Christian Umbach is the Co-Founder and CEO of AutopilotBrand.com, an AI brand operations platform that helps Amazon and Shopify brands manage their catalogs, optimize listings, and win in AI discovery – on & beyond Amazon. Before Autopilot, he co-founded an enterprise a data orchestration platform, built products at Uber’s API Platform, co-created the Lufthansa Innovation Hub, and earned his Master’s in System Engineering from MIT.

He also writes The Order (theorder.press), a weekly newsletter read by 5,000+ eCommerce operators and marketplace sellers. Chris splits his time between Portugal and the US and has spoken at various Amazon industry events and the World Economic Forum. He loves to work at the forefront of what’s possible with data.

Highlight Bullets

> Here’s a glimpse of what you would learn….

In this episode of the Ecomm Breakthrough podcast, host Josh Hadley speaks with Christian Umbach, CEO of Autopilot Brand, about the changing landscape of Amazon product discovery. Christian introduces the concept of a “third shelf” driven by AI tools like Alexa for shopping, which recommends products differently from traditional organic results. They discuss why conventional rank tracking is becoming obsolete, advocating for market share-based measurement instead. Christian advises brands to optimize listings for multiple buyer personas, update content seasonally, and build AI-friendly websites to capture traffic from tools like ChatGPT, positioning themselves for the future of AI-driven e-commerce discovery.

Here are three action items that Josh identified from this episode:

Timestamps:

00:00:00 The Inaccuracy of Rank Tracking
Traditional rank tracking is an inaccurate success metric; brands should use market share-based tracking instead.

00:00:25 Podcast Introduction
The host introduces the podcast and guest Christian Umbach, an expert in AI, data, and e-commerce brand operations.

00:03:04 The “Third Shelf” of Product Discovery
Amazon is entering a new phase of product discovery driven by AI like Alexa, creating a “third shelf” beyond organic/paid.

00:04:25 Alexa’s Recommendations vs. Organic Rank
Alexa’s top product recommendations are often different from the top ten organic search results, signaling a fundamental shift for sellers.

00:06:06 Why Market Share Tracking is Better
Traditional rank tracking is unreliable due to personalization. Market share tracking using brand analytics data provides a more accurate picture.

00:07:35 Alexa Shopping Study Results
A study reveals 64% of Alexa’s top ten recommendations are outside the top ten organic results, creating opportunities for challenger brands.

00:10:47 Can You Buy Your Way into Alexa Rankings?
The study shows Alexa’s recommendations are currently driven more by organic factors and shopper intent rather than paid advertising placements.

00:12:24 Alexa’s Recommendation Strategy
Alexa often anchors its top recommendation with a safe, high-ranking organic result, then diversifies with lower-ranked or off-page-one products.

00:14:46 Examples of Diverse Alexa Results
Christian shares examples for “cotton sheets” and “picture frame,” showing how Alexa’s suggestions differ significantly from organic search page rankings.

00:17:36 How to Optimize for AI Discovery
Brands should focus on optimizing their catalog, especially titles, item highlights, bullet points, and descriptions, to appeal to AI.

00:21:03 Crafting Your Listing for Personas
Instead of keyword stuffing, brands should understand their different buyer personas and reflect their specific use cases in the listing copy.

00:23:09 An Actionable First Step for Brands
Analyze customer reviews and brand analytics data with an AI tool like Claude to identify personas and rewrite listing content.

00:26:20 The Importance of Seasonal Content Updates
Brands should consider updating listing content seasonally to align with changing buyer intent, such as marketing nursing clogs for gardening.

00:28:57 Most Influential Content for Alexa
Written content like the title, item highlights, bullets, and description is currently most influential for AI assistants like Alexa/Rufus.

00:31:12 Ranking in External AI Search (ChatGPT, Claude)
Amazon-native brands are often invisible on external AI search tools like ChatGPT, representing a massive, untapped discovery channel.

00:36:16 The Growth of AI-Referred Traffic
AI-referred traffic is growing exponentially, with Shopify reporting 50% month-over-month growth, making it a critical new channel for brands.

00:37:32 Using “Agent Sites” to Win AI Search
Brands can create “agent sites”—simplified website mirrors—to feed AI crawlers with persona-specific information and improve external AI visibility.

00:44:29 Tracking and Attributing AI Traffic
Tracking AI-referred traffic is still evolving, but Amazon is beginning to build programs to attribute sales from external AI sources.

00:46:26 Three Actionable Takeaways
The host summarizes the key takeaways: stop tracking organic rank, optimize listings for personas, and build an external AI strategy.

00:49:22 Lightning Round: Best Book
Christian recommends “Principles” by Ray Dalio for its insights on building a successful company culture through radical truth and transparency.

00:51:04 Lightning Round: Favorite AI Tool
Christian’s favorite AI tool is Claude, which he has integrated deeply into his personal and company workflows.

00:52:49 Lightning Round: Ecomm Operator to Follow
Christian recommends following Stefan Haney, a former Amazon leader, for his deep understanding of Amazon’s foundational principles.

Resources mentioned in this episode:

Tools and Websites
“Autopilot Brand“: “00:01:14”
“ChatGPT“: “00:05:06”, “00:21:03”, “00:31:10”
“Claude“: “00:05:06”, “00:21:03”, “00:31:10”
“Google Analytics“: “00:39:02”
“Shopify“: “00:49:14”
“Everything Machines“: “00:39:02″Books
“Principles by Ray Dalio“: “00:49:27″Videos
“Ecomm Breakthrough YouTube Channel“: “00:08:43″Additional Resources
“The Order”: “00:01:14”
“Cosmo Paper“: “00:23:36”

This episode is brought to you by eComm Breakthrough Consulting where I help seven-figure e-commerce owners grow to eight figures.

