The Freight Pod
The Freight Pod is a deep dive into the journeys of the transportation and logistics industry’s brightest minds and innovators. The show is hosted by Andrew Silver, former founder and CEO of MoLo Solutions, one of the fastest-growing freight brokerages in the industry. His guests will be CEOs, founders, executives, and leaders from some of the most successful freight brokerages, trucking companies, manufacturers, and technology companies that support this great industry. Andrew will interview his guests with a focus on their life and how they got to where they are today, unlocking the key ingredients that helped them develop into the leaders they are now. He will also bring to light the fascinating stories that helped mold and shape his experiences.
The Freight Pod
Ep. #45: Jonathan Salama, CEO of Transfix
Andrew is joined by Jonathan Salama, the CEO and co-founder of Transfix, a logistics technology company. He discusses Transfix's origins, starting as a freight brokerage business that he and his co-founder Drew built from the ground up. They focused on automating key functions like pricing, scheduling, and carrier management using machine learning models.
When the pandemic hit, Transfix's technology proved invaluable in navigating the volatility. However, Transfix later had to pivot away from being a full-service broker when plans for a SPAC deal fell through. Salama then took over as CEO and refocused the company on its pricing automation technology, which he believes is highly differentiated and can be licensed to other brokers and shippers.
Salama emphasizes the importance of building a strong team and doing right by people as key lessons from his 10 years in the freight industry. He remains optimistic about Transfix's future, believing the company can continue to make a positive impact by helping carriers and the broader supply chain through innovative technology solutions.
***Episode brought to you by Rapido Solutions Group. I had the pleasure of working with Danny Frisco and Roberto Icaza at Coyote, as well as being a client of theirs more recently at MoLo. Their team does a great job supplying nearshore talent to brokers, carriers, and technology providers to handle any role necessary, be it customer or carrier support, back office, or tech services.***
Follow The Freight Pod and host Andrew Silver on LinkedIn.
*** This episode is brought to you by Rapido Solutions Group. I had the pleasure of working with Danny Frisco and Roberto Icaza at Coyote, as well as being a client of theirs more recently at MoLo. Their team does a great job supplying nearshore talent to brokers, carriers, and technology providers to handle any role necessary, be it customer or carrier support, back office, or tech services. Visit gorapido.com to learn more. ***
A special thanks to our additional sponsors:
- Cargado – Cargado is the first platform that connects logistics companies and trucking companies that move freight into and out of Mexico. Visit cargado.com to learn more.
- Greenscreens.ai – Greenscreens.ai is the AI-powered pricing and market intelligence tool transforming how freight brokers price freight. Visit greenscreens.ai/freightpod today!
- Metafora – Metafora is a technology consulting firm that has delivered value for over a decade to brokers, shippers, carriers, private equity firms, and freight tech companies. Check them out at metafora.net. ***
Hey listeners, before we get started today, I want to give a quick shout out and word to our sponsor, our very first sponsor, rapido Solutions Group, danny Frisco and Roberto Acasa, two longtime friends of mine, guys I've known for 10 plus years, the CEO and COO respectively, and co-founders of Rapido Solutions Group. These guys know what they're doing. I'm excited to be partnering with them to give you a little glimpse into their business. Rapido connects logistics and supply chain organizations in North America with the best near-shore talent to scale efficiently, operate on par with US-based teams and deliver superior customer service. These guys work with businesses from all sides of the industry 3PLs, carriers, logistics, software companies wherever it may be. They'll build out a team and support whatever roles you need, whether it's customer or carrier, sales support, back office or tech services. These guys know logistics. They know people. It's what sets them apart in this industry. They're driven by an inside knowledge of how to recruit, hire and train within the industry and a passion to build better solutions for success. In the current marketing conditions, where everyone is trying to be more efficient, do more with less. Near shoring is the latest and greatest tactic that companies are deploying to do so, and Rapido is a tremendous solution for you. So check them out at gorapidocom and thank you again for being a sponsor to our show, a great partner. We look forward to working with you and to our listeners. That's it. Let's get the show on the road.
Speaker 1:Welcome back to another episode of the Freak Pod. I'm your host, andrew Silver. I am joined today by an awesome guest, someone who I've gotten to know just a little bit. We spent an hour meeting each other at Soho House in Chicago. We had a drink and I got to learn a little bit about your business. That was earlier this summer, was that Jonathan? Yeah, it was already a few months.
Speaker 2:Time flies.
