Decoding Amazon's Playbook: A Former 10-year Executive, Andrea Leigh's, Guide to Winning on the Platform

Decoding Amazon's Playbook: A Former 10-year Executive, Andrea Leigh's, Guide to Winning on the Platform Speaker 2: Hello and welcome to The New Frontier. Very excited to have another really high-caliber guest with us today. So today we have Andrea Leigh, who is an e-commerce educator and advisor with over 20 years of e-commerce experience. She is the founder of Ellume Group, where she educates and advises brands on driving success on Amazon and other marketplaces. She was also a 10-year former senior executive at Amazon, joining back in 2005. So we're definitely going to dig into that and get some good stories from Amazon in its prime, we can say, haha, or before its prime, actually, maybe in 2005. But anyway, Andrea, how are you? Welcome to the pod. Speaker 1: I am great. Thanks for having me, Max. Big fan. So excited to be here today. Speaker 2: Well, you listen, right? So you know how high the bar is and hopefully we can, you know, we can hit the bar of the conversation. I'm sure we will. But yeah, it's great to have you. So I guess maybe to start with, we'll start with a question on the kind of the title of the episode. So decoding Amazon's playbook. What do you think is key to success on Amazon for brands? Speaker 1: Yeah. I think there's, I have, there's a list based on a framework that we work off of at a loom group. But I think before you even get into that, there's some things that have to be true inside your organization to be really successful on Amazon. So the difference, the key differences I see between brands that are really successful and ones that struggle more tend to be around agility and curiosity and, and their interests and capabilities around data. Because it's so different from a lot of other retailers, it's a little bit less of a relationship-driven kind of buyer-supplier experience and a little bit more of a self-service model. And it's always changing. It's always changing because it's ecommerce, but it's also always changing because it's Amazon. And so I think that idea of agility and being able to roll with that It's really important for brands. And then being curious to dive into the data and look at trends and find opportunities is also really important. And then obviously your ability to dig into that data and interest in doing so. So I think there's some like cultural things that are important for brands. And then once you get past that, it's really about brilliant basics. And I can't tell you how many times we work with brands who are really interested in advanced advertising strategies or Amazon Marketing Cloud or optimizing some of their content for AI, but they're only filling their purchase orders at 50% or they have some real structural profitability problems. And so the thing that makes, one of the things that makes Amazon so different from other retailers is that every product, every item has to stand on its own from a profitability perspective. And it's a very algorithmically driven retailers. Things like item suppressions and profitability suppressions, advertising suppressions all happen algorithmically. So understanding how all that works is really important. So the blocking and tackling is around Having a really strong strategy, having a strong digital shelf presence, having financial viability both for yourself and for Amazon, structural profitability in the business, having the products in stock, and then receiving traffic and activity. And so before you start thinking about some of those advanced strategies on advertising and even AI optimization, like getting those basics right, like filling purchase orders, making sure the items are structurally profitable is so critical. Speaker 2: You can really tell that you were a early on and senior Amazonian with that answer. Basically, because you started with the culture, which is really interesting because we had over 50 guests on the podcast and we asked them some variants of that question often. Never really had a response which actually starts with the culture that you need to succeed on Amazon, which I completely agree with. And we'll dig into the kind of the changing nature of the beast and the algorithm, etc. But on the agility and curiosity, do you have an example of a brand who maybe it's a client who you've advised or you just admire who exemplifies those principles? Speaker 1: Yeah, I can share an example of what it doesn't look like and what it does look like. I sat in on a meeting with a global CDG brand and they look at One of the leading indicators for them for their business, one of their key metrics is market share. And for those that are less familiar with some of the retail metrics, that's your share divided by the total size of the category. And first, there's some problems with using that metric pretty exclusively at Amazon because the category size is changing all the time because it's not a finite shelf. It's not a fixed planogram. So the category can be growing or shrinking depending on a lot of dynamics, but they were looking at like their top four competitors or five, and then they had another bucket in their pie chart that said other, and it was growing. So looking at the trend, it was growing. And I said, what's in the other bucket? And they said, oh, it's all the ankle biter brands on Amazon. And I think that's a great example of how a lot of brands really underestimate that competitive landscape on Amazon. Sure, those brands are really small right now, but they're growing really fast. They're super scrappy. Speaker 2: These are like private label sellers. Speaker 1: Yeah, it could be a lot of different types of brands. It could be just newly launched brands. It could be digitally native brands. It could be overseas sellers. It could be all kinds of things in that bucket. And I think really understanding that competitive landscape on Amazon or thinking that it should look like it does. I think thinking it should look like it does in a brick-and-mortar environment