and more important over time.

00:15:54.558 --> 00:16:05.458
So I would you know whether, wherever you're at in this kind of AI journey, as far as really excited about it coming, don't really, you know, think it's way overblown.

00:16:05.458 --> 00:16:21.789
Whatever it happens to be, either way, it is a best practice and I think you know more people need to be looking at that flat file to make sure that the information in there is as accurate as possible and as filled out as possible, to index better and really do better overall for your products on Amazon.

00:16:22.432 --> 00:16:24.966
Can I just say, john, you really deviated from the script.

00:16:24.966 --> 00:16:28.235
We told you we were not talking about flat files, because we hate them.

00:16:28.235 --> 00:16:32.573
We probably just lost half of our listener base.

00:16:35.466 --> 00:16:44.777
I mean, that's the biggest thing that I took from that conversation with Vanessa and you know, vanessa actually was the one who who made me not scared of flat files in the first place.

00:16:44.777 --> 00:16:50.177
Like I was terrified of flat files for a very long time and you know, flat files are actually.

00:16:50.177 --> 00:17:02.605
I, like you said, john, I think that flat files are where we'll see changes coming from Amazon before we even notice them on the consumer front end or on the back end of the listing.

00:17:02.605 --> 00:17:15.859
And you know, just downloading your category listing report and which is basically that template already filled out with all of the information that you already have on the on the back backend of the listing, and just reading what I like it's, it's, it's.

00:17:16.540 --> 00:17:24.013
I think that the why flat files are so complicated is because it's really hard to understand what all the different columns mean.

00:17:24.013 --> 00:17:34.626
And if you just take time just to read the definitions tab and what all the different things are, I think you'll learn a lot more about what Amazon wants to see today.

00:17:34.626 --> 00:18:18.585
But then also kind of looking into the future, of what's going to be important in the future strategy in terms of how to write your listing, because, yes, speaking to Amazon's algorithm is important, but it's the consumer that ends up buying the product in the end anyways, and I think the example that we were giving on that call was about some sort of a mug for hiking and just Amazon knowing that this person even though there's all kinds of different kinds of mugs, one that you're going to take on a camping trip is very specific.

00:18:18.585 --> 00:18:26.045
But then there's also you're a very specific person that camps and then are going to be seeing these types of other things.

00:18:26.045 --> 00:18:36.926
So, like for me in my listings, I think I've always kind of had that best practice in mind is that I more I more right to the consumer in my copy than I do the algorithm.

00:18:36.926 --> 00:18:56.587
Yes, I do have my important keywords in the important places, but I've always put more of an emphasis on speaking to my consumer and I think that that's going to put my listings in a better place once this whole, you know shift to AI and Rufus, where they're able to type in contextual things as opposed to just typing in a keyword.

00:18:56.587 --> 00:19:01.204
So that's kind of what the biggest thing that I took from that talk with Vanessa is.

00:19:01.244 --> 00:19:02.829
First of all, don't be scared of flat files.

00:19:02.829 --> 00:19:04.273
Learn them Like.

00:19:04.273 --> 00:19:09.582
It is really something that you should at least understand for your own SKUs whether or not you dig into.

00:19:09.582 --> 00:19:11.666
You know kind of what the construct is and others.

00:19:11.666 --> 00:19:33.628
And I've almost gone too far on the nerd out spectrum and downloaded in another categories and just kind of saw, like electronics for example, electronics listings look so much different than any other listing and it was interesting to me so I went and downloaded the flat file after a conversation with Vanessa and I looked to see like what, what are the differences in the flat file that make it show up so different on the front end?

00:19:33.749 --> 00:19:35.392
Yeah, and I think you bring up a good point there.

00:19:35.392 --> 00:19:43.892
Well, that comes to my mind, which is again and I don't want to, you know, overplug some of our previous episodes, but I think there's some real gems in there.

00:19:43.892 --> 00:19:57.941
And the one that also sticks out my mind on this topic, because I asked him about it, was with Abe Tamale, where I asked him you know, hey, what's your take on kind of this AI stuff?

00:19:57.941 --> 00:20:04.496
And you know he brought up a really good point and I think he's got a really good perspective seeing how, I mean, I think he started selling in like 1991 when he was, you know, it's like order catalogs, almost like this year's Robux day.

00:20:04.496 --> 00:20:13.234
So when you have somebody who's essentially, you know, been doing e commerce before e commerce was invented, that's got that kind of 50 year perspective.

00:20:13.234 --> 00:20:17.488
Well, I guess a little bit of public math, 35 year perspective.

00:20:18.269 --> 00:21:01.369
It does help in the sense of, from a customer standpoint, is it's going to take time for people to get used to putting in more phrases and intent behind things, as opposed to keywords that we see right now, because keywords I think everybody that's probably over the age of 25 is used to this whole keywords concept that we see with Google and Amazon and that type of stuff of looking for it, whereas I think the younger generation, now that we have ChatGPT and that type of stuff that are getting better at you, know more of intent based queries and you know we're probably going to see it move that direction over time, but it's going to take time.