Speaker 1:Yeah, well, I think you had just done your deal with nfi at that point, and um, a lot of interesting stuff. I mean, you have such an interesting business with a long history in the space and, uh, so I'm excited to dive into that. Today my guest is jonathan salama, so I see your last name right. Yeah, yeah, nailed it. French origins Is that true? Yep, born and raised, yep, awesome. We're going to dive into that a little bit. So, uh, he's the CEO, uh, co-founder of transfix, was previously CTO and made the move to CEO, which is another topic we'll dive into. Um, but let's start with your origins. So you know, sure, a frenchie. What was it like growing up in paris, france? Did you? Did you have plans to one day make the move to the states and, uh, become a freight broker, or you just kind of stumbled on that journey?
Speaker 2:no, yeah, like right in high school in france, I thought, yeah, let's go freight broker in the us, it's to work. No, the story is I make harsh and quick decisions and I had the opportunity to study at University of Miami in Florida out of high school. Why not, let's go Moved? Studied there for computer engineering. But I thought, you know like, I'll do a quick internship in the US and then I'll move back to France. And I had one at Microsoft that like didn't work out for me.
Speaker 2:I moved to Seattle from Miami so I went from sunny every day to rainy every day. I just didn't like it there. So I moved to New York. Three months in, I'll find something else. I'll work for one year, I'll move back to Paris and I'll live my life and I'll be happy ever after. And two or three months on the job I met. I worked with this lady that is now my wife and she is from wisconsin. There were no convincing her to move to france with me. Uh, so that was that. I have been here ever since. Um never looked back. But that was, that was a deal. Was another startup in New York, tons of fun.
Speaker 1:And now I'm here Harsh and fast decisions. I'm the same way. When I was 25, I was asked if I'd be interested in moving to Denver because Coyote wanted to open an office in Denver. And I said yes while in the room, because why not? And quitting quitting Coyote to want to start Molo was also a decision I made in less than 24 hours and I had spent 10 years working there. And I'm curious how, if you, if we were to zoom out a little bit and look at our lives and look at, you know, running a business, in some cases I could see harsh and fast decisions being a benefit and in others being a challenge. Can you think of kind of an example where it's been one or the other where a harsh and fast decision has been productive or been a challenge?
Speaker 2:There's so many, but I think there's this. I think you hear a lot from people like oh you have to gather all the data. I made data driven decision and sometimes there's just they're just not there and you've got to make a call. And if you spend one week to make that decision, it's a lot worse than than not, than, in my mind, making the wrong decision. Um, I I love this concept from amazon, um, the one-way door versus a two-way door. I don't know that?
Speaker 1:Explain that to me.
Speaker 2:The concept is so if it's a one-way door decision, you can't roll it back. Simple example.
Speaker 1:Quitting your job.
Speaker 2:Quitting your job, right, you can't roll it back, but I don't know, it's starting to interview at other jobs. You don't have to like bring it back right. So there's this, the concept of one way and two ways, and if it's two way door, I go really, really fast towards that decision. If it's a one way door, it depends sometimes, when you know, you know, I proposed to my wife three months in to her relationship. Really, yeah, I just that's quick. When you know, you know I I proposed to my wife three months in to her relationship. Um, really, yeah, I just that's quick, you know.
Speaker 2:When you know, you know, um, I like I mean she's got, if you ask anybody, she got the short end of the stick. Um, so, oh, uh, you know it is what it is. But um, yeah, um, but yeah, I it. So that I think that I think that's when it's helpful, it's when you can be unlocked and move quickly. I think it could also bite you. But I can't think of an example where I regretted a decision because I made it too fast. Yeah, I can't think of one, okay.
Speaker 1:Well, I like that notion, as I thought about the two-way door, just in terms of trying new things. Like I don't know, I'm in a place in my life where I'm experimenting a lot and I used to be afraid to try new things. I don't know why, I don't know if it's that kind of imposter syndrome or just thinking you won't be good, or thinking I wouldn't be good, or whatever it may be, but I've reached the point in the last six to 12 months where it's like I just want to keep trying new things and testing things. Even something like chat GPT was something that I was like leery of messing around with at first, and now now, all of a sudden, I'm using it all the time, like even this morning I was, I was in the gym and, uh, I was talking to my trainer about my diet, which I have not gotten to a good place yet.
Speaker 1:Um, and he was like wait, you're eating all. He's like what are you eating in the morning? And I said, well, he you know four eggs with cottage cheese mixed in with it sometimes, and, uh, yogurt with protein and a protein shake. And he's like that's way too much protein at one time. Your body can't process it that quickly and I was like, yeah, see, this thing, I don't know what to do, I don't know how to do it. He's like just go on GPT, explain your weight, your height, what's going on, and it'll tell you everything.
Speaker 1:And I did, and all of a sudden I have this plan. So I don't know, I just leaning in and just if, if, if it's, I like the idea that, in a two-way door where there's no risk to testing something out, just do it and see what that, um, what comes out of that. So, speaking of just doing it, uh, you're an entrepreneur, uh, and the CEO of of transfix. So take me back to the origin of transfix and like, how did the idea come to be, um, and, and how did you get yourself involved in that?