I started my business in 2015 and grew it to an eight-figure brand in seven years.

I made mistakes along the way that made the path to eight figures longer. At times I doubted whether our business could even survive and become a real brand. I wish I would have had a guide to help me grow faster and avoid the stumbling blocks.

If you’ve hit a plateau and want to know the next steps to take your business to the next level, then email me at josh@ecommbreakthrough.com and in your subject line say “strategy audit” for the chance to win a $10,000 comprehensive business strategy audit at no cost!

Transcript:

Christian Umbach 00:00:00  Ranking and rank tracking has become very inaccurate as a measure of success for for brands. And on the positive side, brands, at least from a data perspective, have everything in place into throw out their ranked tracking and basically perform a market share based tracking instead.

MC 00:00:25  Welcome to the Ecom Breakthrough podcast. Are you ready to unlock the full potential and growth in your business? You’ve already crossed seven figures in sales, but the challenge is knowing how to take your business to the next level.

Josh Hadley 00:00:38  What happens when you analyze more than 70,000 Amazon Ace ins to uncover how customers actually discover products, and how AI powered shopping assistants like Alexa are changing the entire buyer journey on Amazon. Well guess what? Today’s guest has been doing exactly that, and he’s going to share with you how you can boost your AI discoverability? Welcome to the Econ Breakthrough Podcast. I’m your host, Josh Hadley. I scaled my own brand from 0 to 8 figures in sales, and now my mission is to take it to over nine figures on my journey to nine figures.

Josh Hadley 00:01:14  I bring you the unfiltered conversations with the smartest minds in e-commerce. Past guests include Ezra Firestone, Kevin King, and Michael E Gerber, author of the E! Myth. Today, I’m excited to introduce you to Christian Umbach. He is the co-founder and CEO of Autopilot brand, an AI brand operations platform that helps Amazon and Shopify brands manage their catalogs, optimize listings, and win in AI discovery on and beyond. Amazon. Before autopilot, he co-founded an enterprise and data orchestration platform, built products at Uber’s API platform, and co-created the Lufthansa Innovation Hub and earned his Masters in System engineering from MIT. AKA he’s smart. He also writes the order. The order Dot press, a weekly newsletter read by over 5000 e-commerce operators and marketplace sellers. Kris splits his time between Portugal and the US and has spoken at various Amazon industry events, the World Economic Forum. He loves to work at the forefront of what’s possible with data. With that introduction, welcome to the show, Christian.

Christian Umbach 00:02:26  Thanks, Josh. Excited to to join you here.

Josh Hadley 00:02:29  Well, you’ve got a long history of working in data and systems. And before we hit the record button, I know you’ve worked a lot with a lot of the aggregators. You’ve worked with a lot like 70,000 Amazon ACS, like you’ve seen a lot of data. And so I’m curious right now, Christian, based off of your years of analyzing Amazon Aces and things like that? What are you seeing happening right now? That is the biggest like change that our listeners need to be aware of.

Christian Umbach 00:03:04  The biggest change today that we’re seeing is that Amazon is entering or has entered really another kind of step function in terms of discovery and, user interaction, user experience for, for shopping. we call it the third shelf, basically organic being the first, paid being the second. And what we see with Alexa for shopping now being the third, and you could say as an extension, product discovery within tools like ChatGPT, Gemini and so forth, kind of also falling into that, third shelf framing, where the the way this shelf operates is different from the, the first two.

Christian Umbach 00:03:51  we’ll get into some, some of the numbers and details around that, but I think that’s one of the the biggest shifts because it actually, you know re re bundles reframes competitive dynamics. the, the rankings, the competitive positions that are largely in place today and obviously it’s somewhat fluid become very fluid on on the shelf. which I think is great news for, challenger brands that are sort of up against the the incumbents out there.

Josh Hadley 00:04:25  Yeah. When we were talking previously, what I was astonished by is the stat that you shared that if you look at the top ten organic results and then you ask Alexa for the top ten of whatever it is that you’re searching for, you’re not getting the top ten organic results. You’re getting a whole different list of top ten, whatever it is that you asked about. And so that is going to be a very, very interesting like I almost like liken this to we used to always go to ask Google and more and more very few people are going to Google to just ask the question.

Josh Hadley 00:05:06  They’re going to go to a ChatGPT or a clod, because it’s going to not only answer the question, but then give you step by step solutions to actually resolve or answer whatever it is that you are asking. And I think that’s the same change that’s about to happen on Amazon. It’s already underway. But I think Christian like over the next year on Amazon, I think the way that we’ve been ranking it organically and things like that is just going to fundamentally change in the way people are shopping. Like if you’re just if you’ve built your entire business based off of, oh, I know the A9 algorithm, I know how to get ranked. I think that goes away to an extent, definitely over the next five years. But I think we see significant changes in the next year. Agree or disagree on that.

Christian Umbach 00:05:54  Agree. And on on two levels. One the first one, and we’ve been quite vocal around that for, for a long time. And you know, this is like leave Alexa for shopping aside for a minute.