can be a big misstep. So, on the flip side of that, I think brands that do really well are constantly monitoring the competitive landscape on Amazon. I've worked with another client who they're tracking some of those competitors as they grow and they're even trying to reverse engineer those competitors' advertising strategies on Amazon so that they can continue to compete with them because a brand that we have a client at our agency, they were a newly launched protein bar and they had a lot of private equity backing. And they said, we want to launch on Amazon so that we can have a good business case to bring to Costco and we want to get distribution at Costco. And that plan cost them many multiples of their sales in advertising, right? Like we had to really put our foot on the gas for them, but we accomplished that. And now that's a real established brand that's like at retail. So those small brands might start as ankle biters, but because they're so agile and they're such a good cultural match for Working on Amazon, they can become pretty big competitors. So I think brands that are really successful are the ones that are tracking that and really paying attention to that competitive landscape, specifically on Amazon. Speaker 2: Yeah. I've never thought about it in your framing of agility and curiosity before, but I remember when I started in Amazon, which is a year after you left. I was a strategic account manager and I was managing 10 accounts. Half of them were high street brands, would have hundreds of shops across the UK, like big high street names, and I wouldn't name those ones. And the other half were Amazon, basically Amazon native brands to anchor Geronic, like Sun Valley Tech. And the difference between the two was that when you think about it from a cultural perspective, was that like, The native ones were super agile and curious and very data-driven, as you say, and they ended up being like 10, maybe even 100x bigger businesses on Amazon. Not in total, but on Amazon, then like the competitor high street brands, you just wouldn't want to engage, wouldn't want to look at the new features, wouldn't want to engage in all the different, even add all of their selection on and all of that kind of stuff. I guess the second part of your first answer then was that Amazon is always changing and maybe it's fair to say, probably is fair to say that it's changing at a greater rate than ever before. A, because there's way more people working there than ever before and B, because those people are being or experimenting with all sorts of new AI to make them even more efficient and rolling out all sorts of new Features and algorithm changes and the rest of it. So how should brands, what should they do differently on Amazon to win in this AI-first kind of world that we're now at 2025? Speaker 1: Yeah, I would agree with you, Max. I think the rate of change has increased significantly on Amazon. I think we had a good five years where things were status quo and teams that worked at brands really felt like they had a formula going. And in a lot of cases, those formulas are still working, but I think the pace of change has accelerated dramatically. On Amazon, even just in the last few months, and in our industry in general in e-commerce, because of AI and a lot of the technology that's available both to the retailers and also to just the companies working with Amazon, even just looking at something really simple like writing product page content is a dramatically simplified process now relative to when I ran an e-commerce agency as recently as four years ago and we had editors on staff that were doing all that work and now there's such a faster way But I think while it might not be super disruptive to a lot of brands' businesses today, it presents an enormous opportunity for the brands that are first to leverage the tools themselves and optimize for the changes at Amazon. And so I think the really smart brands are starting to figure out, how do I optimize for the changes in Amazon's search? And how do I optimize for Rufus? And then more broadly speaking, outside of Amazon, how do I think about that outside of just Amazon? I think there's some really interesting things happening with the shopper right now. When we give talks about how shopper discovery is changing, we always start with, do you know what really sucks? It's making a search on Amazon and scrolling through like hundreds and thousands of search results. That's a terrible shopping experience. And if there was something better before, people would have used it. And now there is. And there's a great quote by Scott Galloway where he says, choice is a tax. And I think that really holds true on Amazon. We already see younger generations. They don't shop that way. They don't start on Amazon searching. They're more likely to start their searches on social media. I think it's something like half of people under 25 or start their product discovery process on social media as opposed to Amazon. And all the older generations are still starting on Amazon. And a lot of those sales are still flowing through the Amazon because social commerce is not very developed yet. And we know that Amazon gets a lot of downstream sales from TikTok. And in addition to that, 80% of the wealth in the United States still sits with people over 50 years old who are still going to Amazon and searching. So we're not seeing it in the numbers yet, but we're going to start seeing it. We'll see you really soon. And I think that some of the new ways of shopping are offering an opportunity for true personalization. So this is a sticking point that I find with brands a lot of the time and retailers talk a lot about personalization and how they're personalizing the experiences. But a lot of time we're using the word personalization. We're actually just talking about relevance. So, we're trying to show more relevant ads to a shopper or we're trying to show if they have looked at, if they're in a category, we're showing complementary products or we're showing category-adjacent products. True personalization is when the engine knows you and that is really, we really have more of those capabilities with AI personalized recommendations who remember your history and know you as an individual. And so we talk a lot about personalization. I think usually we're talking about relevance. Ads don't know us, right? Ads running on Amazon don't know us. They might have some basic demographic information about us. They might have a little bit of clickstream data, but they don't know us the way AI could really know us. And so AI is truly a personalized shopping experience. And then you can take that as far as like agents shopping for us. So I think, and then all of the data is showing consumers are shifting over, right? 