Speaker 2:Yeah, and that we'll. We'll roll back to the subject of like size decision, cause that was also a really, really fat decision, but, um, so, as it should be, yeah, I was, um I. So I was at this startup in Nework where I met my wife and we both decided we wanted to try out the san francisco silicon valley um area. So we moved, we moved to san francisco and I joined another startup and that startup didn't work. It didn't work at like. I was there for one year basically the only engineers, or there were a couple others there um and a year in the company shut down, and one thing that I don't blame anybody for that shutdown. But one thing that I was frustrated about is that I, as an engineer, I had no voice. Um, it's a company of four, it was a company of four or five people we're not talking about like hundreds of people, right, and like, okay, fine, I'm one of 50 engineers and I had no voice, and so I came out of it.
Speaker 2:I want the next thing I do is I start my own startup. I don't know what it is, I don't know what I'm going to do, but that's the next thing I do, and this is one of the reasons my wife is a wonderful person. She had to support me while I was doing this is I stopped making, um, I stopped having a salary. I just went in and started studying different, different startup ideas, different things, but I was, I was coming up blanks. I didn't have anything, um, and I was pitching this friend that was all these crazy it was, they were. He was teaching me a lot. There was like some ideas are a great idea, but like, what's the time here? If you succeed and capture a hundred percent of the market, maybe a million bucks, like what do you have? Uh, so starting to learn, like these are the things I'm looking for. I'm looking for a massive TEM, I'm looking for an expert in the field who can catch me up, because I know engineering, but that's about it. And at least I had the time and I want an undisrupted space.
Speaker 2:So I had these three criteria built up and one day one of my friends said listen, you're, you've got, you've got this uh concept and all this stuff, but you don't have an idea. And now you have this friend who's the opposite. He doesn't know how to get it started, but, um, he has this idea for a trucking company. You should listen to him. So, uh, drew drove to Brooklyn. That's where we were. He drove like from I don't know somewhere in New Jersey. Um, we met at a bar. We had a few beers and that night it was like great, so we're starting this, it's happening. We didn't leave the bar that. We shaked in like 50, 50, let's go. Um, we've never met before. So, um, but I, I, I fell in love with the concept, uh, which, back then, it was just you've got this massive, massive, massive space that everyone depends on, absolutely Everybody. Everybody depends on my.
Speaker 2:That first startup I keep mentioning to was a was a fashion company where we're selling handbags and side goes down. It's very hard to feel passionate about. Oh, no, how? No, we're not going to be able to sell those handbags. Um, like, it's very hard. Um, but for um, in this space, a truck driver doesn't get his shipments, a shipper gets his delivery late, like that. There's real world impact to getting this stuff wrong. Um, it's like it fell in love with the space, fell in love with the concept of this. Back, like, we started. I, I'm gonna go on record that we were the first one, uh, to do this right. Um, it was 2013. Um, and and all I thought back then like wait a minute, like if we can track the trucks with their phone, that's already tech that doesn't exist in this space, like this is this is great, let's go Um how did you, how did you learn about this?
Speaker 1:Like, how did you know that? How did you know that? How did you know that people didn't have an ability to track on the phones? I assume you knew nothing. Yeah, and I think what was Drew's background as well?
Speaker 2:So Drew is from the industry, drew's father. So, if I get the story right, the Drew's father had a trucking company that he pivoted into a brokerage back in the 80s, the late 80s. So Drew grew up with his father company having a brokerage and his story is different, in which he went to school, he went to Georgetown, did his four year and came back and thought wait, a minute, somebody is going to disrupt the brokerage space. I need to help my family invest into the business and essentially pivot it to not be just full truckload brokerage. Because he was worried somebody was going to disrupt the space and put them out of business. So he pivoted his company's business and then by the time he was done, realized wait, nobody's doing doing this, so I'm gonna go do it. And that's when we met um and he said to you yeah, cut.
Speaker 1:So how did he describe to you the problem he wanted to solve? Because you know you, you were going to be the guy who was going to actually build the solution. But what was the? The actual problem?
Speaker 2:The thesis back then was the back office at a brokerage, and this hasn't changed that much. But the thesis back then was the back office of a brokerage takes a considerable amount of money out of the transaction, meaning that, like you're, you're scrapping to get your 20 margin or 15 back. That it was like the math was different than today, obviously. But you're, you're scratching to get like your 15, 20 margin and then, once you get your margin, 80 goes away away from your cost. And there has to be a way to automate this.