Christian Umbach 00:06:06  Like just kind of core, old school search basically on on Amazon. ranking and rank tracking, has become very inaccurate as a measure of success for, for brands. and on the positive side, brands, at least from a data perspective, have everything in place, in to throw out their rank tracking and basically perform a market share based tracking instead. you know, the brand analytics data that Amazon makes available, exactly. Allows for, for that type of analysis. and I think it’s important because this notion of, basically leveraging scraper from that’s probably rotating through IP addresses and so forth, like throughout the globe. trying to see like, you know, or are you position, you know, six today or are you position eight, tomorrow. And, you know, what’s kind of changed in the competitive dynamics like that part has largely already gone out of the window because Amazon actually customizes a good chunk of the ranking results based on your shopping history, your location, availability, and so forth. so, there we are already a big fan of the market share based tracking have been like applying that on our end.

Christian Umbach 00:07:35  And then outside of that now take that to the Alexa for shopping side. I think that’s where it gets interesting. And what we found in, in our study, was basically that it is not a reflection of just the organic search results, but basically that. If you look at if you try to get like, say, ten product recommendations from Alexa for shopping, out of those ten product recommendations, 64% are going to lie outside of the top ten organic recommendations, and 40% of your recommendations are even completely outside of the first page results of, usually about sort of 48 or organic or sort of 30 to 48. so that for for us was a big takeaway. And maybe just to throw in a couple of visuals here, here and there as well. kind of let me share one, around that.

Josh Hadley 00:08:43  For those that are just listening to this, make sure you come check out E-com breakthrough on YouTube, where you will, see what Christian is going to be sharing here. And Christian, I’m really interested to dive into this because like my bigger question is, since I’m a seller, it’s like, so how do I influence this, right? Like if the whole like ranking thing is changed, like, what do I do now? So share.

Josh Hadley 00:09:06  Share with me the good word here.

Christian Umbach 00:09:09  So yeah, I guess let’s just set the stage a little bit since we have a few minutes here on how we approach this in the first place. We basically, for the study that we ran, which we ran primarily in June, I think the last week of May as well. So it was post the Rufus to Alexa move. you know, it’s I would say it’s still outdated today because just, you know, Alexa itself evolves and things, you know, without the underlying, frameworks involved as well. but it’s still, I think, a very relevant kind of snapshot in time. And the way we approached it, we took close to 2000 non branded queries. We ran all those queries through traditional product search and through Alexa search side by side. And we basically then the aim was to record usually around ten Alexa recommendations and record the full product results page and try to get an understanding of what’s the the relationship between the two. Like, is there, what’s the relationship between the two results and the organic, results from the search page? Is there a relationship between the advertising side and what are additional cues that we can kind of see in the data? And what we what we found was we’ve talked about this like this bigger split of 64% of the top ten Alexa picks being outside of the organic top ten.

Christian Umbach 00:10:47  and and maybe that’s even more drastic in that sense. Like 41% of the all of our shopping top ten picks being outside of the first page, results overall. which basically is, I would say, a huge emphasis on and a huge opportunity for, for challenger brands because it means that, there are other factors, you know, that around personalization, kind of around shopper intent that are coming into the equation, which are not based, baked into the, the organic, you know, search algorithm today that, that Amazon operates on. you know, then one additional point that we looked at to basically the question of can you, can you buy your way into Alexa and maybe the the answer there is like, you’re going to be able to buy your way into it more and more over time. but at least, you know, as of June here, when when we took this measurement, it was not the case. So especially, in terms of, like, picks that are selections that were, add only and, basically didn’t have an inorganic ranking itself from which we basically concluded that, Alexa at that point was still very much driven by, you know, the organic and the shopper intense side more than by advertising the others.

Christian Umbach 00:12:24  And then just, I guess, sort of seeing this as a, as a scale over time and kind of where, where it anchors like the the top result that Alexa gave was still, I would say within the, organic top ten, you know, about sort of 55% of the time here. But then you would see this go down over and over again, which is, I think, one interesting element of, of this view is it really matters on how Amazon evolves the user experience around Alexa itself. Like, is it designed to give you a single product recommendation is a designed to give you your three best choices, your five best choices. and you know, with that, basically you can kind of get a sense of where this lands on, on sort of the organic slider here.

Josh Hadley 00:13:23  Yeah. No, that super interesting. Well, I love that. It’s like it anchors on a safe winner and then it diversify. You’re saying like the lead slot though. It’s like rewarding like new brands.

Josh Hadley 00:13:35  Like how do you know that. So like it’s like Amazon’s not just relying on the best seller I guess.

Christian Umbach 00:13:41  So the I guess what we’re seeing here saying here is the the lead slot. 55% of the time was, also in the top ten of the organic search. Okay. we didn’t check in terms of, I guess newness or freshness of that product here. but it’s rather if we look at, you know, let’s say the I mean, we for our data here looked at the top ten results, usually for, for for Alexa. But even if you just look at that top five, it’s fair to conclude that basically there is more diversification coming into the picture then, which for us kind of quantifies in that number that lace across the top ten, 41% of the results, are not on the first page at all. So it’s yeah, it’s not that you had to build up your way to be on the first page, but but it’s other measures that are coming into play. and those are very, quite a bit by, by terms.

Christian Umbach 00:14:46  So I can let me jump into a couple of concrete examples here. So, I think one of the most diverse ones, probably in terms of kind of going beyond organic, was the term cotton sheets. so for, for cotton sheets, lane linen kind of was the number one suggestion and actually is or at that time was the number four organic pick, that we received. but a lot of the other ones here that were suggested were not on the first page of results at all. and yeah, so I think this one was like heavily leaning towards, sort of a diversified picture. and then you had other categories that are a bit more like in this case, I guess, a picture frame here. we’re kind of picked within the organic side, but it’s clear that it does not. I would say go through organic ranks, and just references the back, but rather has kind of core underlying mechanisms. And, yeah, I think that’s the full mix of like reviews playing into that. the ratings.