38% of consumers said they are using AI to find better prices or deals. And so I think it's really about figuring out how to be successful in this new landscape. And there's a new playbook, right? There's a new playbook for brands to be successful there. Speaker 2: A hundred percent. And we spend a lot of time, as you probably have heard on this podcast, talking about Cosmo, talking about Rufus. For sure. The personalization is the future and I agree. Were you at Accelerate last September? Okay, fine. So you're not going to help me in this argument, but on the last show we recorded, they announced that Accelerate personalized kind of shopping experiences, where if you're searching for different categories, it's going to show you, let's say pillow was one of the examples, it's going to show you, rather than just like a list of products, describe to you the different Issues you may face when buying a pen are what kind of feather count you need, what kind of shape do you need, and these kind of different elements. And they teased this out in Accelerate last year, and there has been a bit quiet since then, but you can completely see how that's going. And ChatGPT has memory on us all and is giving us more and more personalized experience, and Amazon is 20 more years later than that. So yeah, I completely agree with you on that. Stepping back a little bit from Amazon and from your perspective, I know you work, you're not just focused on Amazon. How do you think AI will affect retail media, Walmart and Target as well? Speaker 1: Yeah, I think there's, it's all speculation, right? I don't have a crystal ball, but I think, I think it's interesting. First of all, we're already seeing a shift. So traffic is shifting from, and it was already on the move away from marketplaces for product search and to social in a lot of cases. I think if we see, if there's a fast shift to using Gen AI for shopping purposes, like starting your search on ChatGPT or Perplexity or Claude, the volume of searches and impressions is going to go down on those retailer platforms. So that's number one. So there'll be less search volume happening there. And I think it has the ability to lower both the competitiveness of search advertising on those platforms as well as the advertiser's willingness to pay. Because if there's not as many people there, it's not as many eyeballs and impressions, their willingness to pay is going to go down. Now, it's interesting because Amazon is, they're blocking a lot of the agents and stuff from going onto the site. If you use Operator and try to get it to go shop for you on Amazon, it can't. And so I really feel like the retail media networks on the retailer sites, they're going to have to reinvent themselves. They're going to have to be more like, they're going to have to rebrand themselves and reinvent themselves as more AI-powered marketing platforms and offer advertisers opportunities to influence Gen AI and agents. So I think that's what's missing right now is on Amazon, we know that if you're running ads, that ad content is often incorporated into Rufus. There's some ads in Rufus. There's not a lot of visibility to the brand in terms of metricing and things like that yet. It's in there and the retailers are going to have to figure out a way to sell optimization for those other platforms if the shopper is still starting theirs. I could see a world where, and this is just speculation, what they're monetizing on the site might start changing. For example, on Amazon right now, they sell ads, but it's free to fill out all the product content. And it's free to engage in the reviews programs. You can pay for Vine, but organic reviews. I could see a world where they might start having different tiers of that and start charging more for providing more content on the site because you would know then that would be more content for the agents to crawl, for example. So I don't know if that's exactly where it's going to go, but I think monetization might start to look a little different. It could look different. Speaker 2: I totally agree. I think the year, the blue links, this is like very 2010 to 2020 when we look back at it and it's like a business model that was great for, you know, the traditional SEO type, put in a keyword and here's 10 blue links that match that keyword or 10 products that match that keyword. It's all the same, but it doesn't work in this new world. And I agree that different models would happen. I mean, I want to get more into your Amazon journey later, but as a former Amazon executive, how do you think that conversation goes? Because the Amazon ads team obviously is a cash cow. And they are, and I imagine same at Walmart Connect and Target. They're big cash generating businesses, but at the same time, the CEOs have to look after or champion the best customer experiences, which may be killing the cash cow and experimenting with new different models. As you say, like maybe you charge for additional elements in the product upload and that, yeah, or whatever it would be like, how would that go? How does that go down? Because these people aren't going to want to like, You know, give up their empire so easily, are they? Speaker 1: No, and I think it's why they're blocking the agent traffic right now, because they don't want to bypass the advertising