Latest / The Modern Supply Chain / Why profitable businesses still end up cash poor with Taylor Holiday

Transcript

Why profitable businesses still end up cash poor with Taylor Holiday · The Modern Supply Chain

• Izzy Rosenzweig (00:00): Anytime I have a customer that starts wanting to spend a bunch of time on assessing my models, I go, "Oh, you don't think about this right. This isn't about how good can we get at guessing. This is about how fast can we react to being wrong and fix it. If I was in charge of this, what I would want to do is how do I build a more flexible supply chain?" And I think that brands are realizing this, it all circles back. It's like full circle to drop shipping as this is actually the right e-commerce business model. Taylor Holiday (00:25): This is The Modern Supply Chain, the show where we break down Modern Supply Chain strategies to help e-commerce brands shift from staying above water to predictably scaling. Everyone, welcome to The Modern Supply Chain. At The Modern Supply Chain, we're passionate about doing everything in a way to help your business be successful in a more profitable way. Today's guest is Taylor Holiday, CEO of Common Thread Collective, one of the most influential growth and financial strategy partners for the modern DDC brands. Taylor works with brands doing five to 200 million in revenue, helping founders move past service level growth tactics and build businesses that actually compound and scale. His work sits at the intersection of marketing, finance, operations, and now increasingly AI. In this conversation, we'll dive into how AI is reshaping e-commerce, particularly in marketing and finance. We'll explore how modern DDC brands can leverage AI, not only to scale faster, but also optimize financial strategies, streamline marketing decisions, and create more efficient supply chains that are fully integrated into the broader business ecosystem. (01:26): Let's dive in. Taylor, the death of growth at all cost. I think you're passionate about it, I'm passionate about it. Maybe for the listeners that are getting into e-commerce more recently, talk us about the old days of e-com, what that looked like and what it meant growth at all cost and how bad that was. Izzy Rosenzweig (01:42): Well, it's funny because I think it's really important that we contextualize that in any market, businesses are asked to play by the terms that are sort of pulled in the direction that the market dictates. And when there is an abundance of cheap capital, cheap on both the equity side and on debt side, what it means is that you're willing to push your return further out on the yield curve. The expectation of return of those dollars, because the cost of capital is so low, goes further and further out, and brands tend to prioritize top line revenue growth, not bottom line, because they have existing treasury dollars to grow with. And so when I look back on that time, it's easy to criticize the Allbirds, the businesses of that era, that we look at and go, they never were profitable. They grew a bunch of top line, but didn't produce a single dollar of profit. (02:37): But I actually think that, and if we look at the actual value realization of a lot of those founders, they actually won the game on the field at that time. It just so happens to turn out to not be a great way to build a business long term. But in that moment in time, the market wanted that and had a belief about these ideas about network effects and economies of scale that they thought would play out that just aren't real in this business model. And so I think in that moment, there was some actually logical behavior relative to the dynamics, but bad underlying assumptions that informed those market realities. Taylor Holiday (03:10): I actually lived in that space. I ran a venture-backed brand for a while and the venture community was A, to your point, free money throwing money. Yes. But the question is, why were they doing it? And the reason why they were doing because they were looking at public markets and public markets were rewarding high growth at all cost public traded e-commerce businesses. And they weren't based on fundamentals. And I would argue you might see a little bit in AI as well. I think there's a couple different versions of AI and we'll get into it. There's like, unlike crypto, which was a trend, growth at all costs had no revenue or no profits. AI is a little business where there's revenue and there's EBITDA, but definitely back in the e-comm business where there is structural costs you cannot avoid, to growth at all costs, venture was looking at further down the line, public markets. (03:54): So, hey, us from the boardroom, we're going to push you guys to go top line. And the second that changed course, now there's no more treasury. You're done. Izzy Rosenzweig (04:03): I actually think that this is a really important point and you're exactly right that the investment community thinks about the end liquidation mechanism and what they're valuing and builds a business accordingly. And that's why I actually think people treat that moment as if it was very illogical. But I actually think there was the behavior of the founders was relative to the incentives of the capital that were put in the organization and how they thought about value realization. And the other thing that happens is if you think about some of the software world, there's a lot of these markets that end up as winner take all dynamics where you create a lot of future pricing leverage and therefore gross margin expansion if you go out and win the market. So Uber is like sort of the classic example where it didn't matter that they didn't make money for a really long time. (04:42): And they're in a position now where they can sort of exert leverage on that capture and create massive cash flows and wins into the future. Shopify is starting to do this now even, I think, where you think about their most recent earnings call and the dynamics that are in play, they now have such market power that they're able to at their discretion when they need to, they can become as profitable as they want to be in many ways. But that is just not the dynamics of a consumer market, especially when there are virtually no barriers to entry in almost any of these categories. So much of consumer e-comm is actually deflationary to price and therefore to margins that those sort of like long-term lock-in values don't play out. And