Speaker 2:And we dig deeper and we talked about scheduling and we talked about this concept to just accept a shipment. Just, there are so many rules from shippers sending shipments to broker and you have the ability to accept them or reject them based on very specific rule the shipper has, and accepting the right freight and and rejecting the wrong freight has consequences. So people take their time and do it and like wait, like we can build rule engine for this, we can do this, we can like, there are ways to do it. Um, and so so we did all right, all these solutions were I don't want to call it obvious, um, but they were. They were pieces of technology that already existed and just had to be put together to offer this solution.
Speaker 1:Yeah, so that makes a lot of sense. And when did you first like how did you go to market? So it was you and Drew. Did you guys raise money first on the idea and then, you know, go hire a bunch of people? No, was it bootstrap?
Speaker 2:Neither. So it was bootstrap for about a year. Um, our our thesis was uh, what this is trucking? It's gonna be hard to get like investors to care about this until we show traction. We, like you, didn't have a product, we didn't have everything. So our, our split in responsibilities was easy drew you go, sell you're out, I'll build everything.
Speaker 1:Um so, and you were when you were building your, your thought was to build the automation functions, or you were building a tms, you were building. What were you building at first?
Speaker 2:yeah, I was. I was going to build like the I didn't even know it was called a tms at the time but like I was going to build my, like my, my version of what I need to operate the freight. And then, drew, before we got started, before anything, the very next week closed Barnes Noble, who was like great, no, we get you, we have no tech, we get you the first one, you get everything. We're going to send you 10 loads a week, figure it out. So while coding, I was actually covering freight as well, which gave me one a great, uh, really great deal of learning on, like how to work with truck drivers, how to how to book a load, how to schedule loads. I had all this like crash course of how to do things while building the platform um, which is why they could like it probably has like some like design um, in the early days there were some of these some design flaws, because they're like oh wait, that's how the carrier reacts, so they must all react this way, so build it that way.
Speaker 1:And then, oh no, they don't Um yeah, because that's the interesting thing, is you kind of simplified, how obvious it seems to automate as much of this as possible, and I think the an overarching lesson that a lot of companies have learned that have been down the automation path is there's a lot you can automate, but there's a lot that you need a human backstop to support in order to ensure fluidity and ensure execution, and it has frankly led to, in my opinion, the demise of a number of companies.
Speaker 1:But I will say it is. You know, it's an underrated, valuable concept that founders and CTOs and engineers and anyone who's coming into these businesses should spend a little bit of time talking to drivers themselves, talking to customers, understanding those different skills. That was something that Coyote did extremely. That was something I give my dad a lot of props for, is he used to make every single person Bill Drieger had to book loads like, yeah, everybody had to book loads, and by learning those things you were just so much more capable of understanding the whole business and what you can do I agree, we did the same, uh, if you didn't matter what you did at transfix, like you book load for a day, um, like we did it at first, we did on your first day and realized, wait, wait, that's too aggressive, you need to understand the business.
Speaker 2:So we did in three months, in 24 hours, you're booking freight, um, and that. That. That is a great experience. But I think, uh, you to go back on something you said, um, I, when I started this, I thought, yeah, this is going to be easy to automate. But I was very naive. This is probably the hardest space there is to automate. It took us over a decade to be where we are today on our level of automations, and I agree. I'll give you a quick anecdote. Our internal tool was called Stark after Tony Stark and the reason for that Was that you.
Speaker 2:Are you a big Tony Stark fan? I'm weirdly obsessed with Marvel and all things Marvel from before the movie, so the movies have been amazing to me, but we called it Tony Stark. I appreciate you sharing that. Yeah, I am. Why not Jarvis? I'll tell you. So it was Tony Stark because we weren't building Jarvis that was going to replace you. We were building, we were the builder of the suit and we needed somebody in the suit. Were building, we were the builder of the suit and we need to see somebody in the suit.
Speaker 2:The goal was never it has never been will completely replace human. Because we've always thought there's one thing that human and I don't care about, like, even given ai, I still believe in this statement humans are great at relationship and there's nothing that's going to replace this and this business is driven by relationship. And so we always thought we'll build you the suit so you can, so the suit can do a hundred scheduling appointments when you like could only do one without it. But you're going to be responsible for that relationship, so that you call the responsible at the warehouse and he hears your voice and he starts knowing you. And when there's the, I only have one spot left at 8 pm. Hey, can I get it right? Robots aren't going to do that and they're not going to get that, but automation will make it so that you're the first one there to ask um, and that that's why we called it start so walking them through the first couple years of the business like what did?
Speaker 1:what did that look like for the organization? Was it like super fast growth and hiring a lot of people like a traditional broker? I did yeah, um, at what points did you start? To what points did the technology start to work in ways like how many? Just give me a frame of reference for just a visualization of what the business looked like over the first few years before you guys raised some cash yeah, so we took exactly yeah, we took exactly one year to raise, to raise capital.