Christian Umbach 00:16:07  and then obviously, some of the core copy around, like, does that reflect shopper intent and, you know, kind of cast that wider net that Amazon can actually see. Okay. Like it’s yeah. Josh is searching for 11 times 14 picture frame here. We also know this about your past purchasing behavior. How does this fit to kind of this broader set of what’s out there. and can be suggested to you as a product?

Josh Hadley 00:16:36  Yeah, I mean, definitely like, this is very compelling because like Amazon as you’re seeing, is not recommending the number one best seller. Both of these examples, it was okay, organically ranked number four. You’re the number one recommended guy. And like your your questions were very basic like 11 by 14 picture frame. Like that’s a fairly broad search term. It’s not like very granular. And then from there you have like organic ranking number 26 shows up as the number two guy on Alexa. The number 12 organic ranking shows up as number three. And then you’ve got two other guys that don’t even show up on page one at all.

Josh Hadley 00:17:13  And they’re they’re included in this list. So my bigger question is this question like obviously things are changing. Amazon’s going to be shifting things a lot. So what should I do as a brand owner. Give me some specifics that like Josh you need to go start updating your title, your bullets, your image. Like what is it that I need to be doing as a brand owner now?

Christian Umbach 00:17:36  Yep. So in terms of the activities there, the the catalog side is something that you have just very easily in your control. There’s obviously things that play into this, like reviews and ratings and so forth, which are in some ways oftentimes secondary effects of your activities that you’re taking. But the nice thing about starting with the catalog is it’s fully in your control. and it does have a major impact on things. one is Amazon. yeah. At the end of July, had the change from long title to short title plus item highlights. which for most setters was more of a okay, let’s take the long title and split it into two for some categories actually opened marketing space.

Christian Umbach 00:18:22  So if we think think about specific apparel products that were capped at 125 characters before they actually won some space, and we overall see good indications that it’s playing out in a positive way from a conversion perspective. but that’s, you know, sort of top of funnel, but it’s a sort of top of visibility. But it is indicating like, I think Amazon’s push for, hey, talk to me in benefits and like, why what why someone should buy this product. Like how does better what is kind of its core feature that’s standing out as a differentiator versus the competition? And, you know, putting that at a very visible top of what you’re presenting from your product and then basically going deep. So we, you know, with our brand partners, we use the bullet points, we use the descriptions very, very heavily. to kind of, I would say cover that we present it as a spiderweb of like, who’s your who are your core buyers? Like, why are they buying it? What are they using it for? you know, know, when are they buying it? What’s the kind of intent buy season and so forth? you kind of which are essentially the underlying like Rufus or Cosmo questions.

Christian Umbach 00:19:47  kind of one and a half year basis cover there to basically, I would say cast a pretty wide net in terms of saying like, okay, these are all great things my product can do for all these different use case areas. and feed that in both, you know, through the description and through the bullets, like in in the text air, but also then in the underlying attributes where, there’s a lot of attribute fields. And I think most setters, the first thing that they need to do is kind of let’s make sure the attributes don’t contradict each other and like don’t create a problem like returns. and then the next step is let’s let’s optimize them and, you know, signal the algorithm that, you know, there are all these different use cases for for the product.

Josh Hadley 00:20:35  So what does that look like if I want to get ranked in Amazon Alexa for 11 by 14 picture frame right. But it’s fairly broad. Like what’s your initial recommendation. Like is it making sure that 11 by 14 picture frame is in my title or the bullet or images like where where are you seeing like the biggest influencing factors of getting ranked in Alexa are.

Christian Umbach 00:21:03  Well, the I think like, yeah, if that’s the product then certainly put it in the title. It’s just kind of from, from a, from a clarity perspective that that will help and clarity drive sales and drive sales ultimately also drive everything that goes into Alexa for shopping. but the the general position I think that I would recommend there like versus traditional optimization is like really go through kind of this thought experiment. Both thought experiment and data analysis of understanding who are your different buyers. and how does again what do they use it for? How do they buy? do they buy for themselves or someone else is gifting? and how does that differ by season? Basically it does. The is the reason for someone to buy, and you probably have multiple personas shopping at any given time of the year. But does their reason to buy change if it’s, you know, October 1st now? November 1st? December 1st. Yeah. March. April, what’s the context that those personas are in? And do you have that context reflected in the listing? Because that’s ultimately like the bigger sort of equation that Amazon is looking to figure out, which is ultimately personalization or kind of the dream of 1 to 1 marketing, which is like, hey, I know so much about Josh and I know, like.

Christian Umbach 00:22:34  I guess from Amazon’s point of view, I know exactly why he’s putting that in right now. And maybe he, yeah, he shopped like a bunch of black frames before, so I don’t have to recommend him a white frame now, but I’m going to prioritize black frames, within the offering then. And then know that you put all the things into your office. so they will recommend products that have a higher indication of being basically being suitable for that use case.

Josh Hadley 00:23:09  So what’s your immediate like? What are the I want to get like 1 or 2 practical cases, like a brand owner listening to this, what should they immediately do today? Is it like, hey, download all your customer reviews, try to identify your avatar, use ChatGPT or Claude to kind of like re-engineer your bullets or speak to them like, give me like something day like today somebody can go do or should do with like their hero skew.