platform. But you could also envision a world where they launch new ad types that are more content-driven, and then those ad types are more easily read and ingested by AI. So, I don't think they're going to let go of advertising. Certainly not. They're already putting it into Rufus. But I think Amazon's always doing that delicate balance of what's best for the shopper and then what's best for their business model. And I think it's best, and I think they realize the shopper is moving. You know, kind of what we were saying before, running a search on Amazon and scrolling through the results is terrible. Speaker 2: It brings a fundamental risk. So I don't know where you got your 38% stat from, but I got from A16Z, it's a big VC fund that did a survey and found that 60% of people were using AI bots to research products and services. So huge amount of people and anyone listening to this probably has done that on ChatGPT at some stage. So is there not a huge risk, which is like the same risk that Google's facing now, this like kind of second order a bit, which is for Amazon, which is? You know, traditionally 80% of people started or recently 80% of people started their searches on Amazon for products. But if it's a terrible experience, why wouldn't you just go, you've got all these ads, you're just going to go to ChatGPT or Complexity Shop or there'll be the model that doesn't give you the ads and just explains products and reviews the content and external factors, magazines, top 10 suggestions for coffee machines and brings that in. That's going to be the most helpful one. So there's a risk here. Speaker 1: Yeah, there's a huge risk that the ad spend is going to get bifurcated to these intermediaries, right? So if you look at ChatGPT, Perplexity, Claude as intermediaries to the shopping experience, it's like when Instacart came onto the scene. They're not a retailer and they're not a brand. They're an intermediary. And so they were able to scoop up a lot of advertising dollars just by being a go between the shopper and the retailer. And a lot of times brands are paying, in most cases, brands are paying Instacart for advertising. And then they're also paying the retailer for advertising that Instacart's getting the product from or delivering the product from. So it's a, it's actually like a four-sided marketplace. So you could also envision, you could envision that Amazon's not going to, they're going to get creative. They're going to make Rufus really good so that you don't need to use another tool. They're going to launch new ad types that are more content-driven, that have the ability to influence some of the intermediaries, and they're going to try to figure out how to show brands how that's happening and prove it. I think that's the other thing. The metrics are going to have to change. How ads are influencing Gen-AI recommendations is going to need to be demonstrated. And how some of the on-site Gen AI tools like Amazon's Rufus, how is that showcasing advertising products or advertised products or how is it showing ads? There's going to have to be a lot more, they're going to have a lot more metrics around that. But I do think we already see, which one is it? Is it Perplexity that's doing ads right now for retail media? Unknown Speaker: They are and they aren't. Speaker 2: So Perplexity did a deal with Shopify. So they use a sort of shop API thing that you can go on to like shop which is its own app and basically shows you it's like a mini Amazon but for Shopify stores. It's not terribly good but the point is they have an API whereby everyone's selling something on Shopify is on this API and they don't deal with perplexity. So they, I think perplexity are even more because they have to be given their like the the upstart here. They are even more aggressive in their experimentations. They're trying ads but also I think they are trying to build this new like perplexity shop which is basically you take all the knowledge of perplexity and all the looking at different reviews plus the Shopify API and you have this kind of new So it's one of many different experiments they're running. They're running stuff on the enterprise knowledge. They're running a ton of different stuff and who knows how well backed each of these are internally. But I think they are very innovative and I'm a huge fan. So I think it's going to be a headache surely for Amazon soon as well as just Google. Speaker 1: And but I think that the pace of rate of pace of change is going to be a little slower than we think. And here's why. Gen AI is not good at shopping yet. It's just not. If you like even perplexity, which is the one that advertises itself as being the strongest, it's only scraping. It's not scraping all retailer sites. It's somehow it's choosing. It's showing you which ones it's pulling from. And it's only choosing from those. So you're getting like a limited, like if you're searching for a black woven leather purse, it's only going to pull from eight websites versus Google, which is pulling from the everything. So I think there's, there are some challenges there and I know a lot of people are using it and engaging with it, but I would like to hear more about how effective they feel it's being for them because I've been using it and I think it's a little bit disappointing. Speaker 2: Have you tried deep research? Are you a pro-subscriber? Speaker 1: I am. Yeah. Speaker 2: So with the deep research, it will literally go to thousands of different websites and whatever. And I haven't used it for shopping, but I've used it for many things. And that's pretty incredible. It is literally like having a mini McKinsey, whatever, like analyst to just do a lesson, take them two days and does it in five minutes. I mean, that's pretty incredible. So instantly you say that. Speaker 1: Yeah, it's still a little bit less of a visual experience, which is my experience with it. So it feels, it depends what you're shopping for, but it feels less fun because you're not having that visual experience that you're having, like