![EP #357] [ENG] - How Rufus is changing the way you sell on Amazon - Andri Sadlak and Vincenzo Toscano](https://i.ytimg.com/vi/ZMB7YK22u2g/maxresdefault.jpg)
Ecom Podcast
EP #357] [ENG] - How Rufus is changing the way you sell on Amazon - Andri Sadlak and Vincenzo Toscano
Summary
To optimize Amazon listings with Rufus, emphasize who the product is for in main images by including packaging details that highlight target demographics and usage, which aligns with the shift from keyword to consumer-focused shopping trends.
Full Content
EP #357] [ENG] - How Rufus is changing the way you sell on Amazon - Andri Sadlak and Vincenzo Toscano
Speaker 1:
Welcome to The Ecommerce Lab By Ecomcy. This is the place for everything related to Amazon private label and e-commerce. Learn exactly what you need to start or scale your business.
Get insights from the top industry experts who will discuss the latest trends and best practices in the world of Amazon. From choosing products and sourcing from a supplier to setting up your Amazon account and marketing your business,
you will hear it here. Let's get started. Here is your host, Vincenzo Toscano.
Speaker 2:
Welcome back, everyone. We're in Milan with Vincenzo Toscano at the Amazon Ads Educator Retreat. And we're going to speak about what changes came into our world, the Amazon world,
and specifically focusing on how do we optimize our Amazon listings visually with images. So Vincenzo Toscano, the CEO of Ecomcy, please introduce yourself.
Speaker 3:
Thank you, Andri. First of all, a pleasure to be here. So my name is Vincenzo Toscano. Founder and CEO of Ecomcy, where we are a full-service agency for Amazon brands,
where today we have worked with over 1,000 brands and with a team of over 80 people globally. We help with everything from Amazon PPC, DSP, Brand Creative, everything under the sun when it comes to scaling brands on Amazon.
I'm very excited to be here today because, as you mentioned, there's a lot of things that are changing,
and I'm sure people are going to get a lot of insights when it comes to things that are coming so they can be ahead of the competition and, you know, crushing in 2026.
Speaker 2:
Absolutely. That's the plan. And my name is Andri Sadlak. I'm currently the founding head of product and strategy at Azoma.ai, which is AI visibility software for ecommerce.
And we actually focus on figuring out ways to optimize for Cosmo and Rufus and showing up more often and higher up in ranking in LLM results, ChatGPT, Gemini, Rufus. Walmart has its own AI-powered bot as well.
So all of that together, learning every day. And we wanted to kind of give you a glimpse of what the visual optimization on Amazon specifically has to look like. Maybe we can start with the main images.
Any advice you have in terms of how to optimize main images, knowing what we now know in 2025 coming to 2026.
Speaker 3:
Yes,
I would say a lot of the changes we're seeing right now when it comes to Rufus is more important than ever to really understand who is your consumer because we are now shifting in terms of when it comes to how people shop with Amazon from being keyword focused.
It's actually being more around who you are as a person in terms of your history of shopping on the platform, and on top of that, how the potential product that you're going to be buying is going to be correlated to your needs, right?
So going back to your main image question, I would say that one of the most important things you have to make sure that you are covering on the first image is reflecting exactly who this product is for, right?
And this could be done by a series of ways, right? So the first one is, As simple as the packaging, right? Because a lot of times I have this conversation with clients and they don't include the packaging on the main image.
And that's a huge opportunity for Rufus to essentially get insights about why your product is in the first place. And the reason is because your packaging has a lot of insights when it comes from a text perspective.
We know Rufus and Cosmo are all now can analyze that information. And it comes down to, for example, your packaging, if you're mentioning things such as what kind of, for what gender this kind of product is,
or what is the age that it's for, what is kind of the usage or the lifestyle, like, oh, so many of the connotation that might essentially make the product very specific.
