The Next Trillion-Dollar Company Won’t Be an App - It’ll Run on Robots and Real-World AI

Speaker A: Welcome to Retailistic with Debra Weinswig, brought to you by coresight Research. In this episode, Debra welcomes venture capitalist Brent Murray to discuss the rapid evolution of technology and how to leverage it for strategic advantage. Brent shares insights on the impact of AI in physical supply chains and the importance of a builder mindset across industries. But before we hear from Deborah and Brent, a quick word from coresight Research.

Speaker B: Coresight Research serves the retail community with timely research on consumer shopping behavior trends across every retail vertical, financial outlooks for industry leaders, and the latest breakthroughs in retail technology. We also develop and maintain proprietary data resources, conduct technology assessments, provide strategic consulting, facilitate leadership communities, conduct seminars and conferences. This week, our premium subscribers will see a behind the scenes of Shop Talk Europe. Visit us@coresight.com to subscribe and gain access to these reports and over 8,000 other reports, webinars, videos and data banks. Brent, thanks so much for joining Retailistic today. We're excited to have you.

Speaker A: Thank you, Deborah. Excited for the chat.

Speaker B: Let's take a big step back and look at kind of what's happening in venture today. It's, I mean, I guess where I'd love to start is how are you staying on top of everything that's happening? Because as I was talking to a startup yesterday, he's like, at 2am I'm talking to folks in Silicon Valley and they're coding all this stuff and he's like, I'm just trying to stay up with what they're doing. So, so how do you stay on top of everything these days?

Speaker A: So great, uh, question. I mean we are, we are in an era of information overload and innovation seems to be happening at an increasing rate. Right. The things that are coming out just tend to one up each other every, you know, every few weeks. So I do a lot of reading. You know, I block off a lot on my calendar, uh, especially in the mornings when my mind is fresh, um, and weekends when, you know, I'm not so distracted and bogged down by the day to day and so I'm reading a lot podcasts on my commute, I do that a lot. And so I stay up to date and then I just spend a lot of my time talking to operators. And so whether I'm investing in commerce, uh, I'm talking to a lot of brands, I'm spending a lot of time in AI in the physical world. And so I'm talking with a lot of warehouse operators, three PL operators to, uh, understand what they're seeing. And so it's a mix of a lot of reading sort of secondary research and then primary research of talking to the operators that are um, implementing the technology and then just a lot of hands on keyboard. You know, I need a lot test a lot of the things that I'm trying to invest in to see what's changed. And it's very hard to have a perspective on something if you're just looking at that one thing. But if you can compare it against the 10 other tools that you have previously used, then it gives you just a much better perspective. But I have no, you know, I've been doing this for about 10 years and I have noticed even in the last six months, you know, a fire hose approach of you just have to be on your toes and on your game a lot better because you know, the pace of innovation is happening so fast. And while that is scary in some regard, it is such an exciting time and such an exciting, you know, thing that when, when you use some of these magical platforms and it just changes the whole way you work. Uh, it, it is a magical moment and so it's very exciting time to invest. But also, you know, you need to, you need to stay ahead of innovation.

Speaker B: It's so interesting, I mean so much of what you said. So I'm, um, I think I mentioned to you I'm, I'm literally not traveling the whole month of June. And uh, what I found, let's say for the first five months of this year that I started to do a lot more building which I really enjoy. And so I've got like a Mac studio and just the, the whole gamut. And the reason I mentioned this is there are so many new startups and I mean new like in the last 12 months, by definition new. And I'm talking to them, I'm um, understanding kind of like what they built, how they built and why they built it, but going back to like testing it, we've actually started in some cases, like we'll wall off part of like what we do to try and actually test on ourselves first. And we've never done that before. And so because it goes back to how do I, how do I really know that what they're saying? And it's not that, it's just kind of like what the actual outcome is because especially if I'm talking uh, to a technical founder, right, I code and everything, but I don't have a technical education. I just want to make sure that what is being said is what's actually happening. Many cases it's better, right, because you know, the builds are happening so quickly. But, but how are you. Can you talk about like this idea around, like testing? Because. Right. We spent a lot of time in logistics with 3 PLs, a lot of time with retailers and brands, and I find this is the biggest challenge that they're facing right now.