so if you end up with a business where you're trying to grow at negative gross margin or a gross margin that's not sustainable for the hope of future expansion, it just never happens and it never comes to bear. (05:30): And I think that's the dynamics that sort of were wrong about some of those assumptions is that just consumer product and also the limitations, the massive just influx of competition into these categories, I think was underestimated in terms of thinking about that future value basis. And a lot of the LTV curves were just wrong as well. And the assumptions of CAC that were built into the future, there was a lot of models that had revenue growing and CAC being constant. That stuff was just bad modeling about some of those assumptions. And so there's a lot to learn from that era. But I actually, as I've thought about it of being an entrepreneur whose job it is to return value to shareholders, in some ways I have an appreciation for those founders that they did that in that moment. Many of them won that version of the game in ways that's hard to discredit. (06:13): And so I think that the market changes a lot and therefore the game that we as founders are playing gets altered as well. And it's easy to be like, " Well, just build a profitable business all the time is the right answer, "but that's really not what the market was demanding from people. Taylor Holiday (06:26): And you know what? And then back to like everyone was looking somewhere, right? So like venture looking at public and trade markets, but what were they looking at? And I think to what you said about Uber was true but Amazon. Someone was going to be the Amazon. Who was going to be the Amazon? Now, they won it. They did it and they burnt money for years. So everyone after that, like, oh, we will be the next Amazon, but there aren't that many Amazon players here. And to your point now, it takes nothing to start a business. Izzy Rosenzweig (06:50): That's right. Taylor Holiday (06:50): You got to run it very differently. Izzy Rosenzweig (06:52): And even if you think about Amazon, if they don't unlock Amazon Web Services and even the advertising business now, I think we think very differently about the marketplace version of that business model as like having played out versus everything that they use that to build around that core consumer interaction that actually drove so much massive profitability. And the underlying thesis that I think ... I even think about this a lot as a service provider is that Amazon is sort of my inspiration that they did this thing which was at the center of it was the customer experience where we were going to try and drive price down and value up forever. And I think a lot of times in my business, in particular, service business, people get this thought that the goal is for me to increase my price over time against the set of things that I'm doing. (07:38): And I actually think this is fundamentally wrong, is that I don't want to increase my portion of the value capture or value creation that happens with my business. I want to drive the cost of my core services down all the time and while also trying to expand my gross margin on that through technology and AI and the things we'll talk about so that the value calculation to the customer increases in return to them. I think Amazon understood that and was able to use AWS and these other things to never capitulate on we're going to be the cheapest, fastest, best way to get you the product. And so the experience is going to be better than everything else. And those other business models enabled that to be so much more powerful than anything else. I Taylor Holiday (08:15): Mean, applying that thesis I think is the reason why you're so successful. I mean, most people do not think like that, and I love that. And but back to e-comm, I think founders that are listening, this is actually really good news. So you are not going to be the next Amazon. Actually, when you start your brand, don't even try to compete with Amazon. You're a storyteller, you're building a brand, but you must build it from the ground up healthy, profitable. This is not version one anymore. Izzy Rosenzweig (08:40): Yeah, exactly right. And it's funny, even the way that you and I ended up connecting for the first time, I think we were reflecting before the podcast started was like two years ago I reached out because I've always had this view that our business through this forced evolutionary function of the drain of all the available capital was going to go through what I called margin innovation, which is that people were going to constantly pursue ways to generate marginal capture off the value revenue that they were creating. And the biggest line items in any e-commerce business are always going to be the cost of goods of the product, the fulfillment of that expense and the marketing. And labor more and more should be lower and lower as a portion of that. But those three things, I am just convinced that we are headed towards ongoing innovation all the time about the way that that happens. (09:23): And I think what Portless was, was like, that was a time where I went, okay, that is an example where somebody is innovating around the core cost center that allows for leverage in all the other ways. Because my world is a bunch of customers yelling at me to try and improve the efficiency of their marketing. And all the time I just go like, "Ah, your underlying need for more efficiency in marketing is actually degrading your ability to grow." And if you would scrutinize these other things with the same level that you are your Facebook Roast, you might unlock the reality that if you could spend more on this channel and still get to the same marginal dollar value, that's how the business might grow more. Taylor Holiday (09:58): Hundreds of brands agree with you. So the hundreds of brands that Portless services. And to your point, I didn't do it because I wanted to do it. I was forced to do this. That's right. I was forced to do this because the entire model was changing. And to your point, Cog is your highest cost and your capital's tied there. So how fast are you going to cash out of capital? How much inventory do you need at any given time? And the world is getting larger. We spoke for a second about one of the OG brands you worked at. It's a big world out there in how you're selling to them. But the next thing I think actually ties really well together, which