browsing more of an image-based search. And then the agents are even worse. That's going to take a little while. Like I had it searched for, I ran out of one of my supplements and I was like, okay, go, all right, go find me the best price on, I used operator, go find me the best price on this. And it, it doesn't understand like price per each or price per ounce yet. So it went and found actually what ends up being the most expensive option. It's just the smallest pack size of the item. And then Amazon's blocking the traffic and then you have to take over and log in. They'll figure out a way around all of that. For sure. But it's not it's definitely not faster or more or more inspiring than the alternatives right now. It will be eventually. But I think in this moment, we're not quite there. Speaker 2: Yeah, fair enough. I want to ask you one more question on AI and then I really, I'm very excited to ask you questions about Amazon because I think it's gonna be good fun. So if you were advising an executive team with a major brand of retailer today, what's the one thing that they need to do now to stay ahead of AI-driven retail? Speaker 1: I think it goes back to the things that make you successful in Amazon. It's like agility and curiosity. I think there's a new playbook and a few things are really different about optimizing for some of the, for AI search, both on Amazon and Cosmos search, but also optimizing for Gen AI. In general, obviously images, you guys talk about this with your company, but having images that show the product in use, who uses the product, we call it the five W's. What are the, who, what, when, where, why. Reviews are fair game. Amazon seems to be using those. One thing that I, and then obviously making sure that the content answers the five W's, but I think one thing that is a little bit overlooked is You're direct-to-consumer site. For a lot of brands, that thing just sits there. It's not really like a moneymaker. They don't do a lot of volume through it. They're doing a lot more volume through Amazon. It depends on the brand, but that's a really overlooked asset right now because that's fair game for being crawled by all of these different Gen AI tools, and that's often where brands have the most robust product information. All the specs, all the technical characteristics of the product, all of the ingredients, the whole brand story, like all that stuff that's hard to get up on Amazon because it has been so much work until now, that is all there. And it's unclear to me, you might know the answer to this, Max, if Amazon is checking the DDC sites in Rufus. Or not, but it seems at some point that would make sense for them if they're not doing it already. I think the DDC site has more, a lot of opportunity right now and it's totally within the control of the brand. Speaker 2: To answer that mini-question before we move on, we've read all of the science papers and in the Cosmo paper specifically, they say they are using external sources. They're definitely thinking about it and they did say that they are using it. Who knows what level of authority they put on it versus the data that they already have on all these brands. But yeah, going back to talking about Perplexity Deep Research, so not only did Perplexity Deep Research come up with the type of this podcast, I was then having a bit of a chat with it because I wanted to get into your journey at Amazon. When you joined Amazon in 2005, according to deep research, the stock price was $1.58. $1.58. Pre-split. Speaker 1: Yeah, pre-split. That's the adjusted price, yeah, for the pre-split. Speaker 2: And 9,000 employees. And it generated in 2005, 30 million in profit. In 2024, 60 billion in profit. This is pre-FBA, which launched in 2006. It's pre-AWS, which also launched in 2006. So, like, just your thoughts on what the company was like at that stage. It's a totally different beast to what we're dealing with now. Speaker 1: Yeah, so I was employee 4,012. I was definitely early in the journey there. We, just to give like some context, we used to hire people out of MBA program. And when I left, we usually had about 500 incoming MBAs every year. And in my class year, there were 13 of us. So they hired 13 people that year out of MBA programs. We all, like in the, on the North American retail team, we all shared a printer. We sat on one floor. It's like a really small team. And I remember letting some of my business school colleagues know that I was going to go work for Amazon. And they were just like, the answers, the responses to that were hilarious. Like I would never give, put my credit card information online. This is how long ago this was. They're not going to make it. A lot of people thought they weren't going to make it. They were having a tough time turning a profit. I'm just asking ChatGPT right now, what was Amazon's revenue in 2005? Because I'm like... Speaker 2: Well, I had all of this lace, but I've lost it. I had all the... Speaker 1: It was like really... Yeah. Okay. Amazon's revenue, it was $8 billion for the year. Now, that's a quarter on quarter improvement for some of their segments of their business. My first job was... I was a product manager for pricing and so we were, the way that Amazon was handling pricing is they were getting these like feeds from companies that would scrape pricing in store and online and then we would upload them. And that's how we price matched. And so it was super error prone. We actually had a number of like pretty egregious errors. We sold like a big screen TV for $10. And so we're like, this is too error prone. We need a better way. And so my job was to scope out an automated price matching system and figure out what that would look like and launch it. And then Fire, the company that was giving us these feeds. So that was, I worked on the price for Amazon's competitive monitoring tool. I have the original business requirements for that. So that was pretty exciting. Worked on Harry Potter book launches. I remember when I worked on the launch of the grocery category in the US, we