Speaker 2:Um, so for the first year it was me and drew, it was just bonds little ball and it was just really the bill and 10 loads like really, like wait, I'm not ready yet. Like this is not going to scale. I can barely survive by myself doing 10 loads a week. While doing this, I was never a good carrier manager, to be very clear, drew had to pick up the phone a lot to make some of those carrier calls pick up the phone a lot to some of those carrier calls. Um so, so anyway, so we grew and then, um one year in, we grabbed, uh, we got the first fundraising and we we hired a few people. We hired somebody who was going to take over carrier managers, we hired an engineer and we hired a designer.
Speaker 2:From there, the first year felt like very high growth, mostly because our customers were very, very large enterprise, which, in hindsight, this is great that we got this done. But why would you give your freight to a company that has so few people and us? I I mean, I'm grateful on the dream, yeah, I'm super grateful because it got us here, but so it felt crazy rapid growth because every account was a really large amount of freight. Um, and so then we we started the machine of hiring carrier manager, carrying hiring operations. We've always had the model of not the what's it called, the model where you do everything versus it's split.
Speaker 1:So a split model versus the model of Cradle to Grave.
Speaker 2:Thank, you yeah, Cradle to Grave. I was losing my words for a minute. So we've always had the split model.
Speaker 1:You had the split model, so you had people who were focused on the carriers and you had people focused on the customers.
Speaker 2:Yeah, they were mostly because of the way the automation was built. It was built in a way like, all right, you're doing scheduling and we're going to automate, like we're going to make your day so efficient, and we thought like, if you're critical to Grave, shirt, um, and we thought, like you're critical to grave, you have to now understand 20 different tools. And that was that. That never clicked in our head. Um, obviously we've evolved and we'll talk about what we're doing now. That can fit everything but um, at least while we were a broker, it was always a split model, and then we grew, and for 10 years we added customers, we grew, we go to business, um, really, I mean to a large brokerage business in the us. That was something we're really proud of, um.
Speaker 2:But to answer your question of of when the technology kicked in, I think it was really when COVID hit that we saw how well our tech worked, better than what we could have hoped for. Because what happened? Covid hit and we got worried. We don't know where the capital is going to come from. We're still burning money. Maybe all shippers are going to stop shipping, who knows. And so we had to take the really hard decision to do a round of layoff to try to survive this phase like this, unknown what is going to happen, and it turns out the opposite happened and we had so much freight, we had more freight than we know what to do with, and the first thing seems like, oh, we're gonna have to rehire.
Speaker 2:And then the tech actually kicked in. They kicked in, you know, in ways that wait, we have actually, we still have bandwidth and we're doing way more volume. Um, and that that really was the first um. No, tech is tech is working. Um, especially, it was interesting back then, cause where it worked really really well was around pricing, automation, the everything that we didn't spot that back then was already fully automated and the tech started realizing what you can increase rate, you'll increase volume, you'll increase everything. These are all the bids we're going to make and everything started working really incredibly in ways that was really really automated and really fun to watch back then of the operation were you able to automate Like when we're looking back at those once it was kicked in?
Speaker 1:you mentioned pricing automation. What other features? And I do want to ask you a little bit more about pricing automation, but before we get into that, what were all the features or some of the features that you were able to automate and found that it was efficient and worked well?
Speaker 2:And if you think of the life cycle of a shipment, I like to think about it that way that it goes in a flow and it started with pricing and we talked about it. We had spot was um I think 98 or 99 fully automated um for um and contract was above 70 automated on rfps um and then so from pricing then you go into tendering, like accepting or rejecting the tenders. That was fully automated. Um scheduling was at 67 70 automation. There's always that. We've always thought.
Speaker 1:There's always that phone call you need to make at some point um and with respect to scheduling, was it because there's different functions within or different I don't know if you call them verticals but different types of scheduling so I think that there's.
Speaker 1:You have email scheduling where you're just emailing back and forth with someone at warehouse. You have phone scheduling where you've got to call and and talk to someone about the appointment you want. You've got portal, um where, whether it's, you know, e2 open and you're going into their portal and doing it, or it's retallix, um retail link, um. What am I missing? Anything else?
Speaker 2:um, yeah, there's one um philip is one um uh open open.
Speaker 1:Open doc. How do you? Forget that which is a portal right that qualifies as a portal. So which of those were you able to automate? I assume the phone call one you couldn't automate.