is your interconnection between finance and marketing. And I think one of the things you said is most brands, some of the best brands, I'll say most of the best brands in the world today are storytellers, meaning they're branding people. (10:39): They create really, really cool products. But as the business grows, if you have to grow and optimize your business, you better start looking at finance. So I'd love to get your perspective. How do you look at the finance department and marketing department as brands grow? Izzy Rosenzweig (10:52): I think as a CEO for a long time, I was first and foremost a marketer that became a CEO, right? And I think for a long time I adopted a story about myself that like, I'm not an ops and finance guy. I hire those people. But at some point along the way, I realized that like the scoreboard of business is in the financial documents. My success and failure is defined by what happens on my P&L, my balance sheet, and my cashflow statement ultimately. And so if I don't own understanding how to affect those things, I'm not denying my responsibility as what my job is for my shareholders. So that was like my first personal foray into deciding that I'm going to think about everything I do through the lens of affecting those arenas instead of these proxy metrics that often we grab onto as leaders. (11:34): And so that sent me on this journey. The other thing that we were experiencing was towards the end of COVID, brands were coming out of this wild demand period and then were coming to us with an expectation that the growth was going to continue. And so we were fighting unreasonable expectations. And if you're not sitting in charge of FP&A, then you are subject to somebody's plan. And so there was this need for us to get back into a position of being able to regulate and level set expectations and anchor them to reality that we had to try and do. So that was like, okay, now we have to get involved in forecasting. And we had to help them. What we like to say is like, our job is to tell you what is likely to occur. You get to decide what you would like to occur and then we try and bridge those two things. (12:18): And so that became our marketing strategy flowing out of an intimate understanding of what the organization craved financially and trying to bridge that gap. And then when you start down this path, what you start to discover are things that are like very simple once you realize it, but aren't necessarily your consideration in the front end, which is like SKUs don't have the same gross margin and therefore shouldn't have the same efficiency expectation. Some products don't have the same LTV as others. The return rates are variable on different SKUs. Even the age of the inventory should have an effect on your willingness to turn it into cash at different price points. These are all parts of, if you're going to affect the cashflow statement and balance sheet and P&L, your marketing strategy changes relative to the lens through which you look at each of those things. (13:02): And so it's just this process of as you go down this path, you begin to unlock different levels of strategic ideation to support different kinds of outcomes. Taylor Holiday (13:12): Love it. And I've said this on I think a few podcasts already prior to this. You said this before, to get started, e-commerce is nothing. No startup. You go to Shopify, it's about 20 bucks a month on what it's going for now. You create a product. It's really, really easy. And honestly, if you're a decent marketer, you'll grow probably a hundred grand really quick, then a million dollars, then $5 million. But at some point you will go bankrupt if you are not looking at your PL&L and balance sheet. And P&L, yes, it's gross margin on the different SKUs, returns, devils in the details. What is your return rate? Selling is a high. And also, how are you dealing with returns? Are you reselling it? Are you liquidating it? That's your EBITDA, right? And then the balance sheet is forget about how much gross margin you're making. (13:49): If you're not managing your cash, you're going bankrupt. So I think that's like the founders that start great, great businesses if you're not looking here, you're not optimizing. Yes, optimize marketing, everyone's going to screen you up to ates marketing, optimize your supply chain, optimize how you're buying goods, optimize how you're using cash. That is the difference of going from one to five million or five to 50 million. Izzy Rosenzweig (14:07): Let me just add one thing here is that ultimately your goal for your business to be to somehow figure out how you end up with money in your own pocket. Somehow, some way through all our labor and effort, that's like what we should be hoping to achieve. And I sat with one of our longest-standing customers is an amazing apparel brand and they got to a place where they were like 50 million in revenue, doing 12 million in EBITDA and had never distributed a dollar. Taylor Holiday (14:34): Wow. Izzy Rosenzweig (14:34): Because every year to grow the subsequent 30%, they had to tie up so much investment capital into inventory that the cycle just never ended. And so they were totally P&L rich and just totally cash poor. And it felt like it was so frustrating. And what I noticed is that they became desperate to try and sell their business because that was the only way they thought they could realize some of the value, but that meant ending up in these situations that they didn't really want with terms that were fairly onerous and lots of debt being placed on the business in various ways. And so learning and sitting with them and helping to say like, okay, what would it look like to alter the dynamics of how cash moves in this business? So you could actually realize some of the value for yourself, I think is a huge transformation of the way that you approach everything down to how you manage the Facebook ad account. (15:23): It changes everything. Taylor Holiday (15:24): It's so fascinating because I don't think that many people fully get the difference between P&L and balance sheet. P&L, great EBITDA business. They didn't have cash. Now I would say this is what Portless does, and I'm not here to talk about Portless, but literally Portless can take a $10 million inventory and remove that down to two million because your next year factor, you could take in goods and sell them and then recycle your cash. So you're either going to