Christian Umbach 00:23:36  Yep. so I would I would go the ultimate the avatar or persona route. which is, you know, take I don’t know if you want to do it in cloud, for example, yourself.

Christian Umbach 00:23:49  Kind of take the the underlying Cosmo paper. There’s like 12 key dimensions that, are outlined in that paper. And for those 12 key dimensions, you know, throw in the demographics which Amazon gives you for, for your account, which are basically your shoppers. and it tells you a lot about their household level, income level and so forth. throw that in, throw your market basket data in. So, you know, an understanding of, oh, when people buy this product from you, they also buy this in the same card, which again tells you something about like, what’s the the persona, what’s their shopping behavior? And all of those are your downloadable reports that you throw into a cloud session or recording session and, then give it the full product attribute data for you and kind of see what it comes up with. And, it will very likely start indicating like, oh, we’re we’re very strong here. But by the way, in the, you know, also in the brand analytics data, we can see that we’re weak on this use case.

Christian Umbach 00:25:00  To give you a specific example, one of our customers are selling, nursing clocks. So shoes for nurses, which turns out are also loved by, you know, folks, doing gardening or, you know, very practical to work in the kitchen, or for, for cleaning as well. and that can basically like the process that we just outlined can sometimes be kind of a bit of an eye opening experience to say, oh, well, actually, you know, I’ve been marketing this product as a nursing nursing clerk all these years, but turns out the biggest market in a much bigger market that we can compete on in March and April is the gardening space, because that’s everyone’s shopping for gardening. Maybe the nursing space is a bit quiet during that time. why don’t we consider reframing the listing a bit? And I would say a soft touch of reframing is going through bullets descriptions. And then obviously you can really start leaning into it once you see some good indicators there. Maybe in your search query performance data say, oh, by the way, actually we should maybe adapt the item highlights.

Christian Umbach 00:26:14  We should adapt some of the images to kind of go broader on on the category.

Josh Hadley 00:26:20  Interesting. So what you’re saying is you should be changing your content like bullets, description, maybe even images on a seasonal basis.

Christian Umbach 00:26:33  Yes, absolutely. Because if you think.

Josh Hadley 00:26:37  Of trip up the algorithm, it doesn’t think, well, I thought you were for nurses. Now you’re saying you’re for gardening or whatever it is, like you’re seeing that it like maintains the context. So it’s almost like more advantageous to be more frequently, like adding, you know, different personas in there because you, you have limited bullet, you know, bullet point space, right? Otherwise, I’d, I’d create a whole thousand page document about each of my products. Right. And put that on there.

Christian Umbach 00:27:03  Yeah, but it’s not it’s not like you’re, you’re saying, you know, yesterday were like full on nursing and now there’s no nursing at all anymore. It’s full on gardening. It’s more understanding that usually your success as a product is defined as like, okay, here are different kind of persona types, different topics that they’re like looking at, let’s say more through the nursing lens versus more through the gardening lens.

Christian Umbach 00:27:30  And think of this as kind of bubbles on a chart like as you slide over time, like the bubbles, they all move, they all get bigger or smarter. and that’s ultimately and those bubbles all have their specific conversion rates and their specific competitor spaces. but that’s at least I mean, it’s one of the reasons why we basically spend three years just working around like, titles, bullets, now item highlights and description and back end search terms. was kind of really honing in on that notion of, you know, things actually move over time. And the core purchasing intent, now in early October is different than, you know, in the middle of November is different than in the middle of April.

Josh Hadley 00:28:20  Interesting.

Christian Umbach 00:28:21  And so, you know, that whole thing obviously requires a bit more of a systematic approach to realize that at scale. but I think the what I would recommend just as a, as an actionable step, for someone to take one of the products that you really care about, and it probably takes about like five minutes to collect all those documents you have, you know, cloud run for 15 minutes.

Christian Umbach 00:28:47  and see what it comes back with. And, I think that can lead to very interesting strategic questions for for the brand.

Josh Hadley 00:28:57  How influence like what’s the most influential aspect? Is it the bullets? Is it the images title description a plus content? Are you finding like one of those being more significant than the other in terms of like, like dialing it in or telling Alexa who you are?

Christian Umbach 00:29:16  So from a like Alexa point of view with my Alexa head on. I would say like that combination of, like this title item, highlights, bullet, and descriptions are probably the most influential piece, because it’s kind of really laying out that core and in a way that, Alexa has a very easy time understanding it. we ran and this was more sort of at the beginning of of Rufus journey. we haven’t rerun this recently, but we basically did a test of, like, let’s put in a use case like one, you know, cleanly into the product attributes such that it’s, you know, should be really easy to pick up.

Christian Umbach 00:30:03  And then we in a different case, we rather put it into descriptions. And then we started chatting with Rufus on the on the detail page. And it was a lot easier for Rufus to react to it and pick it up, pick up on it in the case where it was in the description as a text versus in the attribute itself where, Yeah, I assume that’s maybe been addressed as a shortcoming by now, but, it was interesting at the time to say, okay, like what’s actually kind of a more influential field. And then I do think that images matter. Like for the, the shopping experience and, you know, certainly for emotional products purchases. But at the end of the day for for all of them, it’s just like from, from an Alexa angle, the written content plays a significant role. Yeah.

Josh Hadley 00:30:56  Awesome. Love that. Christian. Anything else that we need to double down on? because I’m also curious to get your take on, like, so how do I rank in ChatGPT or Claude when customers are searching for products there?