00:21:01.369 --> 00:21:10.968
I mean, you know some of us dinosaurs out there may never actually make that full transition over to intent based and still be very keyword focused, but over time we're going to see that shift.

00:21:10.968 --> 00:21:12.590
It's probably going to take a while.

00:21:12.590 --> 00:21:24.857
It's probably going to be years to take time to retrain the US customer base in how they look for things online.

00:21:25.459 --> 00:21:31.921
With that said, I think there's a separate issue of Amazon is how they look at it.

00:21:31.921 --> 00:22:03.429
From a logarithm perspective is already here, and so when we look at you know we talked about a few of those examples as far as you know, the main image that we know Amazon is giving that as input, the flat file but even you know Amazon just recently rolled out a requirement that now you have to have three images, and part of that is because they want to see, you know, other aspects of that product that they're feeding into their AI algorithm based on those three images that are requiring you because they want that more, more of that data to give the AI more cues as to what your product is and where it should be shown.

00:22:05.434 --> 00:22:07.431
You know I actually would.

00:22:07.431 --> 00:22:16.393
I would argue that I think that trend, that transition, is actually going to happen much more quickly from a from a a shopper perspective.

00:22:16.393 --> 00:22:39.394
And I say that only because ChatGPT has become so ubiquitous, and although I understand that there's still a fair number of people who don't use it, who haven't used it, there's actually a much larger percentage of people who have used it than what I think anybody would have anticipated a year ago.

00:22:39.394 --> 00:23:02.665
And the beauty of that is that that is the way that ChatGPT operates, right Like it's not keyword-based, it's asking questions, it's making a query, it's whatever, and so I think the value of that is that everybody who is using ChatGPT for anything is being trained to utilize it in that way.

00:23:02.665 --> 00:23:04.462
In fact, I'll give you an example.

00:23:06.075 --> 00:23:24.460
I was in a chat group just the other day and somebody there was a conversation going on Somebody posted the question to chat GPT and then put the response into the chat, said I went to chat GPT and I said of course, now, why wouldn't you go to chat GPT?

00:23:24.500 --> 00:23:25.000
You should always.

00:23:25.000 --> 00:23:30.087
And he essentially said of course, now, why wouldn't you go to ChatGPT you should always?

00:23:30.087 --> 00:23:32.872
And he essentially said I almost never search Google anymore.

00:23:32.872 --> 00:23:48.395
I just go enter it into ChatGPT and I think that's becoming more and more the case, that people are looking for what they believe is a more robust answer to their questions, and they're using ChatGPT and other AI-based models to do that.

00:23:48.395 --> 00:24:02.839
And even a lot of older individuals are being forced, I think, into using chat GPT, even if it's just within their role at their business, because more and more businesses are basically saying look, this is here.

00:24:02.839 --> 00:24:06.188
We need to be making use of it to make ourselves more efficient.

00:24:06.188 --> 00:24:17.807
If you want to continue in your position, you're going to need to understand how to use this tool to make yourself more efficient, and so I actually feel like that transition is going to happen much more quickly than people think.

00:24:18.816 --> 00:24:34.568
I totally, totally agree with you, and I don't remember who we were talking to recently, in the past couple of weeks that thought the opposite, that it was going to take a long time, and so I'm in that camp and I was going to pose the question to you guys before you basically answered it.

00:24:35.876 --> 00:24:41.808
I can't think of a time that I chose it in the past year, especially the last couple of months.

00:24:41.808 --> 00:24:48.701
I can't think of a time that I chose Google over going to some sort of a large language model to answer a question.

00:24:48.701 --> 00:25:04.538
I personally prefer going to some sort of a large language model in most cases to find an answer to something now and, as a result, the way that I search for something, the way that I ask it a question, is completely different than how I asked Google.

00:25:04.538 --> 00:25:13.025
So I personally am already there, I'm trying to understand, I'm trying to put my mind in the consumer of how that looks on Amazon, because I don't quite understand that yet.

00:25:13.025 --> 00:25:29.249
But in terms of me searching for information, my default has now become AI as opposed to a Google search already, and that that's happened within a span of less than a year for me, google search- already and that that's happened within a span of less than a year for me, yeah.

00:25:29.269 --> 00:25:34.319
Well, I think you know we all agree and I think everybody in the community well, pretty much everybody in the community agrees.

00:25:34.319 --> 00:25:41.281
You know we might disagree on what that timing looks like, but I think we all agree that it's going to have a massive influence on Amazon and and really our entire lives.