And by having that information in a text format, now when somebody, let's say you're selling A water bottle, right? And you have in this water bottle and compared to your competition that's only maybe put in the water bottle,
you're now actually also putting on the main image your packaging. And the packaging,
let's say your niche is people do hiking and this water bottle is for people do hiking because has certain a material resistant for impact of keeping the water colder for longer hours.
If you mentioned that on on the description and the packaging and also even the image within the packaging itself is maybe has to do with a forest or something that can be recognized from an image recognition point of view when somebody is not happy and say I need a water bottle from.
Hiking, most likely your product has a higher probability to be recommended than the person that just put in a plain image of a water bottle.
So that's a clear example of how you can reverse engineer the needs of your consumer and making sure that images essentially make the journey for Rufus as easy as possible to essentially recognize your product as a better fit for the consumer.
Speaker 2:
Yeah, I think you're addressing two important points here. And one is something we tested with the Amazon's lead marketing manager, I think she is, who actually co-developed or led the process of developing this new type of advertising,
which is sponsor brands, prompts, sponsor product prompts. And we spoke about Rufus a lot with her and me, Max, Sinclair and this person from Amazon. We showed our phones,
looked at the same product and looked at the prompts suggested by Rufus and they all differed because of our shopping behavior.
Speaker 3:
Yeah, your history as a consumer.
Speaker 2:
Exactly. So that proves the point that Rufus is here to be your personalized shopping assistant. So it's not the same approach as before where you only go after the bestseller or Amazon choice.
You have to optimize for your ideal customer persona and Rufus will find the match because that's the job of this shopping assistant is to find that shopping intent connected to the Cosmo base,
knowledge base that's built by you as a brand and find the perfect match that maximizes conversion.
And I think the second part that you spoke about is how to include the elements of the image that will help us convert and also help us get recommended by these AI algorithms and specifically Rufus and Amazon.
Because Rufus has what is called OCR, which is Optical Character Recognition. It can read the text on the image. So if you do put the packaging on the side or behind the bottle, And it says what the product is, which is critically important,
who it is for and what situation it can be used. It can be read by AI and interpreted correctly. And I'll give you a quick hack. The easy way to check that is when you go to catalog in Seller Central and you click add product,
there's a way to upload an image and click generate content. What it does is Amazon's AI reads the image, tries to understand what it is, and creates the title.
So it's an easy check to see if your image is clear enough to describe what's on the image. And if it's completely off, or it's not really exactly what it is, look for ways to change the image,
add some elements, clarify things, different angles, different lighting, different packaging, text, or whatever we use as a trick, and upload it again and see if it changes, right?
Any other methods to kind of validate if the image is AI recognizable?
Speaker 3:
Yes, I would say another way, I mean, even within AWS, right? You can open an account right now. You can get access to Amazon recognition, right? And Amazon comprehend and specific for the image of things with Amazon recognition.
What we actually do is we upload the images you can then upload to our listings. Beforehand, an Amazon recognition will tell you exactly what Amazon can read on the image from a text perspective,
from the lifestyle that is being portrayed on the image, even things such as emotions, like a quick example that I can remember.
We had this supplement that is heavily triggered towards essentially helping for sleep, right, and kind of relaxation.
And a huge uplift that we saw in terms of traffic and conversion was as simple as the person that was used on the main image and was also part of the lifestyle that had to do essentially with the feeling of what they feel when using the product.
Speaker 2:
Relaxation, calmness.
Speaker 3:
Exactly. And we put that on Amazon Recognition and we could see how now Amazon was recognizing that. And now when somebody's typing, I need this supplement of vitamin to relax for X,
Y, and Z, or pro was coming up because the image portrayed that feeling. So that's a clear example of how you can reverse engineer, essentially, you know,
testing these images, see what Amazon thinks, and then use that to leverage your exposure on the platform. And I will say a tip to also figure out a lot of these kind of loopholes,
we can call it, to maximize what you can achieve with Rufus. Have a look at your competition. A lot of times, and I think this is a big opportunity when it comes to launching now on Amazon, if you know how to reverse engineer this,
it's like we're entering a phase that is not anymore about organic ranking. It's about who is able to convey the offer the clearest in terms of to the right consumer.
