How Nuvocargo Shifted To Being AI-Native With Deepak Chhugani | EM Group Chat #203

Welcome back to Everything Marketplaces, where we talk about founders and leaders from today's top marketplaces. So this is episode 203, which is really good group chat to have with Deepak Chibani. He's the founder and CEO of Nuvo Cargo. Nuvo Cargo is a leading platform for cross-border freight. They've raised over $75 million in capital from top investors like Y Combinator, NFX, and Tiger Global. So this great chat with Deepak went back to the first steps of starting Nuvo Cargo, some of the early challenges Deepak walked us through, how it's evolved and scaled, did a deep dive into how they're leveraging AI today. We discussed topics like taking the full-stack AI-enabled approach versus vertical agent. Deepak shared some of his key fundraising learnings after raising over $75 million in capital. We also had a great group Q&A. I really enjoyed this conversation. You're going to find it a great watch to the end. So Deepak, as I mentioned to you right before we started recording, you know, Nuvo Cargo has come up in the community a few times now. So, you know, really excited to have you join us today for the group chat. I want to start off by saying, you know, huge thanks for taking the time to do so in advance. And, you know, we have a lot that we're going to dive into with your journey, you know, building Anuvo Cargo. But I thought it'd be great if you could start off by briefly sharing your background for those here that might not know you.

Thanks again, Mike, for having me. So my background is my parents are from India. I grew up in Ecuador, and my dad moved halfway around the world to start a logistics business. So even though the goal was for me to never be in the industry, kind of came full circle back into something Somewhat related to what he built. You know, his dream, like many good Indian parents, was we're gonna send you to the U.S. or went to the American school. I see I see there's someone from Ecuador. ¿Qué tal, Martín? Placer. And and and so basically came to the U.S. for my undergrad in 2010. Started my career on Wall Street. I started in M&A and investment banking at B of A Merrill Lynch in their Latin America team in New York. That's how I kind of weaseled my way in. to that industry. Moved to Mexico City because I didn't get the H-1B visa for all those US immigrants. And then I really just wanted to get into startups and I was lucky to have a couple mentors that pushed me and encouraged me even though I didn't have kind of a technology background or experience starting companies. And I kind of took the plunge. I took a little sabbatical trying to figure it out. I don't recommend doing it the way I did, but it kind of worked out. And then, yeah, I started with one idea, ironically, It was a consumer marketplace idea, was called The Lobby, and then pivoted 6 months after YC into Nouveau Cargo. Nouveau Cargo is about 6 and a half years old, coming up on 7 soon. You know, we're Series B, C, I don't know what the letters are these days. We've raised about $80 million, you know, mid-8 figures of revenue, about 100 employees across North America. And then the company's really evolved. We have 3 products, maybe 4. 2 of them are kind of similar through an acquisition we did this year, 3 products. And it's, it's been a whole journey. So happy to share whatever is useful to you guys and hopefully some wisdom and learnings and tons of mistakes that we made along the way so you can avoid them.

Yeah, certainly. That's a great background and journey. So thanks for sharing with us a lot that we're going to get into, you know. But I guess to start with, you know, it sounds like Nuvo Cargo is a little bit more kind of complex business. You know, so could you, I guess, like give us a brief overview of it before we start to dive into things and what do you do?

Yeah. So again, I always assume the audience doesn't know anything about freight. So sorry if it's very basic for those of you who do. The way I describe it to, you know, my grandma or my friend who's never heard of freight and trucking is every product in the room you're sitting in or I'm sitting in goes through a supply chain, whether it's the Apple laptop or, you know, glasses or desks. All of these things go through supply chains. What is a supply chain? You know, it's from the beginning, from where something is manufactured all the way to where it ends up getting shipped to its end customers or to the distributors that are gonna move those products or sell those products. And so we focus on the middle mile of that process, which is when you have to move products or raw materials in what's called pallet freight. So if it's in a truckload or less than truckload anywhere in North America, that's what we help you streamline. The business model that we started with is basically like a tech-enabled freight broker. If you guys know what that is, it's kind of like an intermediary that will go to brands like some of our customers, whether it's Home Depot or Kimberly-Clark or some of these others, and effectively tell them, I will help you move your goods on these lanes at competitive prices, and I will do it with visibility, with compliance. I'll give you the data that you need to make better decisions about your supply chain, about how you're performing. And so that, that, that's how we started. And we were really focused on the US-Mexico trade lane. For those who don't know, it's the world's biggest intercountry trade lane. You know, it's about $50 billion of freight spend a year. It's about $1 trillion worth of merchandise being moved across the US-Mexico border. And we've since expanded to have multiple products. So, and geography. So we do all of North America now. We move freight, you know, if you're a customer moving a few pallets within the US or a few truckloads within the US or Canada-US or, US-Mexico, we can help you with all of that. And then we not only do executing those freight movements, we also do something called customs brokerage, which is kind of like a legal product for import-export. And our, you know, our, let's call it my favorite product and the one where we get to be the stickiest and most kind of reliable partner to our customers is something called managed transportation, where they're effectively trusting NuvoCargo to run their entire freight spend for North America. And we partner with them on achieving outcomes. Like, are we spending this money competitively? Are we achieving service levels? And increasingly that's all powered by our software and AI agents. Whereas for 5 years it was really a mix of software and humans where that ratio of humans versus software and AI has really inverted to the point where it's really, in a lot of loads, it's 95, 97% technology, which is really, really exciting and something I didn't think was possible That's gonna be a really helpful context.