Speaker A: Yeah, so we have, you know, we actually just uh, went through this exercise as a firm. We do have an internal tech team that is, is very good and they are staying at the forefront and helping us adopt. Uh, but we just had, you know, a, a pretty big meeting as a firm where we sort of walked away with this mantra of everyone's a builder. And so I think historically, you know, it first came up with the companies that we've been talking to, potential investments, where they talk about, you know, these are the bottlenecks that used to exist, used to have marketing that, that handed off design to or handed off an idea to a designer. And then they would have to wait for the queue for engineers to code it in and, you know, make it a reality on their website. And now sort of everyone's a builder. There is no longer that bottleneck. It's the same at a venture firm. You know, historically when we wanted to implement a change, we had send it to our tech team, it would go through a queue, we would iterate with them. But we walked away with this mantra, um, very recently of everyone's a builder at M13. And certainly we have to be consistent across the firm to be all using the same thing. But if we want to test out a new process, if we want to engage with Claude code or Codex or Cursor and build something, we can do that and we have the power to do that. And then we share it with each other and then needs to be that, you know, absolute level of collaboration. But it is very necessary for us to test things and allow everyone at the firm to be a builder. And that's what we're starting to see in the, uh, companies we invest in. You know, whether it's a commerce enablement, you know, agentic first approach, a consumer brand that they're selling to everyone is starting to become a builder. And it's very, very fun to see.

Speaker B: Yeah, you bring up a really important point. You know, we started our AI council in August of 23, so I can't believe we're almost closing in on three full years of these. For the first, uh, like 18 months, we met twice a month. Now we meet once a month, but completely closed door, uh, C suite. Uh, we're having conversations that have truly changed how I think about things and in our last meeting an executive asked the question right like she was asking all of us build or buy? And it, it was a really, it was a really interesting discussion and I, how, how are you thinking about that right now from a tech perspective and how are you infusing that into, you know, your portfolio?

Speaker A: Yeah man, it's such a, it's a great question. Uh, because we are, you probably see this internally yourself. The models continue to one up each other And I think 12 months ago it was, it was pretty clear OpenAI was, was the leader, everyone is using them. I think you know, three months ago it was pretty clear it was the opposite. Anthropic had taken over, you know, Cowork was being used by a lot of uh, the business folks and Claude Code had taken over and now, you know, with releases it might be slightly different. I still think, you know, Anthropic has a good lead but you know, as everyone's been building, they see the power of these models and they see the power of the applications they can build from them, um, and the change that it's made in the organization. At the same time we are all collectively as venture firms investing in companies that are not OpenAI and anthropic, of course a lot of VCs are and they still are and they're pumping money into those, those, those companies which is great. But the, all the other investments that are made, they're being made in hopes that people will buy and not necessarily build in some circumstances. And so we are approaching everything with that lens. And you know, if I'm going to make this investment in an AI application, I really have to be sure that it's not just going to provide a wrapper and a cost markup to the customer, but they in reality need something like that. And so we look for proprietary moats that the companies have that we're investing in. You know a great example, one investment that we made, you know, about a year ago and they just raised a follow on round is business, um, called Cord and Chord Commerce provides the context layer for consumer brands to run their operations. And when we started out with Cord, Anthropic, OpenAI they weren't really on the competitive landscape, you know, yet. Like yes, you could cloud code, a lot of functionality, but everyone sort of knew that you needed a little bit more business context. But I would say over the last three months the conversation has come up more as, as Cord has been selling into E commerce startups. You know, they say kind of like you, like uh, we're starting to build this ourselves. And so we've been doubling down on, um, you know, what is defensible about Cord. And it turns out that having really deep context on a company is not something easily vibe coded. And it takes a ton of expertise on the backend to enrich data, to ingest it, the right way to model on top of it. And the way we talk about it internally is AI provides an infinite number of interns. And that is fantastic, right? You can have an intern doing every task imaginable, but if you don't have deep context and memory on what the AI is doing, it's like hiring an intern without that, uh, few months of deep training the interns need. Interns can come in and do a ton of output, but if they don't have the training, they won't have the context and they'll do a ton of work that doesn't really deliver the end results. And so when you have a context layer like Cord, it provides your interns with that ability to train and understand what are the unique business functions that coresight has that other research firms, uh, and implementation companies don't have Jargon that you use internally, systems and processes that you've refined over the last several years to make you truly great at what you do. And so that's where we've been able to find, and we go head to head with, you know, a, uh, technologist that is saying, oh, I'm building this internally. Then we test out the results, you know, one by one, and they say, okay, query the data, ask it something and then make it, give it a suggestion to you. And time and time again it turns out that, you know, that the layer with, with deep and rich context produces what a experienced operator would produce instead of an intern that just started yesterday and can get you the job done very quickly, but it doesn't really, you know, provide quality output. So those are some of the things, you know, I'm giving a specific example of one of our companies, but one of the, you know, big things we look for is, is what is that, you know, proprietary moat that they have, whether it's data or a context layer or uh, something like that, that they're, they're going to have defensibility against the, the large LLMs.