have to build for an exit, which still needs EBITDA in today's world that is not pipeline, or better build a model where you could actually take off EBITDA off the table because you have cashflow. Izzy Rosenzweig (15:56): And I would even argue that in today's market, if your business can't service debt, your capacity, your options in terms of the buyer pool gets shrunk dramatically because PE in particular, they want to use leverage to buy businesses. They buy Taylor Holiday (16:12): You a debt. That's their business. Izzy Rosenzweig (16:13): So if you can't service, which is just a function of how much free cash do you generate, is your determination of the amount of leverage you can handle, you are going to diminish the enterprise value of your organization and you're going to shrink your buyer pool. So in either case, whether you hope to sell one day or you hope to put a bunch of money in your pocket, I would argue that both require the ability to produce free cash flow. I Taylor Holiday (16:33): Mean, can't get more passionate about this subject and maybe you can stay here forever. But on that note, let's talk about AI. I know you're passionate about AI and everyone has to be. And there's some great articles on X going around that literally if you're not looking at today's version of AI, not 2023, not 2024, end of 25, 2026, the claudes of the world, agentic AIs of the world, there are areas that AI will eat. Now, I'm curious from your perspective, there's a lot of ways to look at AI. It could be from customer support to taking over the roles of humans to the way you actually run your business or the way you sell. Maybe you're selling to agentic AIs, not to go too extreme, but what are your thoughts on how agentic AI plays into e-commerce today? Yeah, Izzy Rosenzweig (17:15): It's a really big question. And I think you sort of hit the two sides that I think are worth sort of talking about. One is like, how do consumers buy and how do we interact with the agents as sellers, and then what is the labor effect of it? (17:27): I think that one of the things I'm experiencing right now that's hardest for me as a CEO is to appropriately level set the expectations of the capacity of my people in this world. How much output should a person be able to do? Because it seems like it's exponentially growing. And in e-commerce, this is really good news because as we talked about, if you can create labor leverage in this business model, you create the opportunity to increase the investment in marketing, which increases the potential for growth and it can also increase your margin capture. And so one of the things that I talk to people about all the time in e-comm is that e-commerce is likely not an FTE business because it's highly seasonal, right? So the idea that your labor force should constant or grow linearly, it just doesn't make any sense in a business where your revenue spikes grow like this. (18:08): So one of the things that I think AI introduces as an initial simple entry point is like seasonal scaling effects for volume considerations of different tasks. So you use the example of something like customer service. The amount of tickets that you get in November and December is not the same that you get in May and June. And so therefore the labor profile should change. And that's like a really simple illustration, but you could apply that to every part of the labor stack and e-commerce. The amount of ads you have to make, the amount of changes that the graphics that you create, the amount of photo shoots that you do, the amount of orders that you pick back and ship. All these things have a scaling effect relative to order volume. And granted, there are some businesses that grow like this, but most don't in this industry. (18:49): There's a lot more seasonality. And so some of these tools I think represent the ability to alter the scale dynamics independent, just adding more or subtracting more people in a way that is really good news. So that's like a simplistic version, but like the ability to take the same unit of production and scale up and down the production massively independent cost changing is like really interesting. In addition to that, the other thing I'll say is that e-commerce is primarily a business that's predicated on your ability to make an assumption about future order volume, place a PO and sell through that inventory just on time. That like is the process of maximizing cash. And fundamentally that's an exercise in some data science, some management, some visibility, relational management on the supply side and different things that I think a lot of these tools are just better equipped to do than humans, candidly. (19:38): They're just set up to process information and effects in a way that I think will help, especially unsophisticated people without a bunch of resources and teams to make those decisions better and more thoughtfully, which is hopeful for a better realization of value for people. So that's another piece of it. The same is true in my world, like in media buying, I don't believe that there's any industry more wrought with just like horrendous decision making on like wildly poor interpretation of data than media buying. And so I think that there's a huge opportunity for some of what we see. And just like we talked about where the capital moves is that you'll see these effects move from like what has happened in trading at hedge funds, is that you've watched this move to like this very real time programmatic, systemic, AI driven buying and selling decisions in hedge funds for stocks. (20:28): And the reason is, is because you can process a lot more information a lot faster, a lot more accurately. Well, the same thing will just move down the stack to areas of lower value capture in forms where you're going to see AI play a much larger role in media allocation and attribution and decision making. Whereas right now that's a giant quagmire and there's incredible amounts of inefficiency and waste that I'm hopeful is going to get a lot better. So there's so many different parts of what happens in e-commerce that are going to be really interesting. But the area of all of that, I still think that that's such a small part of the labor profile that like, or the revenue profile that it's actually not the thing that fascinates me the most. What I'm much more optimistic about is how low could fulfillment expenses get and what is going to actually happen