Christian Umbach 00:31:10  Yep.

Christian Umbach 00:31:12  no, let’s double down on that. and I think that space is an I mean we look at like for the things we do ourselves at autopilot, which is basically core listing optimization. We see, you know, a profit impact sometimes of 15%, 20%. So very significant. If you do it over time and you do it right, and then you have initiatives like virtual bundles or, you know, optimizing parent families and, running AB experiments and so forth. And they all land kind of between sort of a 5%, 2 to 5% profit impact ability or 5 to 10%. and what we started tinkering with about a year ago was, okay, what is AI search? AI answers broadly mean for for Amazon brands. And, I think that’s where we actually see like for us, like what we’ve built in organic listing Optimization is a game changer for for brands, kind of just to run that at scale and to be able to realize from a sales lift perspective of a 20% organic sales lift at scale.

Christian Umbach 00:32:28  but on the other hand, we think the impact from AI search is going to be even higher than than that. So it’s kind of the the next frontier that we started working on. And the reason why I feel like it’s so interesting for, for Amazon brands. And let me just pull up, a specific example for, for that. And these are all amazing brands when it comes to Amazon presence and domination in their category on Amazon. miss Mouth hate stains. which is not part of church and white blue summit supplies Lane linen. They actually came up here. I think they had the leading, listing there, for example, around 100% cotton sheets in the example that we heard earlier on Alexa for shopping. But if you look at those brands from an outside of Amazon perspective of like, hey, let’s go to ChatGPT, let’s go to Gemini and ask about their products, they are nearly invisible. and you know, for I guess as preparation for our webinar, that’s the same run on Hailey Designs.

Christian Umbach 00:33:43  Same thing for you. You built an amazing brand, amazing presence within Amazon. And, you know, just looking outside of Amazon, it’s, you know, walling off all of that goodness for someone who’s using ChatGPT Google AI overviews, to kind of get an understanding of like, what are you offering? And they’re, they’re basically in, in able to find your products. and it’s, you know, the problem lies both with visibility and basically with citations going to to your own domain. I don’t know if you would consider any of these guys. Here are your competitors. yeah. On the left side.

Josh Hadley 00:34:27  Yeah.

Christian Umbach 00:34:30  but but for.

Josh Hadley 00:34:30  For the interesting thing, too, Christian is like, Amazon’s also, like, starting to, like, guard a lot of their data, right? Like, you can’t scrape reviews very much right now. Like it’s extremely hard. Amazon is gatekeeping reviews in a big way. Now in addition to that, like even sending in your email like when you purchase something on Amazon right now, it won’t tell you what you bought via email because they know Google is crawling the emails or Yahoo, whatever it is.

Josh Hadley 00:35:04  and they’re they’re understanding what the buyer wants and then they’re going to like Amazon’s trying to say like, no, e-commerce is going to be ours. While they’re there trying not to feed these big AI machines. So you’re right. Like, I think there’s a big component where these AI systems are not looking at Amazon. And also Amazon’s not being helpful to share that either. So like I’m curious there’s there’s a lot that goes into that for sure.

Christian Umbach 00:35:34  And it’s obviously I think there are certain power dynamics of, you know, I would say keeping the customer and being kind of the entry point for a customer, that, you know, Amazon is playing, that the others are playing as well. I think we’re approaching this here from the perspective of our brand, of you as a brand owner, you’re owning a good chunk of your data here. it’s in your interest. Even if, I don’t know much about your business, but let’s assume you’re, Yeah, 80% plus of your business is, ultimately realized on Amazon today.

Christian Umbach 00:36:16  even in that case, we think that the the discovery can happen in a lot of other places. and, you know, there’s obviously a lot of buzz around TikTok. but, you know, we’ve kind of focused on the buzz, more on the AI, shelf side. And the reason, I guess, you know, we’ve been getting more and more excited about that is like, one is just that the growth, in that space overall, like by now, and, you know, in October now those numbers are vastly outdated already, but, August stats were about a billion weekly ChatGPT users. I think it’s fair to assume that, you know, there’s about 250 to 300 million, product searches across the different models, happening week over week right now, and the growth is enormous. you know, Shopify reported on, you know, roughly 50% month over month growth, throughout Q2. and then, you know, AI referred traffic is already kind of jumping up in very significant jumps with that, as well.

Christian Umbach 00:37:32  I think the so yeah, that got us excited. Our own focus on Amazon with this problem was kind of the next thing to point us in the direction of saying like, okay, like, let’s figure out what we can do about this. and we started on this journey where we basically approached us and said, okay, for, you know, right now, what does ChatGPT do? It’s trying to understand your website, for example. But now your website was built for human readers. so, you know, is it going to go deep on your website. What does it actually encounter there? So one of the things that we did was we partnered up with our friends from everything Machines and, they’re basically an infrastructure provider to build, agent sites, which is ultimately a mirror of the core site that’s, you know, published in a way that, agents have a very easy time reading and consuming that data. And we took that as sort of a surface now that agents got attracted to, agents got attracted to and pushed more and more of that shopper intent information and persona specific information and kind of trust signals from Amazon itself, that that you own, into that or made it available for the brand to, to publish it on that site.