Speaker B: I think that's a great example. And, and I, I was smiling because this whole idea around interns, I've, I've been with a lot of executives in the last few weeks and they've all said that, first of all, I've never Heard of so many high school interns. I was actually at a. More on the medical field, uh, two nights ago, a research dinner. And this, it, uh, was Columbia University and they had two high school interns and they were talking about. And I've heard this, like, this is like, I feel like this is a theme of the last seven days, if you will, how these high school interns are starting to really. And I find like it's almost high school interns, like sophomore and high school, the sophomore in college right now, whatever that age window is that they. Because first of all, they're, they're still very humble and they're, they're asking like such great questions and trying things that I find that their, their value. I mean. And most, most of these are unpaid interns. So you can't even look at like an roi, although ROI on your time that you're teaching them. But they're having, right, an exponential impact on organizations. And so it's really interesting. We have our, uh, I think we have five interns this summer and we have our first ever high school intern. And what's fascinating is we kind of, we do a whole matching process, but we actually have a whole agentic team. And uh, that's where he wanted to work and they interviewed him. They're like, this is amazing. And, and so just kind of this idea as well of how are you almost changing your organization's thought process at a faster rate and who's doing that for you? I mean, it's interesting no matter what any of us do, like, we need that kind of almost. I think that the questions, as opposed to someone telling you when people ask you questions, I feel like you, you change in a different way. So how, how are you finding. Right. And I know that that wasn't the point of your last, but it, but it brought up this idea that's been like. It's actually just been the back of my mind. I'm like, why, why now? Like, why now are we really leaning into kind of this, this younger talent and why are they having such a big impact on all of us?

Speaker A: Yeah, it's a great question. I mean, there is a corollary to what's happened in the enterprise world, right? It used 10 years ago you hired IBM, or this is more 20 years ago you hired IBM because no one ever got fired for hiring IBM, right? It's the trusted sort and it's been around for decades. And you go with tenure, right? And it's this hierarchy. And even 10 years ago, you know, you go with the, the biggest SaaS company because you're not going to get fired for implementing the biggest SaaS company. It's the most trusted source. And now when you're looking at the enterprise level, companies are doing this build, right, and they're dropping their trusted SaaS platform they've been using for 10 years and democratizing the right to do the job to whatever technology does the job the best. And so I see that in talent wars now too. It's not about tenure, it's not about, you, uh, know, experience in the field. Those are all helpful if you can also do things the most efficiently. And so we used to, you know, our evolution 10 years ago when we were hiring at the firm, um, we looked for our interns, we looked for MBA hire that had experience in the field. Uh, they, they could come in and sort of have that maturity level to do, to do the job well. And it's not saying that we don't do that anymore, but that's not exclusively what we're looking at anymore. And we, we didn't hire anyone from high school yet. We'll probably get there, but we used to rarely focus on undergrad, um, because we were looking for that level of maturity. But this year we have a couple undergrads coming in, 19, 20 year olds. And we found that the number one indicator of success of who's going to do well at M13 is a high sense of agency, sort of breaking down those barriers and democratizing the ability to build. Because they're close to AI, they're AI literate, they have a high sense of agency, they want to get things done for the sake of getting done, getting it done. And where they lack in experience, they have that grit and tenacity to um, you know, see things through the end. And so agency is, is definitely where we focus mostly when we're hiring. But yeah, we're seeing an enterprise world and again in the, you know, talent wars, man things people are getting younger and younger because they have that high sense of urgency and they really want to build some really great solutions.

Speaker B: Yeah, it's, it's interesting. There's been, I would say a lot of opportunities that presented themselves in the, in 26 around companies, uh, that we can either help others with in terms of due diligence, we can invest in ourselves, et cetera. And what uh, I'm finding is that the process of almost review and due diligence, right, in some cases it's sped up and others it slowed down. But the way that we even think about putting things into buckets, if you will. Like the buckets are changing. And so as you think about physical AI, a lot of where, where we're spending our time right now, where do you think we are and what is. Let's. I mean, maybe like next six or 12 months. Because I just. We're so much farther ahead than where I thought we would be. How are you looking? Just six to 12 months out.