What I'm trying to say now is a lot of the main bestsellers on Amazon Choice now, they have sometimes basic images. They became very complexion because they've been selling for 10 years now. Getting all the sales, but a lot of them,
they are missing out on the negative reviews where people see that this product wasn't meant for things like that. And you can now reverse engineer those competitors,
make your images really clear in terms of what is the confusion that's still happening on that niche. And essentially, over time, Groof is going to recommend yours over the best sellers.
And that's going to allow you to overcome some of these monopolies that the organic ranking has created in some of the niches, you know.
Speaker 2:
Yeah, I think what we both refer to and what I love about the recent changes and what I actually call the Amazon 3.0 is even if you don't end up on page one for some of the keywords,
even if you're never going to be the bestseller or even get the Amazon Choice badge, you can still be matched With the right audience that you're serving with your products,
if you optimize properly, and essentially how you do it is, first of all, think of Cosmo like the brain behind Amazon. You have to answer 15 relation types that Amazon openly shared in their white paper about how Cosmo works.
If you answer all of these questions as much as possible, as much as it relates to your specific product, then at least with a copy, there will be information about what is your product,
who it is for, in what situation it's going to be used, and with what device it's compatible, and so on and so forth, that answers all of the major questions. Now, the second layer is the visual layer, which is what we're focusing on.
And what's critical, something you brought up, is You have to be consistent, like the whole listing has to match up the information. Everything that you see on the images, say, talking about audiences,
if it is a lifestyle image and we're selling a water bottle for hikers, don't show a picture of an older lady on a treadmill, right? Because it's going to be a mismatch.
And then Rufus reading all that information from the whole database on Amazon, including Cosmo, will be confused. So who is this exactly for? And even if you have multiple different audiences,
you have to be very clear in copy and visuals that it caters towards A audience, B audience and C audience. So there's no confusion that it's not For the person talking to the engine, right?
Speaker 3:
Yeah, so I would say you have to think of this as a dating kind of platform in the sense that let's imagine you're trying to find the perfect date, right?
And you have certain characteristics, certain, you know, things that you're looking for. That's exactly what Amazon is trying to achieve with Rufus.
It's trying to do the performance match between the consumer And the end product in this case, so you need to make sure you put as much information as possible. So when Amazon does this match,
what Amazon is trying to do is make sure the probability of making a sale is as high as possible so the consumer doesn't go somewhere else. So that's why put a lot of focus when it comes to really niching down your content, your copy.
And a lot of times people come to me and say, yeah, but my product can be used for not only hiking but also swimming and other things, like a good hack that we have done this already with some of our clients.
We use a strategy which is kind of. It's been in the space for a while, but now it has to do with the Rufus side of things, which is product cloning, but from a content perspective. So sometimes it could be exactly the same product,
but it could be that one variation is all about swimming, and that's it. And the next variation is the same product, but maybe just the packaging changes, and it's all about hiking. And therefore, what you do is that with each variation,
you make sure that Rufus bring only people that want a swimming water bottle to this thing, or hiking water bottle here, running water bottle here. Right, I'm trying to convey all the lifestyle in one listing, which at the end of the day,
what Rufus is going to end up doing is becoming preferential to one lifestyle and then essentially kind of negate the rest, which could also be consumers for you. So that's a hack.
Speaker 2:
Yeah, totally. And practically speaking, the information that is being taken from the images is text, objects, people, their emotions, as you mentioned.
Speaker 3:
Even the video is actually picked up, like the video.
Speaker 2:
Same with videos.
Speaker 3:
You put a video that's actually also, so if you're saying something on the video or anything, that's also being picked up.
Speaker 2:
Totally. So the goal is answer the frequently asked questions. Not just in your copy, but visually. And very often, if you already use Rufus, you can probably confirm. You go through Rufus, ask some questions, and the image pops up.