Christian Umbach 00:39:02  And what we could see very quickly is that suddenly, even within your own Google Analytics. A new channel starts showing up. AI assistant channel. Like, you know, oftentimes you don’t find it immediately because, you know, it might be on page two of your actual, sources. And, you know, if you do a million, 2 million sessions or so, a month, maybe with your sites, and something starts with, you know, ten sessions, a thousand sessions or 10,000 sessions isn’t necessarily going to show up immediately. But what God is really excited and was just a trajectory behind that data. and, you know, we could see as we started modeling this out, and, you know, in this case, we’re, you know, showing actuals here for a reference brand all the way through through August. before we started this initiative, there were kind of at $400 a month for, for this channel. And then we started the initiative and suddenly we started seeing the jump. And, you know, we could attribute a good a good portion or a portion of that here to Google Analytics and then extrapolate, okay, like where does this go? And just with a pretty conservative growth assumption, can just become an actual channel that scales.

Christian Umbach 00:40:32  And now you could say, hey, 50% is not conservative. But in this case, this growth assumption means that in Q4, basically around 4% of the overall sessions of their site were AI influenced traffic. Now that’s vastly below what you know. Salesforce is telling us that 20% of the sessions are going to be there. or, you know, to annuity coming out with a study that 50% of all of your shopping sales are going to be somewhat touched by AI this year. we try to model this bottom up and understand, okay, like what is ultimately, you know, as we look at generating traffic, as we look at. It all starts with with bot traffic. Then looking at citations, you know, is your domain being cited. It’s your brand being cited. Do you gain visibility? but then if you jump into Google Analytics as a next step, do you start seeing numbers show up? And how do these numbers change month over month? and, you know, that’s just from, from this recent cohort here of brands that we ran over the summer, for example, made us very bullish, major brands, very bullish, on on this space and the way maybe kind of zooming out a little bit on how we approach that is to say, okay, we’re we’re coming out of a place where like our optimizations.

Christian Umbach 00:42:05  Are like our own. Space to operate was basically. We got a bunch of data in from Amazon, DTC, from product sheets, from guardrails that customers have from, like, their legal teams, their marketing teams. That’s kind of our day to day. But we’ve always seen it through that Amazon lens, saying like, go publish it for Amazon and it impacts organic results. It impacts, the Alexa for shopping bots here. but we’ve changed our perspective to say, actually, there’s so much more with that data that can be accomplished. And one channel for us is Google Merchant Center. another channel for us is kind of that surface of an agent side, which basically becomes a place for us to pull together, basically content that’s built both around the product FAQs, product highlights, but also around kind of an understanding of what’s market growth look like. What do personas, what are the topics, what do competitors look like? And back to our example from earlier for that, the nursing clogs. It’s a lot easier to communicate to a bot like, oh, by the way, here are all of the different use cases.

Christian Umbach 00:43:22  and you know, that’s why it makes sense for all of them. Then on a page that’s built for a human shopper, where you kind of have to decide for one, which was exactly your question earlier. It’s like, oh, do you jump between the two? And that’s the nice part of like having content that’s in this case built for agents to consume. We can be we do not have to optimize this for, oh, you know, someone is going to read this one, this little word first and then maybe it’s never going to read the, you know, word number 50 here. but we know that it’s going to scrape the page once It’s fine, it finds the page. And, I think just running this over the summer right now with, with the latest set of brands or that that latest cohort, one of our moments was that, you know, we had always seen OpenAI, meta, anthropic, Google XII show up as scrapers. And suddenly in in August, Amazon started showing up as a scraper on the agent side.

Christian Umbach 00:44:29  and that was kind of an hey, well, sort of full circle moment for, for us to, to see. Obviously tracking is limited like our best ability of tracking today, is, you know, kind of the land on the B2C side. And what happens then on the B2C side? there’s no transparent tracking yet of like, oh, this is actually generating traffic towards, towards Amazon. and what does what does that traffic really look like? But I think that’s a, a space that is evolving as well. Like Amazon just just recently announced kind of a blanking on the name. It’s like a, I guess, partners collaborator program, basically like an influencer program for or I guess AI referred content. and within that you’re basically required to put in certain technical aspects in place, technical products that AWS offers, in order to be part of that program, which ultimately then allows them to exactly attribute, traffic to which which agent that you’re coming from, which is, you know, if we’re there any in a year from now of like Amazon giving us visibility into, oh, by the way, you’re outside traffic came from this agent.

Christian Umbach 00:45:58  I mean, We will love that. We will love to see you tomorrow. and we think that the industry overall is going there with its just, just, just taking some time. So even the, the leading standard, universal commerce protocol just got, I would say proper tracking, bits and pieces wired in, in, in August this year. So it’s really an evolving space.

Josh Hadley 00:46:26  Yeah, definitely. Well, there’s a lot moving in the AI space. you know, we’re running up on time here, Christian, so I’d love to leave the audience with three actionable takeaways from every episode. Here are the three actionable takeaways that I noted. Let me know if we’re missing something. Number one, understand that the way that you’ve been ranking on Amazon is dramatically changing. So if you’re used to, you know, focusing on, hey, what’s my helium ten ranking, what’s my organic position, know that in the next year, I think that changes dramatically. So that means you should begin tracking differently.

Josh Hadley 00:47:04  But like what’s actually being ranked. And that’s primarily through Alexa. So that’s point number one is begin changing your systems to no longer track organic rank, but rather begin tracking where you are getting ranked in Alexa. And I think there’s only more tools to come out to assist with that. Action item number two is to do what Kristian said, which is to take all of the the the scoop data, to take your reviews, to analyze who are the customers that are buying your products, the customer personas, and then dump that into clod and try to address the the common questions that are being asked about the product and the different use cases, and demonstrate through your bullets in the descriptions how your products solve those problems. My third and final point here is to also begin looking at and creating a separate strategy for how you’re going to show up in ChatGPT or Claude or these llms as they begin recommending them. And as you mentioned, like it’s creating websites that are very agent friendly. So begin focusing on that, because if you sleep on that, that’s the new Google.