Speaker A: Yeah, we're, uh, this is, I'm, um, glad you brought this up because we in venture, you're always trying to look into the future and not only what's happening in technology, but the ripple effects that technology has for years to come. And so the best example was in the last technology wave, before AI, you had the mobile and cloud wave. And this was kind of the 2010s. The iPhone came out in 2007 and that was the big technology wave that allowed for the mobile and the cloud era. But it wasn't really until the mid-2010s that that started to take off. And the first ripple effect was the mobile app ecosystem. And you had for the first time all of these mobile apps that were consuming a large part of our days both in work and fun and entertainment. And then from there, you know, the next ripple effect were sort of the new economies that, that birthed, uh, because you could have a taxi sort of on your phone through Uber. This whole concept, the gig economy started to take off and the major ripple effect was now all of a sudden 1099 and contractors were a major part of the workforce. I mean, that's just something that you really wouldn't have predicted, uh, in 2007 when the iPhone came out. But these are the reverberations. And, and so we, we think about right now with LLMs, that is the technology wave and it's going to be the largest that's ever been in history. So what are the ripple effects, uh, that are coming, going to come out of it? And you know, one thing that we're thinking about now is AI in the physical world. You know, if you've gone to warehouses, three PLs, logistics centers, sort of like the, the, uh, the services industry of our economy. Um, for the last several years, especially in Amazon, a lot of things have been automated, right? You have sort of sorting systems, you go to a lot of these facilities and there's machines running around. But autonomy has never been, you know, able to be produced in a quality way because, you know, you just didn't have the LLM systems, ah, that you do today. But now, you know, thanks to a lot of the power of LLMs, you have robots that are able to make autonomous movements and you know, human like movements with either five fingers or two, whatever it is. But it can do a lot of jobs that a human can in a warehouse. Um, and so you're starting to see this, this wild adoption inside of the US manufacturing and supply chain industry of operators. And when I say operators, you know, warehouse logistics, three PLs that want to be on the forefront, they don't want to be left behind. They see what Amazon has done over the last five years and really turning their um, warehouses and you know, next generation type of fulfillment centers. And so now you're seeing that, you know, play out in the physical world with three PLs. And so, you know, we just see like the LLM is the next technology wave and what are the reverberations and the ripple effects that are going to occur. You're going to see a lot more adoption in the physical world. And uh, robotics is a scenario we're spending a ton of time in, specifically in the supply chain sector because it seems to be an area where operators in that field really want to invest in the next 10 years of operational um, efficiencies because you can finally get to very, very good margins when you have robotics working alongside humans to maximize output and throughput of these facilities. So it's a really fun area to invest um, in. Just yesterday I was in Brooklyn, you know, visiting a couple of these, both robotics and some of their customers, the logistics providers. And it's just a very fun thing to see. And um, you know, technology taking over the physical world.

Speaker B: Obviously. I've spent most of my past decade and a half career in supply chain logistics and just the. I have to be honest, I think I thought by this point we would be further ahead of where we are. Yeah, because as I started to walk around at Liam Fungi and just you know, see pervasive manual labor, I mean we're talking faxes and, and fairly manual work. And then because of the manual aspect, if we think about it on fit, which I want to talk about as well. But right, like, so if someone was physically cutting something, somebody in the same factory could cut something inside the line, on the line or just outside the line. So you would literally have jeans is where I think about it. Like coming out of the same factory, same lines, but people like literally like hand cutting in different ways. And so I just. And then, right. We would see a lot of rejects, et cetera, and a lot of kind of rework being Done that was that you know, I would see on the factory floor that was quite expensive. And then as we, as we move forward and we started to work with some, some drone companies more kind of trying to look at in a warehouse setting like trying to better understand the flow of goods. And so we're still. Right if we. And I'm talking outside of the US right More of a global opportunity to you know, digitize digitalized supply chain. Like I uh, like I felt like we were Sexy for like two months in 2020. Like yeah, even pinpoint, it was like mid April to mid June and then it kind of like fell away because it's. So first of all I think the investment opportunity is immense because it is such, in some ways it's such a difficult problem to solve. Whether it's people being resistant or just like really connecting from raw material to consumer. Because that, that's how I define. Right. Until it's in the consumer's hands. That's, that's how I, I personally define supply chain. Yeah. Like all of the touch points and then just making sure that the, the accuracy. Right. They're doing these like digital. I mean I'm on the board of afa, the whole like digital passports. But it's, it's immense. So I, I've spent a ton of time in like lidar and other. Right. Like sensing technologies. Right. Which you know that, that because there's, there's so much. So uh, going back to. For me the big change was in January of 26 this year at CS. The CEO of Lenovo had on stage the CEOs of AMD, Qualcomm, intel and Nvidia. And one keynote at the Sphere and literally of a three hour keynote. I was truly transformational. Talked about how he believes that the biggest impact will be on supply chain and how that kind of comes together I think remains to be seen. But to me it was kind of almost like the demarcation. Right. Like we've all tried all of these things but now as you pointed out, right. We have LLMs and with having all the chip companies on stage. Right. We have the compute because I think part of it, I mean like I was building rendering farms in like 2016. Right. I mean we didn't have the compute power. So if you pull all this together because it goes back to. There's so many different slices of the pie and it's so, so complex. Where do you think about starting? Like where is the biggest problem to solve?