Speaker 2:Yet yeah, this was pre. Uh, all the ai stuff that we have today and always thought I will need the relationship. So the, the, the, the remainder was this every other ones was automated. Email was automated. Um, there was a lot that we send email, send them links, click here If you accept this. This like we had a first tool that would tell us what is the optimal appointment time, based on the shipments criteria, to get the best negotiation from a carrier. We learned very early on the right appointments makes it for an easy negotiation, the wrong one really hurts. So we had tools deciding okay, this is the optimal time and then just email the facility. Hey, click on this link to accept our appointment. But we had email and all the portals. We love the portals. That was the easiest way. We had all of those automated whenever a shipper had those enabled.
Speaker 1:Okay, and with respect to so scheduling, I cut you off for scheduling. Keep going.
Speaker 2:So you get scheduling Once you schedule covering the shipments, so that for that we had either a routing guide, we had a way for carriers to come, we had our own internal load boards, we had ways to automatically email carriers. We had a lot of different ways. Our thesis, whether it's for shippers or carriers, was always we need to meet you where you are, not where we want you to be. And so if you're a carrier ABC, you want to be emailed, we'll email you. You want to be called, we'll call you. You want to call us? We'll send you an email so that you can call us. It was all set up to meet you where you are.
Speaker 1:And the internal load board would be similar to. Was there like a transfix app that any driver could could download and if they in any available load you had was on there and they could just click? I assume there was like a buy it now or a book it now feature? Yeah, they were on it. How did that work?
Speaker 2:so there was a book in now and then there's a button where you could negotiate it and then you would end up negotiating with a bot. This is really cool, but it wasn't by text message, it wasn't. You enter the number, you click the button, you see the we're thinking about this rate, and then we get back to you. It's like we counter at this rate Do you like it or you want to counter more? And then it would basically be an infinite loop until carrier accepts or denies.
Speaker 1:And how did your bot know whether it was a good rate to accept or not? It's a great question.
Speaker 2:So it was all done with actual machine learning. Machine learning is a subset of AI. It's not looking at like, ooh, this is our target. This is part of a lot of what our offer. Like, ooh, this is our target. This is, um, this part of like a lot of like what we're our offer is is our offering today is our models were extremely good at understanding lead time, and so I can.
Speaker 2:You're a broker. You absolutely know If I tell you pick up in one hour, the price is not the same. If tell you pick up tomorrow, regardless of delay and regardless of anything. But what nobody is ever able to tell is like at what point is the inflection point? At what point is it starting to get more expensive, and at what point and by how much? It's a little bit of a gambling dilemma my next hand is going to be better, like my next hand is going to be better, and the bot would know wait, wait. The next hand is not going to be better, starting to get, it's starting, we're starting to get too close. That just, and so it would adjust the rates Like so, if we knew we, if we knew we wanted to hit a thousand and that's what we thought the lane was going for and that's how we bought.
Speaker 2:Our buying power would allow this lane to be at a thousand and truck driver would coming, or or dispatcher would coming and say 1100. The bot would understand why. With plenty of time I can get other offers. I can figure this out. Let's go 900 or it's a you know what. I don't have time. I, if I don't pick this load now, I'm gonna end up with way worse. How about 1100? And even though it would actually bid above our target, it would always minimize the risk of us losing money how do you know it's minimizing it like if.
Speaker 1:What do you have to compare it against? If you, there's not an alternative world where it didn't offer 1100?
Speaker 2:right. Lots of baby testing, lots of games, lots of is it looking at?
Speaker 1:is it looking at historical lanes and saying, okay, six months ago we, you know, at at 1 pm we offered a thousand and we should have offered 1300, because at 3 pm we paid 1600, or something like that? And then it's doing the math on so many of those situations that it has gotten to. Is that how it works, or are there also other sources being pulled in to learn from?
Speaker 2:There's a couple of things. So you're right on the conceptually that's how the machine learns it. But then live, it wasn't open to every single shipment. Lots of shipment was still covered by our team. So it was learning live. It had that information live. Wait, like this broker just booked the truck on that same lane I just had for more money because it's too late, and so it constantly ingests data. It's constantly thinking it's not thinking it's too late and so it constantly ingests data. It's constantly thinking it's not thinking it's a machine, but it's constantly doing the work. And then we get to analyze like wait, it pulled that decision. Our rep did better, our rep was able to get better. The machine was wrong, let's tweak it. And we had 10 years of doing tweaking on this, on this uh algorithm. And it powers the way we price. It powers the way we pull rate. It's powered the way we negotiate with carriers.
Speaker 1:It was powering everything okay, so we got pricing, we got execution with carriers Keep going.
Speaker 2:So then we had compliance. Frankly, I'll give credit to the highway team here. We used to do our own thing. They showed up and were like this is great. So we used highway Highway's great Shout out to highway. It's great, Free sponsor there, Free. Yeah, you're welcome guys.