You did mention that, you know, you didn't initially start with the idea of NovoCargo, right? So could you like walk us through, you know, that kind of pivot and then the first steps that you took to build the business?

As a disclaimer, this is a story I've told 100 times. So if you guys want it in more detail, it's available publicly. But the way that I would summarize that story is, you know, I was 24, 25 when I got into YC and the whole wisdom back then when you didn't know about startups was like, You need to do things that solve your own problems. And I actually think that's really good advice the more I've learned. And the only problem I'd really solved for myself at that point was getting this Ivy League job in banking when I didn't come from an Ivy League university, basically, right? And so I reflected on how I did that and I tried to productize the tactics, which were getting coffee chats with, you know, the person at Goldman Sachs who's willing to sit and have a coffee chat with you, tell you what's wrong with your resume, It's almost like productizing what the elite fraternity brothers or sorority sisters would give each other to get good jobs. We tried to productize that to democratize those insights. We did that through a consumer marketplace called The Lobby. So, you know, if you're a junior in college or a second-year MBA student, you would pay $100 and you could speak with someone who's had the job you want and can almost like coach you through, this is how the interview process will probably work. This is how you should think about selling your story. This is a mock interview. And that's how I got into YC. And I would say that, yeah, it was also a marketplace. It was very complex. I was the product. I would find kids on college campus clubs to buy the calls through me. And then I would find friends of mine in banking and Wall Street to take the calls. And then what we ended up doing there was also super manual and then working with a team to automate that completely and learning very non-obvious things. Like the biggest thing there was scheduling software, quite frankly, it was like a little Calendly version we built for rescheduling. But, but that got some traction, got into YC, raised $1.2 million, realized it was a terrible business, and then pivoted to Nouveau Cargo in early 2019.

That's a really cool story. I think lots of points there relatable to lots of founders here. So, and another thing that you mentioned as far as, you know, kind of being more of like a, like a tech-enabled kind of broker, you know, in the very beginning versus kind of like a marketplace. So, you know, could you like maybe share a little bit more about that and kind of like how it's evolved into more of like, it's like an AI kind of powered marketplace today as well?

I'll preface with, you know, just having invested in a lot of companies myself and knowing a lot of founders, I want to try to be specific because I think that it's hard to lump these things together. I do think there's a different, like when people think of marketplaces and look, we're backed by FJ Labs and FX and marketplace investors have put a lot of literature, a lot of content. So we've debated and talked about these things. at length, right? So what I would say is, I think traditionally when people hear the word marketplaces, they associate it with something like eBay or maybe Airbnb where I can go online, there's so many options, I can just pick it, do it, and you're kind of on your merry way. And the marketplace is the platform that enables that, that has network effects where it's scalable. The more you add on both sides of the marketplace, it gets more and more valuable. But then, you know, we started in this era where there was this thing called B2B marketplaces and people would call it, It was a catchy phrase, but it's very different because for a few reasons. The first reason is you're not going to a consumer who's just gonna make a purchasing decision with 5 clicks online. You're dealing with a B2B customer, right? That's the first thing. The second thing is B2B is complex. Everyone has different needs. And so the way you're trying to lure in that customer on the B2B side means that there's gonna be some complexity right at the tail end of like, how do I even sell this value prop? How do I get them to approve the budgets? Internally. So right there, it's already such a different process, even if everything else was the same. And what makes it even more complicated is when people talk about B2B marketplaces, they're basically saying, hey, there's this thing that companies need. There's maybe a fragmented or large supply base and there's this entity called the B2B marketplace that's intermediating and helping them find that supply, right? Sometimes that supply is not super fragmented. I'm trying to think maybe like in insurance, maybe there's 10 carriers, But then there's companies, there's industries like freight and trucking where there's literally over a million trucking companies in the US, right? So it is really fragmented and you could create marketplace dynamics. The reason that you don't or historically haven't is because there's so much complexity and manual work involved in doing the work for the customer. Okay, I'll help you find the truck, negotiate with the trucker, onboard them, make sure they're not fraudulent. Then I have to make sure that they switch on their tracking. Then I have to give you that tracking. Then I have to, you know, the trucker said they'd charge us $1,000, but then they're saying that you made them wait 3 hours. So there's another $50. And so now I have to like redo the whole invoice because like that's not the number we approved. And so, so B2B marketplace is a magical term that I think possibly hides a lot of the complexity and the differences between a traditional consumer marketplace, right? And so I would say that you almost have to, if you want to innovate on processes like those, you almost have to start as a tech-enabled broker, or you see that there's enough of these brokers and you sell tools to them, which people have done in our industry as well. We decided to be more like a tech-enabled broker. And what I told Mike before this session is you're catching me in a moment where it's been like 2 months where I finally feel like we are a marketplace because it's taken us so long to automate every little thing I just told you guys where now I'm literally, I can't show this because there's customer data, but there's a screen and we just have bids coming in. AI agents are negotiating them. Tracking is done automatically. It feels like magic and it finally does feel more like a marketplace. But what we offer to customers hasn't changed. They work with us because we offer a wide selection of Trucking companies they could work with. They rely on us to do the vetting, the negotiating, the procuring. They rely on us for tracking and visibility. And I think what AI has done is just enabled the possibility for these things to get automated where they didn't feel possible a few years ago. So maybe I'll close by saying, maybe it's possible, at least we're feeling it's possible, we're seeing it, that you can get some of those exciting scalability and marketplace dynamics from consumer marketplaces. 10, 15, 20 years ago, maybe in B2B, but it's extremely difficult. It's not been easy. It's very painful to do all this like operating and automating and change management. But today from where I'm sitting, it finally feels real, right?