Speaker A: Yeah man, great question. Let's see. I wish I Had a crystal ball? No, we've been taking it. Uh, you know, I always like to, to bring in analogies and I think, uh, because when we first started looking at robotics, I had some of the same thought as you. Well, I've been investing for a long time and you know, one of our portfolio companies that just announced an acquisition, they were acquired, is a company called Passport. And Passport, you know, helps US based brands, Consumer Brands International. And they started in 2019 and you know, just ended up having an exit. And over the last several years they've been building a lot of really interesting solutions for US based brands that want to go international. It started with shipping logistics, but then, you know, expanded to duty and tax calculation, customer returns, reverse logistics, even marketing inside of those international, uh, countries where you're trying to ship product. And you know, when we invested in 2019, going into 2020, there's a lot of talk about autonomy, uh, in their network, right? Then their network being sort of the logistics providers that were helping fulfill that package. And I was thinking, you know, in 2020, 2021, wow. A lot of people are investing in robotics and it just seems like everyone's going to do what Amazon has already done and automate a lot of their processes. But when you, you actually get into it, it's sort of like the, the digital application of LLMs. You know, several years ago it became really easy to do a research project on anthropic or OpenAI. Right. You type in a question, it would pump you out a research report. But then the harder thing is providing context on a business, helping take that research and coordinate tasks. Right. And code. And there's a lot more expertise and complicated solutions that go into the next wave of AI. Well, it's the same with, uh, AI in the physical world. Like when you actually go to a factory floor or you go to a 3 PL or a warehouse, you see a lot of things that have been automated for several years. Sorting a box to go left or right on the conveyor belt, closing a package and putting tape on it. These are things that simple robotics, not even autonomous or LLM power, needs to do. But to your point, some of the more complex tasks need an immense amount of data and training data to be able to get good enough to that they're better than a human. I mean 100% success rate where they have run times that are multiple shifts, two to three shifts, something that a human can't. Like a human can't plug in for eight hours a day, like one shift, and just be on for, you know, all those not even lacking a single minute in the eight hour stretch. And so my point is, when you go to a lot of these factory floors, like we were just at somewhere yesterday, but you know, a lot of these boxes, you don't really appreciate how complicated it is to fold some of these boxes. Right? Like some of the boxes you get on your doorstep, the Amazon package is the easiest. But a lot of these DTC boxes have these like 100 types of folds here and there. And you're watching the machine do it and you're thinking to yourself, I couldn't even do that with five fingers unless I was like really, really trained on this. But you, you, you understand the thousands and thousands of hours that are fed into the, to allow a robot to have that exact type of specificity that a human would have after thousands of thousands of hours of doing that. And so like my, my broad point here is it gets really complicated inside of a warehouse. It's not just, you know, a simple sorting something on a conveyor belt. There are a lot of complex human tasks and that's why there's a lot of humans right now in the warehouses. But over time, with immense amount of training data that is only made possible through powerful LLMs, you can actually produce that outcome at scale. And I think that's where I get excited. It's like the complex intricacies of the US supply chain. You're starting to see the power of LLMs. With immense amount of pre training and post training data be able to produce the output that we've seen for a number of years with humans. But now you have somebody that doesn't call in at 8am and say, you know, I'm sick or I'm out or you know, warehouses are famous for having like 50% turnover every single year for, you know, hourly wage workers. And now you can have something that that's, you know, a lot more dependable.

Speaker B: Yeah, I mean it's, it's interesting because, you know, I spent many years in China and right, we had fully automated warehouses and right like you would walk into a million square foot facility and there's like three people there. So I mean truly highly automated. And what was interesting is the, and right as you said, I mean I think 50% turnover might, might even be a low number because these jobs are so first of all there's like injuries. You're walking on concrete, right. It's a very physically taxing job. And so what we saw is right, like some of these folks end up moving into like the knowledge economy and how that started to change organizations. Right. Because when people have like from a contextual perspective understand all of these processes and then they can help. It was like the example you gave in terms of like research versus contextualizing. That I think was excellent because okay, now someone understands all of this. How do we take, right, like going back to raw material sourcing all the way to the consumer and because right now there are so many different solutions talking jibber jabber right to each other. It's, it's still, there's so much, I think lost revenue if you will. And then you mentioned it before, right. Like the reverse logistics piece which I think is going to get worse before it gets better. Um, which we can talk about because of GLP1. And, and what do we think are the, if you will, the, you know, we like I look at it, we've got right the whole kind of gen AI agentic, you know, to AGI at some point. Right. We've got that changing and then we've got this massive thing happening in health care, health and wellness with GLP1. That's, there's a lot, I'm seeing unbelievable change from, from both and right. Of course there's a lot of other things happening with GLP1. The number one thing I see right is people are changing size and we're seeing an uptick in returns. How are you thinking about the impact of GLP1 on either your portfolio companies or on consumer industry at large? And what do you think are maybe some of the near term and long term either impacts or opportunities?