So the answer comes visually from the listing images that you as a brand uploaded. And that's where you want to be. Because an image is worth a thousand words. Video is probably worth a million words.
So that's the way to optimize going forward. It has to be human. And the way to think about Rufus is like a shopping assistant. Back in the day, we would go to a department store and we would talk to whoever works there,
like what's the best What a bottle for me. And they'll ask you questions, and you'll answer and they'll narrow it down to what works for you. That's the job of Rufus, which I think is genius.
Even from the last conversation, Andy Jassy, the CEO of Amazon had recently, he shared crazy numbers like 250 million people have used Rufus already in 2025. And we're not like Still have a month and a half.
And 60%, he said, this is insane, 60% of people who use Rufus converted better. Yes. Something like that. I don't remember the exact phrasing, but obviously, if you talk to Rufus, you're going to be more confident about your choice, right?
Speaker 3:
Yeah, yeah. I think that's essentially the conclusion we're trying to get to,
which is you need to make sure that we start focusing on And I guess it goes back to a lot of what we have most likely some of you see in the space the last couple of years,
a lot of the strategies came down to try to be appealing to as many people as possible, keyword stuffing, target every single keyword, go as broad as possible. But unfortunately, that doesn't work anymore, because if you do that,
what it's actually going to cause is create more confusion. And over time, that's going to allow you, even from a ranking perspective, get less visibility. And I think a tip actually that has to do a bit with this,
a lot of times, Simple things such as A plus content. We see this mistake happening a lot that a lot of people when it comes to A plus content, especially people have a big catalog.
Right now they're doing the mistake that in the A-plus content, they are not doing product specific A-plus content, but they're doing brand overall.
And what we're seeing, actually, that's also creating underperformance event when it comes to roofers because if you're in your A-plus content now, on a SKU, you're mentioning multiple products from your brand or maybe different lifestyle,
different things that have nothing to do with the SKU that you put in the A-plus content. We're seeing Rufus also actually not being performing very well. So that's another tip I advise everybody.
Like, make sure you look at your A plus content and make sure that now it becomes skewed specific, just as the example I gave you. Like, if you want to target multiple audiences,
go down the route of creating other variations and make them only those variations specific to those different.
Speaker 2:
Yeah, I love that advice. Speaking of A plus, another thing that we know Rufus is looking at is the alt text. So behind every image you upload, There's a 100 character limit field you can fill out explaining in human language,
don't keyword stuff it, in human language you explain what's in the image. And that trains Rufus as well because it's looking at that old text and the image to understand exactly what's happening.
It's training on what every phrase, every situation looks like visually in your specific brand style. And it's able to learn and recommend products based on these situations. So if you have, say, hiking, seeing on an A+,
explaining how good the water bottle is or whatever the benefits are, make sure to explain it in the ALT text as well. You reinforce the value and make it easier for Amazon to recognize what's happening, right?
Speaker 3:
Yeah, and I think another advice I will give when it comes to the optimization of images as well. A lot of times when it comes to the process coming up with these ideas, people also struggle when it comes to, you know,
where are the things in the first place that you have to be adding to this image because yeah, it's easy to say, be specific about what kind of a person I should put, what kind of Emotion, I should reflect on everything.
So I just want to double down that, of course, one is sources reviews. That's the first one. Analyze reviews of the competitors and also some of the pre-made questions Rufus is doing on our listens. But the second thing I love to use as well,
and this comes down to bringing also external traffic, is how I look at, for example, on things outside of Amazon, like also research a bit. What are some of the trends that are being Heavily relate to your type of product.
Maybe you're seeing that for your product. What is working? The best is people that niche down to in the height in the hiking kind of approach for the water bottle. Make sure you use ChatGPT. We use it a lot.
We're actually an advisor when it comes to ChatGPT. We create custom ChatGPTs for every single brand that we manage. And we upload not only the reviews, that's the first part of the advice, but also insights from social media, trends.