had, we didn't know what we could sell. We didn't know what the fulfillment center could handle. And I had a client who had Bertolli's olive oil in a gallon plastic jug. And we were like, can we put that on a conveyor belt? What will happen? Will it tip over? We don't know. And our operations team, it was on a wiki, an internal wiki. It was like, you have to do a drop test. I'm thinking it's this big sophisticated thing, like I have to have the vendor submit it to this center or something. So I called the supply chain department, which you could do back then, and they sat in the next building and they were like, oh, just meet me in the garage and bring the olive oil. And so I went down into the basement and my supply chain friend, stand back. And then he held it up above his head and he just dropped it on the ground in the parking garage and it didn't break. And he's, cool. Unknown Speaker: That was it. That was the drop test. Speaker 1: And I remember being like, Wow, this place is really put together with duct tape and chewing gum. But yeah, it was just a really different place and going through those specific 10 years where they just experienced such extraordinary growth. We got a campus and started hiring from like really important schools and the quality of the talent and caliber of the talent increased and the stock price increased. Everyone got new cars. Some people retired. I remember there was one year where the stock like really went up and like almost the whole parking garage turned over. If you're in your 20s and you've got a bunch of money, what are you going to do? You're going to buy a car. Speaker 2: That's amazing. My first question when obviously is, did you have any interactions with Jeff? What was that like? Did you see him walking around or any big meetings with him? Speaker 1: Yeah, we saw him walking around. He didn't have like security detail back then. He used to eat up at the PACMED building cafeteria every day. So if you wanted to see Jeff, you could just go up there and you could walk, go up and sit at his table and pitch and people would pitch him and stuff like that. I saw him up there a couple of times. Our all hands were really small back then. Speaker 2: Yeah. Speaker 1: That was always cool. You'd be in the Fifth Avenue Theater, one of our local environments here in Seattle, and get to hear him talk. I saw him walking around a couple of times. I didn't ever have a meeting with him. There was one time I was supposed to. I was working on the launch of Harry Potter 7, which was our highest-profile commercial launch back then. And I was supposed to be, I was supposed to be presenting him like our whole project plan for that launch. But then this guy, Diego, who used to run retail, he subbed in at the last minute. So never had a meeting where I had to present to Jeff. Speaker 2: That's a shame. It's so interesting how, in my opinion anyway, the When you have an enigmatic founder, the culture of the company is just different. I remember when I joined in 2016 and we would still do all hands. In London, people would start around now because they're based in Seattle. People would be sitting in a room watching the screen with Jeff talking at 6, 6.30 at a time in London, which is after work, basically. You just couldn't imagine anyone doing it. For the Andy Jassy ones, people would even bother watching a recording. It's just a totally different culture and to be in that in Seattle in the early days. Speaker 1: When you never knew what he was going to say, that was the thing. I think in later years, he really got a lot of coaching, but he, there was always something that came out of those all hands where like someone would ask a question and he would really, he would really speak his mind. Like I remember one time someone raised their hand and they said, are you worried about the stock price? Cause it had gone down a lot that year. And he was like, no, if I, if, what did he say? Or he's, no, they said, are you happy about the stock price? It had gone up. And he was like, No, we're not any smarter when it goes up because we'd have to be dumber when it goes down. It just speaks to the inputs culture, right? The inputs focus culture. Speaker 2: One of my mentors who was a TA, the technical assistant to Jeff for a period, told me that he would love to encourage arguments in meetings. Like that was the thing he would do is like trying to get the two different sides of the team. So in the example we talked about, head of our Amazon ads versus head of Amazon Rufus or whatever, and get those two people to really like fight it out. And this is something he says quite publicly. He would say, if I choose between arguments and conflict, I would take conflict every time it yields better results. It's publicly stated that he did this, as well as inside privately. So what was some of the biggest like debates and strategic kind of conflicts that you saw from your time in the early days? Speaker 1: Yeah, I think he was also pretty famous for, and I think it's the same line of thinking, having different teams work on the same project in complete isolation and seeing who comes up with the best idea. And I had that experience in two different roles there. One where, so we were trying to figure out how to do This is early, before Amazon had that as part of a prime benefit. And I worked on a team that was, we were doing it through Amazon Fresh. So we were using Amazon Fresh as the mechanism to do that. And it was called Amazon Now. And that was my program. So I launched that. And then someone else was working on Pantry, which, or not Pantry, they were working on a tote program, which was like another basically same day delivery program. And then someone else was working on Prime Now. And it wasn't called Prime Now, it was something else. So we had three teams that were all working on solving the same problem. How do you get non-grocery items delivered same day? And what's the best way to do that? And we all were taking like completely different approaches to it. And we all