to supply chain. (21:07): And I'd be curious about your thoughts around this. We live in a world where like my son 3D prints really cool objects in his room. And so I just wonder how much further down the line we're going to get on robotics and AI affecting those core parts of the business as like where I actually think the effects for e-commerce going to be realized way more than in the software and labor side, because it's just not as much of a profile. Whereas my business, human services, 70% of my cost is human labor. We are going to feel lots of effects. Our business is changing really, really fast, but e-commerce, it's kind of weird because it's only 10 to 12% of the revenue that's actually tied up there. Taylor Holiday (21:40): It's a fascinating subject and I go a couple differen directions here. So my first macro question is, is this going to be a deflationary advantage? Meaning will costs go down, therefore prices of goods go down? I don't know. I don't know. I mean, maybe I think if you go really far in the future and you're looking at robotics, looking at optimists, meaning could the factories in the States potentially be cheaper than the factories else in the world? Maybe, but they also have raw material problems. Where's raw materials coming from? I think there's a possibility at some point really, really in the future. Honestly, I don't how far in the future, pretty far where there's enough robotic expertise that literally local manufacturing plus the robots building the raw materials locally, finding fields, planting cotton, like really end to end. I think there's a long time before that. (22:24): But when you get there, I do think we would drive down cost tremendously. On the upstream, the current supply chain model, there are areas of robotics, which are by the way already pretty good as it is. The modern fulfillment centers are very robotic. I mean, you look at Amazon and these of the world, like holy crap, unbelievable. So then there's where else can you apply that? So I think the inventory predictions I think are really good. To me, if you tie that with a different business model, agile just in time inventory versus the traditional inventory, but I think the labor will go first, to your point. I think the areas you talked about, customer support, media buyers, how do they, again, leverage. They will still, I think exist for a while, agents, creating agents might get a little scary, but I think it'll take more time for the real world. (23:08): I think for atoms to change, we'll take some more time. Izzy Rosenzweig (23:10): Some of the areas where I think we're going to see prop up and I'll be curious about this. So one experience that really transformed, like was a whoa aha moment for me was when we started working with ColorPop. Are you familiar with ColorPop? So ColorPop- Taylor Holiday (23:22): I appreciate that. Izzy Rosenzweig (23:24): Big makeup brand, they're a subsidiary of a manufacturer called Seed Beauty that also does Kylie Cosmetics. They have a huge facility (23:31): In Oxnard, California. So we went to their office for a meeting and they walk us around and they take us into their warehouse. And inside of their warehouse, there's this laboratory that's fundamentally, think of it as like a micromanufacturing facility where what they do is Ulta.com launches new SKU, it starts trending. We can make a batch of like a hundred units here. We can get it on the website and we can test the velocity and the demand and then make our larger PO to our scaled production overseas. And what I think that is going to happen is that people are sort of realizing that product expansion, that almost every, one of the problems with e-commerce, I think e-commerce is generally deflationary because it's the competition, every marginal dollar gets competed down to zero in almost every bit of fixed digital real estate on the internet, especially in like an environment like Amazon where it's more commodity driven objects. (24:23): And the more commodity is, the more that this plays out. But what happens is when that happens for a business, their only growth mechanism is either broadening distribution, so retail, international, whatever, or category product expansion. And I think a lot of businesses are realizing that like product expansion, that the investment against that cost relative to the value potential is often way, way better from a ROIC perspective than is the next dollar into Meta. And so they're starting to see that like it's actually way better if I try and develop a new SKU. And so what that will do is cause innovation around how low can I drive the cost of launching net new things? And so some of this, I imagine if you give an AI a balance sheet and you ask it to make considerations, and this is the thing that I think the biggest mistake people make with AI is they try and take their present workflows and they like try and create efficiency in the present workflow. (25:12): And I think that's like really dumb versus like, here's the big picture. Now, what am I missing? Where would you invest or allocate capital as the business owner? And I think what that'll lead to is like these different ideas about how we test and build and launch things that will lead to much more rapid iteration and testing of net new products in entirety that in a physical world felt abnormally impossible. And you're going to see this like manufacturing base that isn't about trying to do large scale supply chain as much as it is about a hundred units fast. And I think that there's this thing there that I think is going to be an example where it's going to be this like AI defining these new category opportunities based on trends that they're from data that they're sorting from TikTok and all these other places to say, "You should launch this SKU now and then how fast can we get that done?" And I'll give you an example where ... So we had this customer called Breaking T, and the print on design world is sort of a proxy for this all the time because they have basically no base manufacturing costs, it's all made on demand inventory. (26:12): And so they built this thing that was basically a social listening tool for X that as a sports event happened, they could determine the velocity or like the volume of demand that would come out of that moment and therefore how many designs of shirts and what the expected sales volume would be on it. And they built that like as a human model, just