00:25:41.281 --> 00:25:58.714
So, with that said, as we kind of wrap up for this episode, for people that are listening, what's one tip that you'd give them on being on the earlier end and kind of taking advantage of what we're seeing as far as AI for Amazon right now?

00:25:58.994 --> 00:26:06.252
Similar to a lot of the things that we talk about, I think, not being scared of something that's new and doing the things that we've already talked about.

00:26:06.207 --> 00:26:07.821
I mean, I think that we've about, I think, not being scared of something that's new and doing the things that we've already talked about.

00:26:07.821 --> 00:26:17.648
I mean, I think that we have all just kind of scratched the surface of what's even possible right now to test inside of the AI that Amazon's already developed.

00:26:17.648 --> 00:26:28.479
I think that you know being the first mover on Amazon on anything that they give us, being the first mover on Amazon on anything that they give us.

00:26:28.479 --> 00:26:39.915
I've watched it happen time and time again how you know, when you're first to a different ad type or a targeting strategy or something along those lines, a lot of things are easier until the masses start moving in that direction.

00:26:39.915 --> 00:26:51.148
So, similar to everything else that we talk about, I think not being scared of it and jumping in to learn as much as you can about it now is going to put you light years ahead of the people that are resistant to change.

00:26:53.155 --> 00:26:56.786
And I would say that, first of all, I agree with all of that.

00:26:56.786 --> 00:27:05.009
I think it's absolutely critical that we don't shy away from those things that, right now, we kind of fear because we don't know what they are.

00:27:05.009 --> 00:27:06.258
We don't really know how to use them.

00:27:06.258 --> 00:27:33.544
It's all very nebulous, but the thing that I would recommend, in addition to that, is structuring your business now, today, in such a way that it makes room and it makes time for you and for your employees of various types to begin to play around with these things like that.

00:27:33.544 --> 00:27:46.978
They have time to learn how this stuff works, to test it out, to see what works, what doesn't, to research new tools that are coming out and new ways to do some of these things.

00:27:48.080 --> 00:27:56.135
I feel like if you don't make space for it, it won't happen, because everybody's just going to be focused on their little thing, and this is the way we've always done it.

00:27:56.135 --> 00:27:58.643
This is the way I'm going to continue to do it, because it gets the job done.

00:27:58.643 --> 00:28:17.385
I think you need to start making a point of making room in every employee's week that they have time, and also in your own week that you have time to address this, and maybe that means you're bringing on additional personnel, whatever that looks like.

00:28:17.385 --> 00:28:25.680
You need to figure out a way to make space in your schedule so that you're doing it, because if there's no space, it's not going to happen, no matter how much you feel like it's important.

00:28:26.342 --> 00:28:27.865
Yeah, it's not a shiny object AI.

00:28:27.865 --> 00:28:31.806
I'm very hesitant to jump into things because I don't.

00:28:31.806 --> 00:28:43.020
I have a tendency to chase shiny objects, but AI is not a shiny object and it's something that you need to be figuring out how to leverage in your business, because you will be left behind without a question.

00:28:44.384 --> 00:29:01.314
Yeah, and so I you know my tip will kind of build on, I think, on both of what you guys have said, which is, if you don't know where to start, start with you know a couple of things Amazon focus that we've talked about, whether that's AWS, comprehend and running.

00:29:01.314 --> 00:29:17.196
You know your title and some of your other text through there to see what sort of results you get back and how essentially that learning model feels about how you describe your products and does that match with you know your target audience, whether that's.

00:29:17.196 --> 00:29:20.445
I don't think Rufus is fully rolled out for everybody yet.

00:29:20.445 --> 00:29:24.250
I think that, but if you do, I mean you should have the app on your phone.

00:29:24.250 --> 00:29:29.565
You should be looking at your products on the app at the very minimum, but then also test out Rufus.

00:29:29.986 --> 00:29:47.662
Like you said, mike, carve out some time to not only look for your products but make it an effort, if you're buying something off of Amazon, to try and find that product or that solution via Rufus, because I think you'll be very surprised and get a lot of great consumer perspective insights.

00:29:47.662 --> 00:30:19.489
And then, at the very least, take that, create a product and test out the AI builder, whether it's with your current main image or a new product that you're launching, and I think you'll be shocked if you put that main image in what Amazon gives you back in some cases, of what it thinks your product is, and pick one of those, spend some time with it and learn from that process, because it's only going to get to be a bigger influence as far as AI, amazon and really everything else we do going forward.

00:30:19.489 --> 00:30:23.912
Yeah for sure, all right.

00:30:23.912 --> 00:30:26.103
Well, that's a good place to wrap for today.

00:30:26.103 --> 00:30:34.201
I encourage listeners take away one of those tips and dive into an AI tool and learn something to be ahead of that curve.

043: Tactic Tuesdays: Amazon AI Optimization

Listen On

Recent Episodes