Speaker 2:I should start charging them Carrier's in sending rate confirmation, doing all the paperwork that I think today is fully automated. Back then it wasn't. We had to convince Carrier. It was okay to sign on a phone or sign on a website, but now it's like a stable stake and track and trace along the way. Interestingly enough, we've always had more success with our app than integrating with ZLDs. We had all the ZLDs that you can think of, but we found that our app ended up being uh, more accurate and give us more data than than those elds, but it was fine either way I feel like tracking trace is almost a given today is there alignment between?
Speaker 1:so if I look at convoy's app and the driver kind of profile that they have publicly talked about and I look at ubers and and some of of profile that they have publicly talked about, and I look at Ubers and some of the profile that I know they've talked about with their broker access, is it fair to compare that to a similar where you guys very heavy with owner operators and small carriers then as well?
Speaker 2:We I mean small carrier, yes, but not owner. It's not that we would say no to an operator, they can work with us. It's not that we would say no to an operator, they can't work with us. But we found that our bread and our better ICP, if you wish, for carriers were between the 5 and 20 trucks we were really. Our thesis was we want to make sure, we want to reuse your capacity, we want your book of business to be 90% Transmex. And because we're giving you round trips, because we're giving you a late or not exactly round trip, but like trying girls or like squares, you tell us like, hey, I want to be back in Chicago in two weeks. Cool, take your truck for two weeks. That that, that's what really was trying. And it's really really really hard to do with owner operators Close to close to impossible.
Speaker 1:But we, we had yeah.
Speaker 1:Well, I'm sorry, I just it feels like it would be easier or maybe I'm missing something that you can clarify. But the when I'm imagining an owner operator, he's got his phone, he wants to load, he opens the transfix app or transfers app or the combo app or the Uber app and he books it and then it's really easy for you to track him because it's it's in the phone and on his phone, if it's. If I'm a dispatcher and I'm booking the loads for my 10 drivers, how do you make sure that they're getting tracked on their phone?
Speaker 2:So technically technically, uh, yeah, that's a little bit easier on their phone. So technically technically, uh, yeah, that's a little bit easier. What's much harder is getting the right shipper that has the right freight concentration to give you those those trips. To begin with. And if we have, I don't know, a hundred shipment going to laredo and 20 shipment picking up in laredo that day, right, you can't. It has to be the right location at the right time. But it's like you're, you're, you're have a dart and you're shooting at a map at the U S and you have it foot.
Speaker 2:It has to be at this like pinging the map at this point in the map, at this time on the map at scale then becomes really hard. You have to. You have to grow your shipper base really um, aggressively. If you want to do that with owner operator versus the 5 to 20 truck, have dispatchers they have.
Speaker 2:Okay, I know I'm gonna have a truck empty here. I know you don't have shipments, so I'm gonna, I'm gonna find myself a shipment and, look, I found one. I'm now going to be in california in two days. Right, okay, we can work with that um. So it was in the tech that was difficult to scale at this point it was making sure we had all that demand um, and we found that it was much easier to do that with um the 5 to 20 trucks companies so something that I don't think any broker that I know of has has been able to manage super efficiently is getting the right freight from the right shippers to create the optimal network and and why that is is partially because every shipper acts independently and frankly, I don't want to say they don't care about the network you're trying to create.
Speaker 1:But especially for brokers, it's not exactly like they ask you for a wishlist and plan to give it to you.
Speaker 1:I've definitely had shippers who care.
Speaker 1:I'm not trying to say they don't care, I just it's hard to make every broker happy, especially when they want to generally feed their assets first, and most brokers you know there's certainly strengths that you develop because of the relationships you have.
Speaker 1:But there are also just lanes that are easier to move because there's more capacity on them and most brokers are going to put those. So if I'm looking at an entire nation of network I'm sorry, entire nation of lanes that make up a network and it's all drive and loads and my options are to prioritize the loads that ship out of Florida going to Chicago or the loads that ship from Chicago to Florida. 99% of brokers are going to pick the loads that ship from Florida going to Chicago because they're just easier to move and anybody can find capacity to move them. So if we all made that our wish list, shippers don't necessarily comply. But I'm curious, were you able to find a way to leverage technology to more effectively bring these networks together, to say, okay, if we pair this with this and that with that, and then get shippers to give it to you, and then get shippers to give it to you.
Speaker 2:So we have a. I want to say, yes, like we did, it worked great. But practice, if I'm being honest, it required, like some, like shipper partners. They came with us, they looked at it. We had a few example of this, but it wasn't. We had a couple triangles that were really cool, but I don't want to give any location.
Speaker 1:I appreciate the honesty. I didn't yeah, I didn't expect you to say you had a perfect answer there. I just was curious if there was something. Because I it's so hard, because it's just each. Each shipper cares about their own network and I'm sure they want their partners to have an efficient network. But it's hard to the tricky part the some shipper.