some basic ML stuff. But I think that that as an AI listening that says, "This is taking off, make this skew, get it live on the website, batch a hundred units, put it live." Humans might not even be involved and there might be products for your brand that you didn't even know how it got there. That level where AI can absorb so much constant purview of information is where I think there's like some really interesting things that could occur. Taylor Holiday (26:54): So much to unpack here. So first of all, the idea you said like workflows or your current workflows are defined by literally you're living in a box. Software was created in a box and you are following the software's rules to create your workflows. With AI will help you create what you need for your business. Not that you're living in their box of a software provider, AI will create the software you need that fits within your workflows to make it run really, really well. And then regards to the predictions versus reactions, I, this morning spoke to LeFung, right? Leafhung is one of the largest trade partners in the world. They see all types of manufacturing. Until today, projections manufacturing is the way manufacturer is done. You got to predict four months, six months, eight months ahead of time. That's a ton of money upfront. That is a ton of money lost in excess inventory and liquidation, and just the way of the world till today. (27:43): But the world is moving to a reactive model. And again, this is the world I live in. If you think about Sheen and the Teamuts, shots on goal as fast as possible for the lowest cost possible with no timeline delays. Because if you're going to do shots on goal and wait for two months to get your goods, the season's done. The trend has finished. You need to be reactive to the market now. There is an X trend, there's a TikTok trend, there is a new design. You're producing 50 units, 100 units, you're testing, you have data, you double down, you're scaling, and some of those stuff will be future essentials, which then you go law, big manufacturing, South Korea, wherever you want to go. But you need to be world of a reactive on demand type of business model. And that's where the world is going. Izzy Rosenzweig (28:29): What you're describing, so all the time, because we exist in forecasting, everyone, anytime I have a customer that starts wanting to spend a bunch of time on assessing my models, I go, "Oh, you don't think about this right. This isn't about how good can we get at guessing. This is about how fast can we react to being wrong and fix it. " When I think about inventory forecasting, people will want to go big brain forever into trying to guess right. And I go, "If I was in charge of this, what I would want to do is how do I build a more flexible supply chain?"That's actually how I could be right more often is that I don't have to take these guesses on these really long horizons where the number of variables in the model become innumerable. It just reaches a point the further out you go that your ability to actually include every potential factor is impossible. (29:11): And so you don't want to play that game. And I think that brands are realizing this, and in some ways it all circles back. It's like full circle to drop shipping as like, this is actually the right e-commerce business model. There's a reason why this was the thing where it started is because it's pure. It de- risks the biggest problem in the Taylor Holiday (29:29): Whole thing. It's truly intense. It's truly direct consumer. And it's funny, Drop to me has this bad connotation because the OG was, I'll express crap products, but really it's about, no, take beautifully designed products, take quints, quince level products and ship it as agile directly to the customer from where it's being manufactured. I would love to cover scenarios. We just talked about reducing labor costs, however customers for probably marketing efficiencies, probably product design, tech back creation. There'll be designs in general and website software costs. How about the buy side? Do Do you see, and I guess this is very debut area, PBC, Meta, TikTok, kind of still Wild West TikTok, but are people buying differently now? I know people talk about OpenAI. Is that real? Is SEO needed now there's ads in OpenAI? Where's that going? Izzy Rosenzweig (30:15): Yeah, I'm a big believer that in what I call omnipresent commerce. So the people think about Omnichannel, but that's just a mapping to the present points of distribution that we understand. Whereas what I see happening is that I have this sort of underlying principle that I believe that transaction always moves as close to the point of discovery as possible. And the biggest illustration for this is Apple had this incredible innovation in their retail stores where they brought the register out from behind the register. And it's like the innovation was like, "Oh, the customer's right there. Let's transact as fast as possible between the point of desire creation and the ability to purchase." Now it's TikTok shops. TikTok shops is I'm watching a video and I can click and have it right now and be able to transact immediately. And if we think about the way that the world will interact, what will the platforms of the future be? (31:01): I watch my children in my house, so I have twin boys that are 12 and a daughter that's nine. If we were to get into an argument about a fact, like my boys are arguing about how many home runs did Cal Raleigh hit last year. You know what I would do? I pull out my phone and I Google the answer. You know what they do? They ask Alexa. Their source of truth is not my source of truth. Where they go to find information is completely different than me. And ChatGPT will just be another layer on this and then our glasses will be another layer on this. And in all of those arenas, there will be an expectation that I can transact in that medium. To me, the website is like a really archaic idea that I have to see a product that I like and then I have to pull up my phone. (31:41): I have to open this thing called a browser. I have to type in a URL. I have to go to a place and then I ... That sounds fast, but that's like a thousand steps. That's insane. Versus if I have an omnipresent agent and I actually saw somebody do this, they created an app for the Metaglasses, Meta Oakleys where it was like they're in the world, it stored my credit card and I go, "Oh, hey, see that