Yeah, no, certainly. So yeah, I think a lot of times as founders, you know, we're trying to think about like, you know, kind of marketplaces or like more tech-enabled brokerage. And then when it comes to fundraising as well, right? Like getting a lot of kind of input on that. So. Yeah. So you did mention like a lot of the different, I would say like solutions that you offer, right? For customers today. So, you know, how do you think about like kind of sequencing, right? So like how much to kind of take on and then when it comes to actually building out solutions for that?

Well, if you ask my team, probably the biggest feedback I get is that we do too many things and we have to do less. So I don't know if I'm the right person to talk about sequencing. I think founders naturally are very excited about new ideas and new products and trying to like eat the world and solve customer problems. I think we're in a reasonably good balance now where maybe in '21 we were going to do everything for everyone and be a fintech also and whatever. Today, I think we look more like the incumbents of our space 20 years ago and the solutions that we offer are solutions that scaled customers expect from you, right? So for example, they need, they want help moving, finding capacity and moving their shipments. So that's an offering. They might need, they might wanna work with you because they value your technology in a more involved way so that, hey, I spend $10 million a year on truckload spend instead of just giving you 10% of it. I like your technology, I like your visibility. Why don't you just run all that for me? That's our managed transportation offering. That's something scaled players have like Transplace that Uber Freight bought for $2 billion or C.H. Robinson, which is a $20 billion company. These are offerings that customers understand and take time to build up. And then the third offering is customs brokerage. And the reason we have that is because we started in US-Mexico freight. So naturally a way to get visibility and improve the experience at the border is controlling more of that import-export document process. And so we, so we started with freight brokerage, we added on customs because it was natural. And then as our technology got better, we added on this idea like, hey, I think the technology's good enough that You could leverage our technology not only for the loads you move with my brokerage, but with any provider you work with today. And then trying to latch onto a business model that customers understand. That is not without its challenges. Like I would say that people still get confused sometimes, like our sales team, like what's, what's this brokerage thing versus this managed trans thing? And it's, it's not easy to have multiple products. I've spoken about that with a lot of other founders and operators. Like it's the moment you add different products, you add complexity into your business. Like what does the sales rep sell? How do you comp them? What are their goals? How do you train them to sell this? And then how that changes as your product and engineering team ship more new features, especially in the age of AI that's moving so quickly. So the sequencing was trying to be grounded on a mix of what customers need so we can be a deeper partner. And then also business models that we felt were well understood in the market and that would obviously increase our value as a business in that order, really starting with like what the customer needs, right?

Yeah, that's really interesting to hear about. It's definitely a topic that we're always discussing here in the community with founders and teams. So, and I did want to spend some time talking about AI and how you shifted to being AI native, but it's also worth noting that you're around kind of like the pre, you know, ChatGPT kind of era, right? Yeah. And the new kind of capabilities. So could you kind of walk us through, you know, how you've approached AI and kind of shifted to being more AI native?