Josh Hadley 00:48:15  And there’s never a better time than to be an early adopter in AI, as we know that’s going to be the future wave. So, Kristian, as we wrap things up here, anything you would add?

Christian Umbach 00:48:26  Yeah, I think maybe just to do the third one because sort of I guess creating websites obviously sounds like a daunting task. I think there I would say it’s already been a best practice in some ways to have a Shopify store, or some, some B2C domain that kind of helps you, funnel traffic into into Amazon, if that’s your core, outlet core channel of, of selling. so the agent side concept is basically a rather simple mirror, which takes about 10 to 15 minutes to properly activate and start feeding in with content. so it’s not a whole it doesn’t have to be a whole big, website building project basically to, to get that off the ground.

Josh Hadley 00:49:14  Love it. Good, good. Quick win. Very actionable. So, Christian, as we wrap things up, we’ll go through the lightning round.

Josh Hadley 00:49:22  Give me number one. What’s the best the most influential book that you’ve read and why?

Christian Umbach 00:49:27  I think for for me, at least like a top one there is Ray Dalio principles, where he talks a lot about, like, Bridgewater’s culture. Like the company that he’s built, which was kind of a industry leading one. And the investment space and, specifically goes into, I guess, how radical truth and, radical transparency are essentially like a core component of building success, building like a successful team, a successful company. And, just seeing basically that, I guess, as core principles of a company culture. Like I could, when I read it, I thought back to my time at Uber, for example, where, I would say the leadership principles were framed differently, but the ultimately those two parts were part of the day to day. and, Travis Kalanick at the time was really kind of leading with and leaning into to that notion. And it created a very, very strong cultural, that was, I think, the the core force behind, Uber and how it’s, it’s impacting all of our lives.

Christian Umbach 00:50:52  And then kind of just like when it when I went into the ring to, to build companies, then like that became an important part. Just on on the culture that we built with our teams.

Josh Hadley 00:51:04  Love it. Great book, great recommendation. Question number two. What’s been your favorite AI tool and how have you been using it?

Christian Umbach 00:51:13  I think the the favorite one is probably a rather kind of common one, around cloud. Well, basically cloud, big in in camp cloud there. And the I think my favorite way of using it is like really leaning into the, the mic and kind of traditionally I had kind of my stack of applications where a lot of those applications have basically become, you know, basically through MCP directly connected into cloud. I could control the workflows there. I haven’t had success with the tools that I would say promise to run your own. Your entire company, with a set of virtual agents. so you can kind of, I guess, so far have defaulted back to, to cloud and, have kind of grown grown with that.

Christian Umbach 00:52:12  I’m planning to play around with, with grok bot next, which was just released and, promises from friends. I’ve heard promising things around, I guess, how it’s impacting, like their workflows just through the, the easier ability to spin up, certain agents and actually, orchestrate them. but, yeah, I think, you know, if we just look at the last six months, last nine months, like, cloud was certainly for for me personally and for us as a company, the most influential tool.

Josh Hadley 00:52:49  Totally. I think everybody could agree with that. question number three, who is an operator or, maybe a person of interest in the e-commerce space that other people should be following and why?

Christian Umbach 00:53:03  Yeah. I’m a big fan of, Stefan Haney, who also happens to be one of our advisors. but he Stefan was, had like 15 or 16 years with Amazon. He was the leading force behind building out, Amazon Seller Central. then took on kind of an operator hand after that, after his time at Amazon.

Christian Umbach 00:53:29  And, I really enjoy his point of view. And, you know, he’s communicating that on LinkedIn and so forth. podcast as well. And you should have him on the pod. because I it ultimately is a perfect combination of like, hey, here’s kind of the underlying. Fundamentals on like one how Amazon things which are also the fundamentals that have built a massively influential company. and you know, bring some of that into our own lives but also lead it, use it as a kind of as guidance for our customers, where obviously we’re focused on organic optimization, sort of on a very particular set on sort of on the PNL, like with very specific dials and measures. But I think the way, you know, Stefan has taught me a thing is like, let’s lay out the full picture. you know, we’re owning these dials, but, let’s make sure that they have visibility over these other dials as well. And, even though we’re we don’t want to drive that conversation, they’re basically be in a position to say, okay, here’s kind of the questions or the question marks that should be arising on your end at certain parts.

Christian Umbach 00:54:50  And I think, you know, his way of thinking kind of have gone through the Amazon school and then putting that into practice outside of Amazon, as well has is a really interesting point of view.

Josh Hadley 00:55:03  Fantastic. He would he would be great to add on to the show. Well, Christian, thank you so much for your time today. I know we’ve uncovered a lot in AI, has lots of rabbit holes, but I definitely encourage you all to go check out autopilot brand.com, see what the cool work is that Christian’s got got going on. And we’ll also link to his website in the show notes and his contact so that you can reach out to him. So Christian, thanks again for your time today.

Christian Umbach 00:55:31  Thanks, Josh.

MC 00:55:32  Thank you for listening. Visit Ecom breakthrough com for more information. If you’ve enjoyed today’s episode, the best way you can show your appreciation is by clicking the subscribe button and quickly leaving a review. See you again next time!