Speaker A: Yeah, I mean everyone saw the, the headlines where you know, GLP1s were starting to have an effect on airlines. Right. And fuel costs were going down because the weight of the airplane was lighter. And that's just something that was really fascinating to read that first time, you know, we think. And then the impact on food and the snack and beverage industry, you know, plummeting sales from, from some of the unhealthy type of treats that you've been used to. Um, we've seen the impact on a few different levels. I think one is on consumer. Right. Things have been trending more health and wellness, sort of a companion type of stack. If you're taking a GLP one, you know, grooms had an incredible rise, you know, in a matter of three years, you know, selling for over a billion dollars. But you're just starting to see this move towards companion snacks for, you know, when somebody is on uh, a GLP one. You need to get your nutrients in you need to get your protein, your macro. And so there has been a move there, you know, David, protein. A lot of these are just like complementary to the rise of GLP1s. I think we're, we start to see something on the health and wellness side. You know, we have a very large healthcare practice at M13 and you know it's been, it's been really interesting to see the uh, that the push towards continuous care. And so we have, you know, GLP1s have come up and now peptides are sort of the next frontier and we're all going to see what happens there. But that has been in response to uh, you know, some of the largest downstream burdens on the US healthcare system is you know, the cost associated with obesity and chronic conditions. Right. It doesn't just show up when, when you're, you're diagnosed, but for years and years to come, insurance is paying out for the costs associated with, you know, some of these, these larger impact chronic conditions. And so if you can stop that earlier on in somebody's life, it offsets a ton of costs later on. And that's why healthcare and insurances is willing to pay for a lot of these, you know, GLP1 treatments because they know they're doing that, you know, cost, uh, analysis internally to understand that it is something they should pay for right now. But we're seeing the move towards, in the healthcare system, the move towards this push towards continuous care. It's not just about reaction when you're going into your doctor's office, but it's the whole lifestyle of prevention. You know, if you get on a GLP1, um, are you going to stay on that forever, uh, when you get off of it, how do you continue that life cycle of preventative care so that you're not always going to the doctor to get treated, but you can go to the doctor to check in on how well you're doing in preventing a chronic condition or something like diabetes or obesity. And so we have seen a push on the consumer brand side. It's kind of fun to see the snack brands that are coming up. But also on the healthcare side this push towards continuous care. And then, you know, the last piece is using AI to really optimize that um, to track different signals from your body.

Speaker B: Well, of course we've always had a foot in health care and a foot in the consumer and I think like you kind of meet in the middle around health and wellness, if you will.

Speaker A: Yeah.

Speaker B: And it's interesting, right, they used to call it, you know, Compliance and now it's adherence and this idea. Right. That the consumer is making the decisions. But I think the biggest difference between compliance, adherence and you know, how however we want to think about that in the past versus Right. With with other drug and other kind of diagnoses is that with, with GLP1 right. It's almost the immediacy. If you don't take it within a week you see a difference. And some of these other drugs that you take that are, that are long term, you don't necessarily see that impact so quickly. And so I, I think that number one we're going to see greater kind of compliance. Number two, what we're seeing in the data is that it's often taken for whether it's a uh, 1C pre diabetic, diabetic, et cetera, but that it's now positively impacting a lot of other diagnoses. And so this idea that maybe over time, right. The consumer is more as you mentioned focusing on whether it's prevention or adherence and then what can they do now because their, their, their mind is free to focus on other things and. Right. And they can, they can be much more proactive in taking care of themselves in, in kind of like how they spend their, their leisure time. And just there's all of these other areas of impact that, that we've seen. Right. Where consumption is down but people as they change size. Right. I think that while yes of course returns are up but I, I personally think. And what we saw in the fourth quarter in apparel was unbelievable and that the physical traffic into retail has. I mean you talk to any mall owner. I was just at icsc. I mean people are just gobsmacked and what's happening in physical real estate and it goes back to. Right. People are trying new brands. You want to try the product on. And so there's, there's all of these uh. I mean there's a lot of really. I mean outside of fmcg. Right. And I think what we're seeing there is right. Going back to. We're seeing new brands. I think we're going to see a lot of acquisitions and we're also seeing new products that are very protein laden that are, that are coming to market. But, but if you think about right this, this fact that we have. And I agree with you, Peptides are next.

Speaker A: Yeah.

Speaker B: Or he in it next in a. Let's just call it. And as it develops into more of. More of like kind of GLP1. Right. Kind of. You're the doctor's office and that's more in their, their kind of, you know, toolbox. Um, how, how do you think this, this change in the consumer psyche combined with how they're spending their, their leisure time, where, where does that kind of, if we look at the end of 27, which I think is going to look very different than where we are right now, how does that impact how you think about investing? Right, because you're investing for the longer term. Yeah, it's, to me it's, it's the most complex but also most interesting it's ever been. So how do you take all that, right, like what which is out in the future and bring it back to today? Because so much of it is, is really way more unknown than what we've seen in the past.