Speaker 2:
Social listening studies.
Speaker 3:
Yeah, anything that has to do with essentially the reflection of the market.
And we use that brand to come up with templates of what the image should reflect and everything because it's a good starting point to essentially make sure it's a target as well.
Speaker 2:
Yeah, I think to wrap this up, I'd love to ask your opinion on the adoption of Rufus. Say 2026, we're almost there. What's your bet? Will people be using it more and increasingly more? Do you think it's going to slow down? And knowing that,
how would you prioritize your attention and maybe even budgets in terms of system optimization on methods that help optimize for Cosmo and Rufus versus traditional branded, nice quality content?
Speaker 3:
Yeah, so I would say, as a prediction, I would say yes, people are going to use it more than ever for a simple reason. In reality, most people are lazy, in the sense that when it comes to shopping,
people don't want to spend a lot of time scrolling, trying to find the perfect product. So we're seeing that's why Roof is having such a high adaptation rate, because people just want to say, what is my issue? What is my need?
And just who I am as a person, not too much. I don't want to be looking at reviews, questions, like no, yes, this is what I need, find it for me. So that's why it's going to get more and more prevalent. And I would say in terms of focus,
I would say First plan of action I will do before the year ends. Create an in-depth analysis of what the competition is doing with the images, what the market is saying on the reviews, reverse engineering this with, as I mentioned,
with your research outside of Amazon with things such as ChatGPT or anything that will give you insights about the trend.
Create like a baseline of what the avatar should look like and use that as your foundation to start split testing your images and everything into Q1 2026 because That's another thing I wouldn't advise to be super experimental right now with Q4 because it's a business time of the year and you want to work on what is already working but for sure starting January you need to take action as soon as possible.
Yeah, that would be my take.
Speaker 2:
Totally, yeah, I love that advice. I think something to add to it, with the current pace, like if we even compare, say, ChatGPT adoption to, for example, mobile phone adoption.
Speaker 3:
Yeah.
Speaker 2:
Mobile phone reached 1 billion users. Believe in 10 years and ChatGPT is at 1 billion coming up to three years. So under three years, so it's going to be a billion in three years, basically.
So it's pretty fast, which is a signal that humans started interacting with the web in a different way. So a lot of people are now just talking to their phones, even to Rufus, saying, hey, I have this problem. Suggest what is the solution?
And they take a picture of And I think they've seen somewhere and uploaded to Rufus asking, hey, find this product or something similar. I want to have the same setup.
And Rufus finds that and puts that into your cart, which makes it so much easier. And as you said, people want limited friction as easy and fast as possible.
An average person, I think, Spends about two minutes to find the product from the idea to buying.
So you have to kind of play along and be optimized for the new ways people communicate with the web and the way that the trend is shifting towards more gigantic shopping. Yeah, I think this is a great segue into the question.
Where do people find you, Vincenzo?
Speaker 3:
So, yeah, you can find me as Vincenzo Toscano on all the social media channels. And if you want to have a look at our agency, ecomcy.com,
More than happy to jump on a call with me or my team and figure out how we can help you with everything related to Rufus and Amazon overall. So thank you for having me, Andri.
Speaker 2:
Amazing. Yeah, my name is Andri Sadlak. I'm the only one with that name in the world. You can find me on LinkedIn as well. I often share stuff I learned and I'm definitely excited about the agentic shopping.
Speaker 3:
Thank you.
Speaker 2:
Cheers.
Speaker 1:
Thanks for listening to The Ecommerce Lab By Ecomcy. Be sure to subscribe so you don't miss an episode. While you are at it, we would appreciate it if you could leave an honest rating and review on Apple Podcasts,
Spotify, or wherever you listen. That will make it easier for others to find out about the show and benefit from it. Want more? Visit our website at www.ecomcy.com where you can get your first consultation for free or find us on Instagram,
Facebook and LinkedIn at Ecomcy.
This transcript page is part of the Billion Dollar Sellers Content Hub. Explore more content →