You know, it's rare to see a big platform shift or a technology shift while you're building a company. And one of the biggest learnings for me is, you know, leaning into the uncertainty and ambiguity as much as possible. So what did that mean for us? First of all, like the cool thing about AI and ChatGPT is it's kind of your own tutor. So even as confused as I was, I would use it to educate me like what's an LLM? What's a data center? don't understand any of these things, quite frankly. I feel like people don't admit that. I had no idea what anyone was talking about and I would just go through, look, let's have conversations. Let's understand. That was one dimension that helped as I started getting a grasp of what's going on and why this is a big deal. Second, I think you want to think about how this could impact your business and your customers. And again, the cool thing about maybe the way we were set up is we're a YC company. I've angel invested a bunch and we've always tried to position ourselves as someone who would help new startups. So one of the weirdest things that was really crucial to us getting smarter about AI faster was any startup that wants to pitch us instead of like, oh, leave me alone, I'm busy. No, like all of you pitch me and I'm going to be on the first call. What the hell do you do? I don't understand. Why is this different? And what you learn about early stage founders, if they're doing it right, is they just have this magical ability to intensely try to explain what they do to you. Even if they're rambling for like an hour the first time, if you're willing to ask the right questions, they will start educating you. And so for me, it was a mix of being that annoying CEO in the company that would just start dropping like, this is a new startup, this is a new article, this is a new thing I read. And everyone in the company, including like people I deeply respect, were like, Oh, this is Silicon Valley's new stupid hype cycle. This is like crypto. This is like VR. And it's almost like this thing where you feel resistance. But then as you start grounding in like real use cases, and that's why the startups were interested. Hey, here's a startup wants to demo how document processing can be better. And we have a team that does 5 hours a day of document processing. And then they're like, oh wow. Oh shit. Like this is different. Right? And that took a while, by the way. Things were not You had ChatGPT. It wasn't like that. You didn't have all these startups that took, I would say, I don't know, 6 to 15 months. I don't even know until it was just like every week there's a new startup that's claiming to do something where we have human staff and where we think we thought it wasn't possible, like voice AI to negotiate with trucking companies or, you know, LLMs to think of different variables for quoting shipments or understanding where the market is. Or things like freight audit and pay. So for me, it was just like, felt like a slow journey. And then I looked back and spoke with friends who hadn't done any of this. And I felt like I was 3 years ahead of them. And all we had done was feel like a complete idiot all the time. I don't understand this term, ask questions, speak with the startups, have them pitch our teams, have your team tell you you're an idiot, you're a hypey CEO who's just reading stuff about AI. And I went through this with crypto and VR and you're wrong. And then keep pushing that information and try to ground yourself on things that impact the business, like make humans more efficient, make your customer value prop better. And then I think at some point when you've seen enough of it, you start drawing patterns of, okay, this is what the technology can largely do and it's getting better. And these are the 5 to 15 areas of the business that I think it's gonna have an impact. Then you can start taking actions like, am I systematically Creating a culture where my team gets it, they're bought in. Do we have this muscle of evaluating these new technologies and using them so everyone can be more efficient? And then you start seeing these little things like the same people that were criticizing, oh, look at Devin AI, it's like a terrible product that makes all these mistakes. Those same people 4 months later were like, oh my God, I used Cursor and I just saved 5 hours of my day and this is insane and I don't need to ever hire anyone again. So it was like finding those moments of people getting it and then it starts permeating in the culture. It was super difficult. And then finally, and I know I'm saying a lot here, the holy grail, I think as a CEO is that you understand this stuff so deeply and you've seen it play out in certain parts of the business. Now you have to sit and think, how does this change my value prop and my business strategy so that I can keep up with the startup who's pitching someone right now to be in my space as the AI native player, right? And I can give you an answer for us. It would be a little long, but long story short, you need to have an opinion on that. And I think the only way you get there is by understanding what people are saying who are at the forefront of the tax. This is what can be done that couldn't be done in the past. And once you get that, you have to modify your business strategy, which is uncomfortable because it might mean everything you're doing is wrong, or it might mean you have some structural advantages that you have to explain and learn how to leverage with your team, your employees, your investors, etc.

Lots of great points you mentioned and also 2 tips for founders here as well. So, so if we kind of dig into a few specific examples, you know, there's a lot of, I would say, probably like a lot of coordination and kind of operational complexities to your business. So could you maybe share some examples of like how you specifically leverage like maybe like agents that would previously kind of be like more, I would say, like, you know, maybe like labor intensive to solve a problem?

Think about it this way. When you're, when you're doing a tech-enabled brokerage or tech-enabled services or whatever, There's a transaction, a core transaction end to end. And if you've spent years like we did thinking about how technology transforms it, we already have this muscle of where can software and technology make each part, the life of a load in trucking, life of a shipment, how can it be better? How can it be more automated? How can we— that's from the experience, like what can be automated? And then also you think about metrics like the cost to serve. Like, hey, I could change this industry if I had a lower cost to serve. than the incumbents, right? So basically in the life of a load, there's 7 or 8 areas where we had concluded in 2023 or 2022, look, this is as good as software gets. And then the humans are just going to do this part. And I just have examples where now AI agents are doing most of the work. So that's the framing. I'll give you a specific example to answer Mike. Let's say that you need to go out and find, actually, I think I can't even share this one. So I'll do a simpler one. Let's say that you get a tender from your customer. That's like one of my customers. They say, look, here's a tender. I've selected you to move this shipment from Detroit to Miami and we've agreed upon a price. And so we had already pre-negotiated that. So here's a tender and it's an email with a PDF with a bunch of information that normally a human would have to go and type in to our platform. And by the way, a lot of brokers don't even have a platform, so they'll just like manually do a bunch of stuff. To create the shipment. So in our case, you had to still, even though we have a bunch of software we've built in our platform, Nuvo OS, you had to manually create that shipment, those facilities, those lanes. And then that open shipment would go in something called the load board where people would start procuring and calling truckers. So not to talk about the rest of the process, but just that initial thing, it wasn't really so easy to think, oh, you can just add an agent to an email. It will parse out the PDFs and the LLMs and it'll automatically be able to structure what the tender should be and create that shipment. And right there, you know, if you're doing hundreds of shipments a day like us, you've eliminated hours of someone's day, right? So that's one example. I have like 9 or 12 across the life of a load and they really compound over time. And I think that the cool thing about how we built it is because we have this system of record, we really create very deterministic and clear locks. So if an AI agent doesn't do its job well, it cannot move on to the next step. Whereas I think a lot of these pure AI agent startups are running into the problem of, oh, I'll do everything with AI agents. But if your agent gets it 5% wrong in step 1 and 5% wrong in step 3, by step 7, it's like 60% accurate. And that just doesn't work in B2B, right? Or really in anything, right? And so that's how we're excited about the system we built. And then using AI agents in a very controlled and thoughtful way is a way to get that automation. And then you have your humans troubleshooting when the AI agent hallucinates or gets something wrong or needs further training.