Speaker A: Yeah, our, our internal mantra, you know, we talk to a lot of founders about this too is as an investor and a founder, you sort of have to have a microscope in one eye and a telescope in the other. Right. You have to be incredibly in tune with, with what's happening today and be very analytical but at the same time understand that today is not going to be the same five years from now. And so there are some, you know, key truths that, that we think about what, what's going to last, what's going to change. And you hit on a couple points that, that, that's interesting. I think with, with the healthcare, you know, the adherence is a really important piece and some things that we've found in our healthcare portfolio is if you can really nail a B2B2C approach, it is so much more valuable for, for a consumer acquisition and retention. And you have a lot of D2C brands that sell GLP1s or health and wellness. And it's up to the consumer to maintain adherence to that. And you know, whether they continue to subscribe or churn, you sort of have to continue to market to them and you know, show that, that your product is valuable with versus that B2B2C approach, you're selling to a provider that is then selling on your behalf to the consumer. And if my doctor is prescribing something and sending up follow ups and making sure that I'm trying to adhere to this thing and it's for my health, I'm much more likely not to churn from that. And so we've loved investing over the years in B2B2C approaches, uh, because to your point on adherence, it really ups the level of retention with some of these consumer products. And that's been a very, very good and Efficient acquisition strategy for us. And then the other thing. Yeah, go ahead.

Speaker B: No, I think that's a, that's an excellent point and just one that you know, we should think about as we're kind of like moving forward.

Speaker A: Yeah. And then the other part is on, on physical, physical retail. I think it's, you know, even now as I'm, I spend a lot of time talking to agentic commerce founders, um, and brands that are looking to get much more into agentic commerce. And you know, there's just an article posted the other day that online traffic has now reached a tipping point where agents are accessing webpages, ah, at a higher rate than humans. So more than 50% of website traffic is now being done by agents or bots, which is the first time in history that's happened. And that proportion, it's just going to continue to get more, you know, it'll be 60% in the 70%. Agents will be, will be browsing the Internet a lot more than humans in the near term. Um, and I'm very, very excited about agentic commerce. I'm very excited about in the future, you know, putting my telescope on. I'm going to provision an agent to do purchases on my behalf. I'm going to provision and a uh, shopping agent that knows me very, very intimately that asks my personal shopper to get things for me and you know, I can be kept in the loop but it'll do a lot of it autonomously. I do not believe that that is going to be a hundred percent of purchases. I think there's, you know, I've talked a lot about this in the past, but there's innately human aspects of shopping that we can't discount. And part of it is physical retail. Like at the end of the day humans like to go shopping in physical retail and that is never going to change. I don't see an oasis in five years from now where we're in pods just having things delivered to our doorsteps. No, that, you know, discounts a lot of human attributes like social, wanting to go on a treasure hunt, entertainment, you know, passing a day at the mall. These things are always going to be very, very human. Even online browsing. Like sometimes I actually do want to go to the website and so I don't believe in the death of the website because humans just want to browse. And even after a long week of work sometimes it's a coping mechanism. Right. Scrolling through and doing your online shopping. And so I do believe even that I'm immensely excited about the future of magentic ecommerce. It's only going to increase the value of in person shopping in human directed e commerce shopping. And so I'm very, very long term bullish about the, the human in the loop of, of shopping because it's just such a human behavior to want to shop.

Speaker B: It's interesting. We uh, just hosted a webinar this week with the CPO from Resolve and we did a, a very, very detailed look at the size of the agentic market. And uh, right. If you use the terms whether it's orchestrated, inspired, you know, kind of what that looks like. And so it, it was unbelievable. Just kind of how, how big is big? And we came up with slightly under a trillion dollars by the year 2030 because it goes back to right if, if you're even right like on a website searching and you're like okay, I'd like to think about it this way, right. Like most likely that that kind of assistance. Right Is being done agentically now. And I think that the consumer expects an increasingly frictionless experience which once again is almost kind of by definition agentic. And what I think we're going to start to see is more you know, Bopus, which is why like RFID and whatever is very important because making sure you know what inventory you have that you can actually sell. And so I, I think this idea around a lot of the discovery and a lot of the kind of shopping journey may be done through agents will go into the stores because there's, and a lot of this will be kind of put together for us. But ultimately we want to look at the, the color, the fit, et cetera. And you also want that human aspect of, you know, what do you think? How does this look? It's. And I, I feel like we're in this and we've seen that in grocery, right. I think that's the best example where we've seen there's. I, I do think the estimates were that, I mean My gosh, 10 years ago, 15 years ago it was like 50% of grocery is going to be done online. And even before the pandemic we were in the low single digits right. Now we're in like the low double, like the low teens because people still want to go. And I mean it's, it's interesting. We do a lot, you know, from an ethnographic research perspective, right. We're like when you watch people shop, right. It's, it's even different how they, they explain it themselves, right. Like I think we don't often understand even ourselves how we're shopping, let alone how the consumer is.