right there? Can you ship that to me? " And it's like, boom, it's at their house 24 hours later. Where did they transact? They don't care. I don't care where I bought it from. It's like, find the best price and the fastest delivery and have it at my house tomorrow. Done. It's Taylor Holiday (32:14): Funny. By Gen Z versus millennial comparison, like millennial, we could do small transactions on a phone. Big transaction, I need a computer. Yeah, exactly. Gen Z is like, it's on the phone. Versus the next generation is like, it's all open the eye. It's Siri. I don't care. Izzy Rosenzweig (32:31): It's funny. I even think about, so I've interacted, this is in different ... The financial world is a very fascinating version of this like you're describing where ... So I play in the crypto world, I got caught up in the NFT phase and the amount of money that I would be moving around and now it's like if I want to wire something out of my JPMorgan account, my banker has to call me. And I'm just like, how do these two worlds exist simultaneously? Where on any day ... I saw the amount of money that was bet on call sheet on prediction markets doing the Super Bowl was like this insane $2 billion or something. And you're just like, and yet over here, my JP Morgan banker wants to call me about a $1,200 wire I'm saying.You're just like, the world is like so insane in terms of where it's all headed that I think that we're going to have an agent that's shopping for you. (33:19): They're going to be interacting on your behalf all the time. How you prompt and get them to begin acting, I think is going to happen in all sorts of different mediums, including ones that are yet to be discovered. And eventually we're all just on Neuralink and who knows. The idea though that my eyeballs are discovering a thing in the world, how quickly can that thing get to me is ultimately the game. And it's also why I think that Amazon is in a good position to win a lot of these feed-based demand things because ultimately what the user's going to want is access to the object fast. And if brands don't begin to contend with that capacity, I just don't know why the AI would ever choose not same day delivery if I have the option for an object. Taylor Holiday (34:01): Totally got it. And then this area that like brands are very good at marketing ideas, et cetera. So I think a two-part question. One, I think you've talked about ideas versus systems as a topic and what that means for an entrepreneur. But as a kind of connected to the last part of the AI before we move out of like, I guess AI and back into systems, where should people go? Is it Cloud? Is it OpenAI? Is it certain softwares that are focused for e-comm? I don't know if there's a Claude for e-com yet that I know of. Where do you look at as an agency that you've probably seen the most of those? Izzy Rosenzweig (34:32): Well, so I think the challenge right now is that the prompt interface is a bad thing to help an entrepreneur because it necessitates asking the right questions. And the problem is actually people don't know what to ask. So until there's this more proactive engagement into the problem that you are actually ... To me, the Shopify of the future is me saying, "I want to build a store." And it starts telling me everything that it's going to do on my behalf more than it is like I'm dictating the things that I want to do. It's like I'm going to have this general desired outcome and I'm going to be sort of interacting verbally around like, I have this vision for a brand, it's like a kid's sports training product and it works like this. And it's like as I'm interacting and verbally communicating, it's just being created. (35:20): The supply chain's being set up, the website's being built. It's all sort of happening in a way that I think to your point, does the cloud of e-commerce exist yet? I don't think it does. But yeah, so where should people go is an interesting question. I don't know that there's an obvious really immediate leverage point that I can say definitively is like solving something for everyone yet. I just don't know what it is. And you look at these things- They look Taylor Holiday (35:45): Like a microservices. It's customer support, there's marketing, there's tech pack. Izzy Rosenzweig (35:49): Yeah. It seems like most maybe adopted is probably like a rich panel for customer service feels pretty, it's getting close to ubiquitous in a way that's like, "This definitely saves me money. It definitely is more efficient." So that's happening. I think that we're seeing a lot of work in the creative side, but there's almost nothing that I like there yet that I've seen that has been really impactful to people because they're all constrained by the governing ideology about how many ads you should make, what the quality should be, how that workflow happens. They're like really gummed up by people's present expectations of that. So I don't know. I don't have a lot of things that I say this is the definitive way to do it. What we're trying to do, I know what we're trying to solve for, which is like build an intimate connection between the systems of the organization, the systems of record, if you will, between marketing, finance and operations, where to me, those things being disparate and giving the capacity to organize it such that you can create AI interactions on top of it is a problem too where I think e-commerce suffers really bad data infrastructure as an underlying problem that I think is going to be a big business model in the future is just sort of this like someone helping build organized data structure to this information where like COGS still lives in a lot of spreadsheets in places that are, and POs are fragmented and the balance sheet hasn't been updated in eight weeks and the P&L's wrong. (37:10): And there's still a lot of that in e-commerce where these AI tools are like, they're correctly reading your bad information and providing you back advice on the top of it. Taylor Holiday (37:17): You're only as good as your data. Izzy Rosenzweig (37:19): That's right. Taylor Holiday (37:19): And a great agentic in the future is going to find the right data, know that it's wrong if it's wrong, match it to PO. That's Izzy Rosenzweig (37:24): Exactly right. Taylor Holiday (37:25): And be proactive versus reactive. Izzy Rosenzweig (37:27): If there's a thing I think I could encourage brands to do, it's actually this idea of like, I think