That's definitely a great example. So thanks for sharing with us on that. I'm sure we could have an entire kind of group chat just going through all the different examples and kind of solutions, right?

So yeah, totally.

If we kind of like zoom out from like a high level, you know, what are some of the biggest challenges, you know, operating in the cross-border freight space?

Well, the first thing I'll tell you is that we have to do a good job of explaining this. We're no longer just in cross-border. We're doing all of North America. But in general, cross-border is very complex because you're dealing with regulation, you're dealing with import-export, you're dealing with different currencies, different cultures. So yeah, a lot of the complexity of our business boils down to that. You know, we had to build teams almost from day one and in the pandemic in the US and Mexico, different cultures, different employment law, different invoicing processes, different, yeah, just different cultures even within freight and logistics, right? And then in general, logistics is one of those crazy industries where things change all the time. Like there's a war in Ukraine. You wouldn't think war in Ukraine hurts your B2B SaaS business, but for us, war in Ukraine means fuel prices went up and some craziness happened with supply and demand or interest rates go down and then the stimulus is crazy. So then demand goes up and truckers are buying trucks like crazy. It floods the market with supply. So there's endless stories about how logistics cross-border and domestic is an extremely dynamic market. And, but that's also why we add value. We can help customers navigate those crazy changes that happen really all the time, right?

Yeah, certainly. I'm sure it's a challenge just, you know, trying to stay up to speed on everything as well.

So yeah, it's difficult. It's difficult. It's challenging. We've had, you know, there's a port strike, there's a government shutdown, there's You know, there's, I don't know, some crime issue happening on the US-Mexico border. There's just so many things that happen all the time. But, you know, 6 and a half years in, you develop a muscle and you realize that all those things can also be opportunities. Yeah, even just today, we're talking, there's some regulatory change with Mexican customs and we have to react to that. And you almost build a culture that's very dynamic and reacting to all those things. But it does create more complexity than a more simple and static business, right?

Certainly. I wanna make sure we have time for the group Q&A, so we'll get to that here in a second. But, you know, right before we do, I did wanna ask, about the fundraising journey, because I believe you've raised over $75 million in capital from some top investors, right? And you mentioned a few, so what has the fundraising journey been like? What are some of your key fundraising learnings?

I think that all young founders go through like this immaturity of wanting investors to validate them instead of using them, instead of using capital for what it is, which is a source of fuel for the company to get to the next stage. And that most founders won't admit it, but it leads you to make pretty dumb decisions. Like I'm going to optimize for this fund's thing versus that one, or I'm going to take more money when I don't need it. I'm giving up more of my company. I think we were quite fortunate in that we hit the ground running very aggressively and then it was '21, '22. So we were able to raise capital at very attractive terms, kept control of the business and with people who were really excited to work with me. And so I've seen a lot of my friends from '22 and '23 who raised a lot of money just deal with a lot of drama from working with people they shouldn't have been working with at their board and in the investor level. So a lot of my takeaways are, you know, being very thoughtful about why you're raising the capital, who you're raising it from. And then once you know you need it, there is a whole, almost like a game or an approach to fundraising that really pisses people off. But when I'm privately helping them, It's all counterintuitive.

Yeah.

Like it's not about the perfection of your financial model. It might be about getting 3 people to make a bid and then you get the best price and the best terms. And a lot of people hate that because it's like FOMO and dating and games and all that stuff. But it really does matter when you're operating in a private asset class where the main emotion that investors are operating under is fear. It's fear of missing out. the next big thing that will be 1,000x. They're not worried about your business. I miss out on an opportunity to 2x my money safely and conservatively. That's not venture capital. You can get that with private equity. And that's why I said it really matters to understand the kind of capital, why you need it and be really clearheaded. But I think my unique take is a lot of people don't realize how much of that fundraising journey is really just people going through their own therapeutic version of like validating themselves and Outsourcing that to investors with spreadsheets, which is so stupid if we all think about it, right?

Definitely. That's going to be helpful tips here. So cool. So we're going to get to the group Q&A. Hey, Michael, saw you raise your hand. You want to help us kick things off?

Thank you so much, Deepak. This is fascinating to hear, especially with AI, what you're tackling with such complex logistics. So my question is in that area. So you're now adding AI agents to things that were previously human handled, probably. Probably. How have the customers adopted this? Any learnings in that area?

Do you—

are they— are these things visible to the customer or are they more behind the scenes? Just anything in that area that you can share would be great.