knew, everyone knew that there were other teams working on it. It's like this race to see who can get there first. What they ended up doing was it ended up being like a blend of all of the programs. It had features of each one and that's what they went with for the same day delivery program. But yeah, I think the program did end up being successful because they were able to take the best bits from everything that we were all working on as opposed to putting us all on one team and making us subject to groupthink and compromising and things like that. Speaker 2: I definitely, that is definitely a feature of the culture and I saw that a few times as well. Just as you're saying this, I triggered in my memory of definitely teams working on the same thing and totally in isolation. So what was some of the biggest challenges you saw as just like this crazy scaling, like historic levels of scaling for a tech company? Speaker 1: I think one of the challenges, so I worked on the retail side of the business my whole career. I was a category leader. I worked on the grocery launch for U.S., Canada, and Mexico, and I worked across some other categories too. I think one of the challenges that was front and center for me is that we were leaving the vendor community behind. So we were scaling so fast, the suppliers didn't know how to keep up with all the changes that we were making constantly. And so they were just in this reactive mode for a few years. I think a lot of them are still working on that. But because Amazon's just so different and operated so differently from other retailers, I think that was a big challenge back then. Oh, we want to sell all your products. Great. Give them all to us. And then it was like, but actually they all need to be profitable and they need to be profitable on an item level, not in a portfolio level. And we don't do a loss leader philosophy the way other retailers do. And the way we approach negotiations is different. And operationally, shipping into Amazon was really different. Small quantities, frequent orders, as opposed to big truckloads of stuff you'd ship to Walmart. So I think the vendor community, that was a big struggle. And then I think the culture there was, I loved my time there. I had a great experience. I feel like I learned so much, but it's certainly not for everyone. You also leave people behind. When you're scaling that fast, if what's required to be successful in your job today is going to be different than what's required to be successful in your job in 6 or 12 months, not everyone can scale with that. I don't think Amazon did a great job of helping people find something else or there's no one looking over your career there. You're on your own. There was an HR department back then that they were a little less They were definitely more reactive than they were proactive. It seems like a lot of that's changed. And also the pace of change there is different than it used to be, at least in retail, but so much was changing so quickly. And I think when that happens, you just, you end up leaving some people behind, both like partners, vendors, and also some of your staff. Speaker 2: Jo, that makes sense. And you're obviously now a founder of various businesses since. Which leadership principles have you taken forward that you're now putting to your own companies? Speaker 1: Yeah, that's a great question, Max. When I joined Ideoclick and we ran that company, it was really small when we started, like about 15 people. And we actually, as a leadership team, sat down and said, we were all ex-Amazon, and we said, okay, what leadership principles we want to be deliberate about? Which ones do we want to keep and which ones do we want to let go of? And I think the ones that we really liked were, and we really like to dive deep as a client service organization. We were Amazon Managed Services Agency. You need to be able to really look and be curious into a client's business and figure out what's going to drive good results for them. You need to earn trust because we are in a client service business. I don't even know if all the leadership principles are still the same as they were when I was there. Those felt important. The ambiguity one always felt a little weird to me. What we decided to let go of was there was this concept at Amazon. I don't know if this was still around when you were there, the business athlete. Everyone had to be good at everything. And if you were not good, like maybe you were better with the vendors but less analytically driven or you're more analytical and you were not as good with the vendors, there would always be this pressure to improve the other area as opposed to just like leaning into what you're really good at. And that was something we decided to let go of a little bit at the agency because You have the ability to assemble a team based on people and leverage people's strengths and really lean into their strengths and hire around that. And so that was something we let go, that pressure of like everyone has to be a business athlete and good at everything. Speaker 2: I think that definitely started to change in my six years because in the early in my journey, it was very easy to move between departments and they had that kind of like business athletes, as you say, and you should be good at everything. So you could, like I did for a bit, like you're going to go from Marketplace operations and doesn't matter that you know nothing about operations like you can go and do it because you're super smart and that that did change towards the end and you were you were hiring people like it was a specific. Kind of roles and like they weren't you they couldn't really easily move teams if they weren't fit to the team they'd be they would they wouldn't easily be able to shift around. So my last question I think we can tie it up because it's been a great conversation. In 2005 it obviously wasn't clear that Amazon would come to dominate and just wipe eBay off the map and now it's bigger than Walmart and become like outpaced Walmart. But specifically with