Speaker A: Yeah, I mean it, it's a great example. I think online shopping and specifically grocery just spiked during COVID out of necessity. And I think it got as high as it was, like in the high 20% or something like that. But, you know, as excited as I am and as much as I've invested over my career in commerce solutions, it's still only 20% of all of, you know, retail in the United States. So e commerce, roughly 20% is done online and 80% is done in a physical store. And just shows you like, to your point, the power, the staying power of going somewhere, shopping in person. And it's not just about convenience. If it were about convenience, it would all be online. But that's not all that shopping is. We don't just want the most convenient path. That's one aspect of shopping. There's a lot of others that actually in person is, is the ideal way to do it.

Speaker B: I think that's, that's a great point. All right, so we, we get to my favorite point of this podcast, which is a little less, uh, formal in our lightning round. And so we have lots of questions for Brent, as many of you do, as I'm sure as well. So we'll, we'll, we'll, we'll move through these quickly and if you want to pass, just say pass and we'll go on. All right, number one, most underrated consumer

Speaker A: brand today I would say Google. It was kind of left for dead a few years ago. OpenAI anthropic. But the staying power, the distribution, all of the different ass search and YouTube, it just has ultimate staying power and it's still the verb even though, you know there's others that have come around it.

Speaker B: I love that. Number two, most overhyped technology trend.

Speaker A: Oh, uh, man, I might get in trouble saying this, but agentic commerce, kind of going back to the thing we were talking about that, I guess. Oh, let's be more specific. The death of the website.

Speaker B: Yeah, I like that. Number three, one startup category you wish you saw more of.

Speaker A: AI in the physical world. Solutions for supply chain logistics providers. I think it's the next wave. And I want to talk to everyone in the category.

Speaker B: Favorite AI tool you use personally.

Speaker A: It's gotta be anthropic, Claude. Personal use, business use. But everything from scheduling to researching to doing tasks that our analysts were doing 12 months ago, it's doing for us every day.

Speaker B: Uh, one book every entrepreneur should read.

Speaker A: I think it's called Setting the table. By.

Speaker B: Yep, I know exactly.

Speaker A: Danny Meyer. It was the. The biggest takeaway is a lot of people can do things great. You know, a lot of restaurants can, can serve the same type of food. But uh, hospitality, the way somebody feels leaving your restaurant or the conversation with you or the meeting with you goes a long way.

Speaker B: It's. It's funny I have to comment there and just give a shout out to Rick Darling. So when we spun out of Liam Fong, he said you should sleep with your phone because if a client calls, you want to pick it up no matter what time it is, day or night. And he's like, that's one of the most important things. And so I think he. I'm, uh, m. I'm forever grateful he taught me that very early. A, ah company you admire that is not in your portfolio.

Speaker A: Pass. I. I'm only, I'm only talking about my portfolio.

Speaker B: Good. Okay. Biggest mistake founders make when pitching investors,

Speaker A: not spending enough time on them. You know, earlier, earlier stage, you're making a bet on the team and talking about your superpowers doesn't come naturally to everybody because you know, some people don't like to hype themselves up but not spend enough time on the team and themselves.

Speaker B: The next trillion dollar company will be built around what physical economy?

Speaker A: Uh, AI in physical world.

Speaker B: Coffee meeting or zoom meeting?

Speaker A: Coffee every time.

Speaker B: What is the one thing the venture industry gets wrong?

Speaker A: I think even how high valuations are getting. If you look at the best successes in venture, everyone always underestimated the outcome of those successes. And so I think, you know, we're always taught about the downside and that's great. But I think sometimes we fall into is underestimating the potential upside of uh, an investment. And I think that's a mistake a lot of people make. It's not. Not enough risk sometimes.

Speaker B: Agreed. And uh, last kind of lightning round question and then we have a final fill in the blank. The future belongs to companies that, that are AI native. I agree with you. All right, so this is our last question for you and thank you so much. This was so enjoyable. You spend your time looking for what's next. What is one trend technology or shift that every CEO, retailer and brand leader should be paying m. Attention to right now that they are not.

Speaker A: I do, I. I've given a common answer. But I do think LLMs are, are an incredible powerful point as it relates to the physical world in a physical economy now. And so, you know, we've, we've been focusing on a lot of digital applications to date. And the next five years will surely be AI applications in the physical world. Um, and. And that's an area that I'm. I'm extremely excited about.

Speaker B: Great. Thanks so much for joining us. We really appreciate it.

Speaker A: Thank you, Deborah. Thanks, Deborah. And thank you for joining us. If you enjoyed this episode of Retailistic, please check out our full catalog of interviews with retail industry leaders and tune in next Tuesday for another great conversation. Have a wonderful week.