Great question. So the short answer is no, it's not been visible to the customer until now, like until right now. So It's important to separate what I'm talking about now versus the past. We literally had a moment where we decided to deprecate all customer-facing software because we realized this is a hard thing. Like, oh, it's sexy. I saw on Twitter they have this dashboard. Like, you have to think in the nuts and bolts, like, what are you offering? I deliver outcomes for the customer, finding the cheapest truck, onboarding and vetting, negotiating, making sure all the data and visibility for tracking is there. So we realize if we don't get our internal system right, to deliver on these outcomes, nothing matters because we would have customers logging into a dashboard with incorrect data. So a lot of our company focused on the inward software and making the internal humans more efficient. And then once we got the data to a threshold where, hey, this is really good, this is better than the industry, now let's start exposing that to the customer. And so you're catching us in a moment where we just hit some kind of like exciting inflection point of automation and AI working, and now we feel comfortable to expose that to the customer so they can see something no other broker in the market can show them, which is like every agent, every negotiation, every email thread, everything is available. And we're going to be launching that sometime next year. But I can't stress enough, none of it would have mattered if we wouldn't have gotten it to deliver the outcomes the customer paid us for internally first. None of it matters, at least in our model. Maybe in another model it doesn't matter if they're just hiring you for AI, right?

That's a great question. I'm gonna get to Lisa and then Margaret. So I know you both sent me questions in advance.

I'd love to get some insights into your go-to-market on that side of things because from what I understand and in my experience, it's very similar to what we are dealing with. Don't have systems analog resistant almost to technologies in some cases. So I just love some insights on the go-to-market there.

Yeah. It, it, it's, again, it's a super hard thing to answer concisely. One thing I'll say that confuses a lot of people is We have the brokerage model, but we don't sell to brokers. Brokers are our competitors, right? So there's 2 entities here. There's the shipper, which is the company that needs their freight moved. And then you have the broker who is the intermediary between them and the trucking companies. And then obviously you have the trucking companies, right? So we, we have acted as the broker historically, and only now with our technology are we in a place where we show up to the customer and we can tell them, hey, maybe you don't even need brokers if you use our technology. We'll give you all this muscle. Available in software. Now, to answer your question, the shipper, they, some of them use technology, some don't. Like, just like with anything, there's a spectrum. The small and medium customers, there's something called the TMS, a transportation management system. There's some parallel to this in every industry, an ERP, loan management system and lending, whatever. There's always like some ERP someone is using for their business. In freight, the shipper who's most sophisticated and advanced is using something called a TMS. Where they get to see all their shipments because they've decided they have enough freight that it's worth investing in that SaaS product. And then they start adding other tools and integrations. So I would almost put it on a spectrum where you have the company that's young and is doing everything manually with emails. And there's still a lot of that, which is still shocking to me in our industry. And then you have people at the very, very high end who've adopted the biggest CMS. They've added 9 integrations. They'll show us their diagrams. And so I have to calibrate the pitch on the go-to-market depending on where they are in that spectrum. And that's a muscle that just takes time. You have to understand what else is out there. You have to understand like, does your product actually replace those softwares? It's really difficult to summarize today, but I would say that depending on where you add the most value, maybe trying to be ruthless about filtering customers on that journey, if that matters, or you just start adding value. Like if a customer doesn't wanna adopt my software, they still need somebody to move their shipments. So we would move the shipments and start the relationship that way. And then hopefully I can get them one day to see that if they just ran all the loads through our software, their life would be better and they could get rid of 10 other SaaS tools. But it's a, it's a journey and it's really difficult to convince someone to make that big of a change if they've never heard of you and never worked with you and you're a small startup, right?

Thanks. That's a great question. Hey, Margaret, do you want to jump on here?

Thanks for doing this. I was wondering how you organized your go-to-market and sales over time, especially like with limited resources in the beginning. It struck me you have a lot of similarities to my marketplace, which is a semiconductor marketplace, that you've got an axis of geography, the buy, the side of the marketplace, and then also like this concept of getting them into the marketplace versus long-term support. How did you decide what axes to divide your go-to-market on?