the kind of other online markets, like the specific ones like Pets.com, specific niches and the other generalist ones, what do you think, with the 10 years of reflection maybe, was the factor that led Amazon to win and everyone else to fade into irrelevance? Speaker 1: I feel like a broken record, but I think it comes down to the culture. I really do. It's not any one business decision that they made. It is a culture of embracing failure and choosing to see failure as an opportunity to learn something. As opposed to something that's maybe more negative and less useful. And that's the big difference that I still see between across the clients that I work with and companies that I work with. That sets apart a really innovative and fast-growing company from one that is a little more stagnating. And it tends to come down to how they treat failure inside the company. And maybe if you're listening to this, thinking about the last time someone on your team failed, like how did your company treat it? How did you treat it? What do people say about that person? At Amazon, it would be pretty common for a project to completely blow up and be really unsuccessful and that person to go on to take on a role of bigger responsibility and maybe get promoted later. Because the learnings that came out of a failure, just that someone was willing to put a business case together, be thoughtful about it, launch something, try to learn something and move on. I worked on tons of failures while I was there and failing is really how you learn and how you get better and how you figure out what's resonating with your shopper and I think that companies that are more fearful of failure and are less willing to try and experiment. And then, but Jeff used to talk about one-way doors versus two-way doors. So a one-way door is an experiment or a business idea or something that you do that you can't walk away from. Like you, you can't undo it later. That might be like a, an acquisition or like a, something that's more difficult to pull yourself back out of. And so all of the opportunities, we tried to look for two-way doors, but something we could try and if it doesn't work, we can walk it back. So I think that distinction can be pretty useful for companies as they're thinking about where they might want to experiment. Speaker 2: Just because you teed it up in that, I have to ask the final question and we can move on to wrapping up if you don't want to answer it, but what was one of the craziest project failures that you worked on in your 10 years? Is there anything you're able to talk about? Speaker 1: I have a couple, but they're not very interesting. The first one was when we very early in the private label journey, I worked on the baby category and we were like, we're going to do like private label baby bedding. And we didn't know anything about sourcing, manufacturing or anything. We ended up with just a lot, like everything ended up on Cadence Markdown. Unknown Speaker: We didn't sell any of it. Speaker 1: We totally underestimated how important it is to build a brand and then the name that the agency we hired to work on it came up with was absolutely hilarious and I cannot believe we went with it. It was called PINZON, P-I-N-Z-O-N, which sounds like pins on, P-I-N-S-O-N. It was PINZON baby. It was betting. It was like the most ridiculous name, like you're sticking pins in your baby's body or something. Then the prints we chose were all wrong and we didn't understand about building a brand or sourcing. We realized after being working at a retailer for so long, we actually knew very little about retail. That was a big bust. That was a pretty big one. Then there was another one where we wanted to sell ice cream. Oh, yeah. Financially, that does not work. Speaker 2: As in through the Amazon? Speaker 1: Through the Amazon.com platform, we tried to sell ice cream and it's really, surprise, surprise, it's really hard to keep it frozen. Speaker 2: What is actually hilarious about both of those is that it comes back to the multiple team things, like one of my friends in Amazon ran a private label. And then he, they closed down the program. So they tried it again. Didn't work again. This is like in 2023, 4. And the whole team had to find other roles in the company. They were given six months to find other roles in the company or had to leave. And then again, with the ice cream, like we worked for frozen delivery as far as I worked, but in Gracer. And we, yeah, we were looking at different like freezer models for delivery. So it's funny that the same thing's coming around again and again, they're not scared of. Speaker 1: There's not a lot of ice cream, apparently. While I was failing with pins on baby bedding, my friend Adam Getsch was working on Lawn and Garden and they made private label patio furniture. If you're going to just experiment with private label, it doesn't seem like where you'd start, but whatever, that's where they started. And they made like colored cushions and stuff. And he had this, after it failed, he had this whole thing that he used to say all the time. And he said, fuchsia is not a primary color. Stick with the basics. Don't get too cute with your private label. But that was back before they had the team working on it. And you just had the vendor management teams trying to figure it out. Speaker 2: Andrea, thank you so much for coming on. It has been a great conversation. Hopefully, you've lived up to the bar that you've heard, but no, I'm joking. It's been fantastic. Do you want people to reach out to you? And if you want them to reach out, how would you like them to? Speaker 1: Yeah, you can find me on LinkedIn. Message me there or you can get me off of our website as well aloomgroup.com. A-L-O-O-M-E. Speaker 2: We will link those below. So yeah, thank you so much for coming and I think you're in short talk. So if people are listening, are you speaking at short talk? Speaker 1: I'm not speaking but I'll be there if anyone wants to know. Speaker 2: There we go. They can find you there as well. Yeah.