there's going to be a big service around, I call it like database by design, which is just basically bringing into the purview of the AI all of that information because it can't solve for the PO that exists on your computer that's detached from its access point. So somebody's just, because I agree that it's not about like ... It used to be build the data table. I'm not saying that the game in the future is build a specific structured database, but the AI has to have access to the availability of all of the information in some way in order to interpret any of it. And so I think that that piece of like, where are the POs? Where do COGS live? What does the inventory count? (38:07): All of these things that are necessary, I think somebody's job it is to organize that. Where are all the marketing APIs connected? Okay. For ones that are manual IO, because people still buy podcasts on like manual IOs where literally it's just like an email between you and the podcast provider that's determining the purchasing and the days that they're due to QuickBooks. And that organization I think is going to be really important for brands to think about. Let's imagine there was a superhuman agent that knew everything and could give you all the advice in the world, would they have access to the thing that they need to solve the problem? That's like an important question to ask. Taylor Holiday (38:42): Yeah. And in today's prompt world, that doesn't exist. All right. So I think you got tailored from the marketing side, you got me from the supply chain side. You can't find two biggest advocates on all parts of the business. Never been a better time to build a business from. I Izzy Rosenzweig (38:55): Agree. Taylor Holiday (38:56): But build it from the ground up in a healthy manner, EBITDA focus, cashflow focus, and it's a big world. You call omnichannel, omni distribution, and globally, like at least on poor list for huge fans in global. There is a big freaking world out there. You got a great product, tell your story and do the distribution. Prediction time. Five years ahead, what is the thing that is in the e-com market that's helping brands scale? AI, not AI, whatever you think. Izzy Rosenzweig (39:22): I think that we are going to see compute move a lot closer to our physical presence, whether that's glasses or some future device. And I think what that's going to create a prioritization of is like, where does your product exist in people's worlds such that they encounter it? So I think that we're seeing this emergence of ... One of my favorite examples of this is like high rocks as a e-commerce business model where it creates the physical reality that you exist in as much as the product that you use. I think golf is another example of this where there's this overlap between the culture of your life and the product category itself, such that your discovery and interaction happens in those spaces. Right now, the problem with why the phones are such an addictive behavior is because they unlock this incredible world, but we have to come out of presence in order to get there. (40:09): And so I think as we move compute closer to a lens of reality where reality and that lens are sort of more in sync with one another, it actually could bring an increased importance to the physical world in a way that we're not being pulled out of that into our phones as much as there's this layer of what I'm experiencing all the time that's bringing in other indications, but it's an amplification of my reality. I like the idea of AR. I think there's still so much to be discovered there. So five years out feels like, I don't know, we could all be slaves to the robots too. Yeah. I think the idea of the robot in the home too is going to become ... I'm a big believer. I can't wait for one. I would love to never do the dishes again if I could. (40:50): So I think that will unlock so much change in what we do. I heard a great thesis. I forget who it was. I want to give them credit that basically if you look at what we did during COVID as an example of what happens when all of the work gets sort of stripped away and you saw this massive influx in outdoors hobbies and things like camping and paddle boarding and all these activities as like a proxy for what happens in a world where robotics and AI do all of the manual labor stuff, I actually think that that's sort of true is that if you look at what the behaviors were when all that stuff got obfuscated away, I actually think that that could be an example of what the future might be like. There you go. Exactly. What do you think? Wait, I want to hear. (41:34): Izzy, what do you think? You've got a view of the future. What happens in five Taylor Holiday (41:36): Years? Yeah. I think from a five-year game in the future, I think manufacturing will evolve a lot. I think to your point in the e-commerce world, where do robots take it? I think there's only so much kind of software engineering reduction you could have, and that world will be its own chaotic world. And where people go for jobs in a while is its own question. But for e-commerce specifically, I think a lot will ride on the evolution of robotics and the optimises of the world and what that will do to supply chain. I'm hoping deflationary, and I think it is deflationary, and I think that's the world that Elon talks about. If you reduce costs so much, everyone is wealthier because you're reducing the cost of actual labor. So that to me is like a better world there where costs are going down and brands are still running healthier businesses. (42:22): Taylor, this is incredible. Very, very few people I think in the e-comm space like yourself where you have marketing of finance operations, people want to learn more about you. Where do they find you? Izzy Rosenzweig (42:32): Yeah, commonthreadco.com is where you can chat with us if you're a consumer brand looking for some services in these areas. If not, I'm at Taylor Holiday on X. My DMs are open. I'd love to meet people on there. Taylor Holiday (42:42): Awesome. Taylor, thank you so much. Izzy Rosenzweig (42:44): Thanks, Izzy. Take care, man. Taylor Holiday (42:48): Thank you for listening to The Modern Supply Chain. If you have questions about anything we talked about, you can find me on LinkedIn. And if you're interested in learning more about Portless, check out our website, portless.com. As always, hit that follow button so you don't miss an episode. See you next time.