At the beginning it was very simple. And, you know, the more time I do this, simplicity is underrated. It's hard to keep. So at the beginning, it wasn't the sexiest, but it was like, let me hire people from other brokers and let's sell brokerage. How do you get someone to give you loads and then learn about that process? Have them set up meetings, speak with customers. Some founders don't like this if it's not in their personality, but even till today with this new kind of AI, agentic AI end-to-end product, I'm doing all the sales. I have 9 customer meetings this week because I can't count on my reps to understand everything the team built and then learn and then get the customer feedback. You know what I mean? It's a tall order and that's one of the prices you pay for a complex business or a world where things are moving so quickly. So I think a lot of it should be founder-led sales. And if your network is not able to get you a lot of calls every day, then you might have to hire 1 or 2 sales reps from the industry. And their job is just to get you on the phone with customers and you have to get good at asking the probing questions and figuring out clear next steps and really simplifying How can you start working with me? How can I explain what we do in 3 to 5 slides so you understand the value I add and why you should care? And yeah, if you're doing a marketplace, presumably there's a clear need you're solving by helping them fulfill that need or find those products like in the semiconductor space. So who's the stakeholder? You know, again, you just create a go-to-market motion. It's either a senior person who can build it or you build it and you find reps that will go with you. But you just have to get a lot of at-bats and you hopefully start codifying it somewhere. I sell to this person at this type of company. This is what I ask them in the first few minutes of the call, depending on what they say. These are the value props I talk about for my product and my offering. And then if it resonates, this is what I do with next steps to start working with them. And that's like your V1 of a go-to-market motion. And then you can spend years optimizing each side, like getting 1,000 leads or learning how to sell your product better or improving your demo motion or Hiring better salespeople, comping them differently. Like it becomes endlessly complex. But when you start founder-led sales, a few reps and just try to codify who you're selling to, why they give a shit, how you make it easy for them to understand, and then try to have next steps and clarity so that it drives growth. And the last thing I'll say is I think one thing that YC did for me early on that was great, just like pick like one KPI that is a true measure of improving your business. And I think for marketplaces, that's especially true because you can go crazy trying to please both sides. In our case, it was gross revenue or bookings or whatever it is in marketplaces and just try to ruthlessly grow that number week over week, month over month. And in the pursuit of growing that number, you should learn about the blockers. Like my product's not good enough. I don't get enough meetings or I don't have something on the supply side. And that just becomes your go-to-market, but also kind of the roadmap for what the business needs to do to keep growing, right?

Awesome. That's going to be real helpful. I know, Deepak, I know the conversations can start to get very nuanced very quickly, right? So we talk about marketplaces. So, and also to a related topic, and I know we were, you know, kind of mentioning this prior to the chat, is around, you know, taking the full-stack kind of approach versus the more kind of like vertical agent. And you were mentioning, you know, discussing some of the pros and cons. So can we, you know, spend a little bit of time discussing that?

Basically what I told Mike is, again, part of my job I feel like is just to understand what's going on. And the cool thing about Nuvo today is we get pitched by all these AI startups. I invest in some of them, and then we're building and training agents ourselves. So we kind of are seeing the whole gamut now, whereas a year ago or 18 months ago, I was kind of confused as to what the hell is going on and how things are going to shape up. I don't feel that way anymore. And so I can see, I can say that there's a few things that I see happening in my space. I'm not speaking for yours, but I assume similar dynamics would ensue. There's a whole class of AI startups that are selling agents to companies like mine, the freight brokers, right? Hey, I'll help you negotiate with your truckers better. Hey, I'll help you process your documents better. I'll help you build agents for automation and workflow. And that's one approach. The other approach is full stack, which is what we do or what, I don't know, companies that are like, hey, I'm a law firm. I'm a law firm. I'm not a software company, but I'm powered by AI and you're gonna see it everywhere. I'm a freight broker or whatever. And so I think there's a lot of pros and cons. I think one thing I am noticing because I'm getting pitched by so many is 2 years ago there was one voice AI company selling to freight brokers. Now there's like 15 and they're all, and now they're offering me a year for free and they're all driving their prices down and I don't know if they make money. And so I don't know, that's kind of crazy. I really do see that some of them as somewhat of a commodity, somewhat, not completely. They're better at prompting or they're better at go-to-market. But the technology edge goes away if someone, because markets are efficient, you find something where it's easy to get a lot of customers, they're willing to pay, and then 10 people get funded to do it, right? So I think that there's something powerful about that. The simplicity of I sell an agent, I get a bunch of customers, I grow super fast, and then my business strategy allows me to expand from there. But if you can't do that, I would be worried about those companies in terms of their structural advantages or moats. In our case, we're trying to take what we've built. We tell, we're trying to tell our shippers, you don't need to learn how to train AI agents or monitor them. We'll just do the work and we'll show you what AI did. And there's some benefit to that approach. We'll see how it goes for us. I'm very bullish on that because the customer doesn't have to think about it so much. Because if you think we are all playing catch up on AI as technology founders, just imagine people in a lot of these industries that are just trying to like move their freight or buy their semiconductors or whatever, you know what I mean? So I'm bullish on that approach, but the risk of that approach is when you sell, you still sound a lot like the old school version unless you're very good at explaining what you do differently. So a lot of it becomes explanation and demos and creating value props that are maybe a game changer so that you don't fall into that commoditized slow grind of a sales process, right? Yeah.

No, definitely lots of great points there. So I think, Abigail, I think you actually had a question somewhat related to this, which is around, you know, competition and defensibility in AI as well.

Deepak, great conversation. Thank you. I'm building a marketplace in excess inventory and logistics, moving it into nonprofits. And so my question was just that, how do you build your moat and how do you stay defensible against competition?

I think every business has a different answer. My answer is not super crisp for a podcast, but again, speaking with a friend, I would just say, hey, I'm seeing that the legacy players, the big logistics players, they're really struggling to deploy these technologies effectively because they have to like learn how to use AI. Then they have to layer it onto their ERPs and TMSs, which they honestly don't, most of the time they don't own, or it's a 30-year-old system. And then they have to do change management and then they would have to fire a lot of people. And it's just really difficult to do at scale in a slow-moving incumbent. That's one. And then the AI agent startups, as I just told you, I'm seeing them pitch me and they don't want to be, most of them don't want to be full stack. If they want to be full stack, they'd compete with me in my years of like learning the edge cases and building it into the TMS and learning where the AI agents work and study the cost structure, or it's pure AI startups that to me are becoming somewhat of a commodity for how we use them. And I've heard they also struggle with a lot of our competitors because they're much less tech savvy. Like even these companies raising $100 million, getting deployed and doing these big splashy press releases, you find out in the background, oh, they tried to switch them on for this one workflow and they have to be shut off because it's really hard to make this stuff work. So what I'm realizing is a moat for a company like ours is like all these cycles of executing this and then making the technology work. To actually automate, it sounds easy. It's actually so so freaking difficult. That's some version of my answer, right? I'm sure there could be more, right?

Awesome.

Godfrey, do you want to come on? We'll try to get in on one last question.

At the beginning, when you started Novocago, did you try to hire people from the brokerage industry, or did you go learn it yourself? I hired people from the brokerage industry, and I've had. If you interviewed me every 6 months for the first 3 years in the company, I had a different answer to it. I want people from the industry. I never want to deal with people from the industry. I want people from the— it's a terrible, unfortunate thing that like you just have to get good at both until at one point you have a team that's homegrown that has been with you for 3 years that gets what you're trying to innovate and also understands what they need to understand from the industry. But early on I needed people because I sounded like a complete idiot on these sales calls. I'm like, I'm innovating in trucking and they start talking about, Lumpers and FTL and all this jargon. I'm like, what the fuck are you talking about? So you needed a salesperson to help you not sound like an idiot, basically.

Hey, Deepak, actually related to that, do you have a few tips maybe for founders going from 0 to 1 and today, and then when it comes to eventually kind of team building and hiring?

Look, on team building, very generic trope that is not hard advice to give, but super hard to follow. Your first 3 to 5 people, they should be people that understand they're signing up for an insane amount of ambiguity and uncertainty. The moment you're negotiating with someone about, you're spending your time negotiating their comp, so it feels like they're at Google and you're explaining to them why everything's going to go smoothly. Like that's the wrong person for an early stage startup. It's just going to be painful. It's going to be expensive for you. It's going to be annoying for them. You're going to hate them because like, I'm just doing my best as an early stage pre-product market fit founder and this person wants Google. No, your conclusion should have been, They should go to Google. I figured it out on the second interview. And I want someone who's excited about the zero-to-one phase where they know they're signing up for uncertainty and ambiguity and they're generalists and they're just excited to figure it out with you, right? I think that's the first one. The second one honestly is the first 3 to 5 people will dictate your culture to an insanely uncomfortable degree. Like I'm still fixing things in my culture that I cannot get rid of in year 6 and a half. that come from those early hires where I screwed up. Like it's just, and everyone is going to screw it up. No one has 100% hit rate. Maybe the most actionable advice, there's this article called Gene Pool Engineering from Vinod Khosla. I don't know if you guys have seen it, where it basically says, take the biggest 5 risks to your company succeeding. Each of the risks, find 3 companies that have solved it effectively. And then for each of those companies, write down like 5 names. I don't know the math, but basically you get like a list of 75 people that you can try to hire. If you don't hire them, you get referrals, but hiring should be people that can help solve the biggest risks in the company and then understand that early stage. And I think that can create a really powerful early team. Yeah. And then early on, it's always a chicken and egg. I don't have the money to hire them, but I don't have the traction, but I need them for the money. Like, unfortunately, too bad. Like, that's the magic. That's why the founders will make crazy amounts of upside because you're figuring this shit out that seems impossible. Like, how can I do X without Y? And I need Y for X. And like, that's entrepreneurship, right?

Certainly. Got a lot of head nods for that. So those are going to be great tips. Also too, great way to wrap things up. So I know we're out of time here, but yeah, once again, I really appreciate you taking time to join us here today. It was great to hear about all of your experience building Novo Cargo and some of your learnings. And I appreciate helping to answer some of the questions that we had here during the group Q&A. And actually, one last question. This is my typical kind of parting question, which is that if you could go right back to before you started Anova Cargo, you know, what would you tell yourself about marketplaces?

About marketplaces, I would say they're very difficult, basically. They are harder than B2B SaaS. Yeah. And I would say that you need to be extremely thoughtful about what value you're adding across 2 stakeholders, which is not easy. that's gonna drive growth and differentiation and is gonna create a flywheel that's really exciting. And that's really not obvious for the most part, or it's a little muddy 'cause you know what clicks for one side and not the other. Yeah. But in general that they're just, they're hard businesses, but they can be very rewarding when they work, right?

Yeah. That's gonna be good parting advice for us all here. Now, last but not least, time for a quick plug. Where can we keep up with the app?

Yeah. So if you guys wanna sign up to the Substack, I think you might find it interesting. Like we're talking about stuff we're building. Which we share stuff on LinkedIn and Twitter with our company as well. But that's basically it. Yeah, LinkedIn, Twitter, and Substack, I guess.

Awesome. I'll include links in the description of this. So you have lots of marketplace founders coming your way. So once again, really appreciate it. Thanks for taking the time to join us. And thanks everyone for the great questions.