Speaker 2:It worked really well. The other one is like what? You're telling me that if this facility of this other guy takes too long, you're going to be late. Now, right, like I want you. You're like I'm using a broker because I want you to have backups, because I want you to be able to grab another truck if you need to, so you can be on time. And it's that's what breaks. That's what breaks the perfect routes. It's time, it's efficiency at those warehouses and I mean it's trucking. Things happen, there's an accident, there's a flat tire, there's whatever, and then that makes that trust. You're using a broker to make sure it gets picked up, and that's why it's that's always been that trust you're using a broker to make sure it gets picked up.
Speaker 1:That's always been our. This is 2020, when all this technology is starting to click. At this point, the company is seven years old. I think by now you've done a few raises. I think you did 22 million. I think I saw saw was a series b and 16 um. There seems to be a connection well, I don't know if connection is the right word, but a theme with a few of the companies that have been bucketed in the digital freight broker camp and you guys were bucketed in that camp for sure um where these companies built a lot of automation and technology for their organizations and it was really good technology that definitely was better than what you know the traditional broker was operating with. But at the same time, the trade-off seemed to be that a lot of these companies struggled to build profitable, sustainable brokerages that could keep going. Was that a part of the problem along the way?
Speaker 2:So it's hard for me to tell what's their problem. I wish no one's harm and I honestly wish I feel I wish everybody succeeded. At the end of the day, I'm much more prone to uh encourage people. Like people are trying new stuff, so like might as well go for it, uh that then then bringing them down for trying right, but so I can talk about, like what, what happened with us? And yeah, no, that's what?
Speaker 1:what I was asking? Sorry, I wasn't asking about others. I was just asking if that was kind of a challenge you dealt with.
Speaker 2:Yeah, and so for us, specifically, what we thought was we want to build if you just grab the broker, the broker should be profitable. The brokerage If you remove the tech, the brokerage should be profitable. The brokerage if you remove the tech, the brokerage should be profitable. That that was what we like, we, we, we got there. Is that I forgot?
Speaker 1:what year is that? Because you won't always have to pay, you won't always be carrying that much tech cost, because you're not always having to keep building at such a it was too full of depth yeah, it was too full as a feast.
Speaker 2:The first one is like I don't think, uh, for the, the tech we built over the year was really really expensive in term of like the engineering time that went into it. And thinking that you're going to be able to um to um to make the whole company profitable using the brokerage margin is always something that was like we. We thought we could make it, make it happen. If we really get there to a billion dollar revenue like they were, they were a path, but we knew it was always going to be hard. We wanted the brokerage to be successful. The other side is it's still it's, it's a business, right. And if we couldn't make the brokerage to be successful the other side is it's a business right. And if we couldn't make the brokerage profitable as an entity by itself, then why is the tech good for? If we can negotiate the right shipper rate, if you can negotiate the right carrier rate, what is the tech good for if you can't make money out of the brokerage Plus? We've always thought internally at least, we thought we delivered a little bit of a different experience to shippers and to carriers. Right, like it wasn't the same sale. So we have a differentiated sale. We have a differentiated product. We hope you like the product, you think it's better, but we can't make money out of it. That would be bad business. Um, so so that's really like the the.
Speaker 2:That was core focus since day one. Like we're not, we're not we're. When we, the few times where we had negative margin shipment, we it was all hand on deck. Why is it happening? What are we doing? We doing, why do we make it? To not have that happen again. It wasn't on purpose. It wasn't to grab wallet shares. We did it a couple times to get in with the customers. Sometimes I feel like in this space, you just have to.
Speaker 1:Yeah, we did that a few times To clarify. It's not that you're like trying to lose money, but you want to make sure you get your foot in the door so that you can prove your capabilities and your intent is to make money on the account, but sometimes the only way in is to give an attractive enough price to get an opportunity yeah, so, so.
Speaker 2:So that's what. That's why we really really went for right. We wanted that. We wanted to build a business. We wanted to prove that the tech was working, and the only way to do that for us was to have a successful brokerage by itself, and it puts us into a great position when, fast forward, we talked to NFI, it put us into a really interesting conversation.
Speaker 1:So at what point did you become the CEO?
Speaker 2:You started as the CTO and Drew at some point left the business and was replaced by Lily, and then yeah, he stayed into the business, but he went more into a sales role and Lily came in as our COO and he was CEO and then she moved into CEO, he moved into heading sales and so we tried to SPAC in between all of this.
Speaker 1:Oh I forgot about that. I didn't know that. Yeah, that was fun. The timing was really bad, right, awful timing. So the thing is it takes quite a bit to get ready. It blew up.
Speaker 2:Yeah, sorry, go ahead.
Speaker 1:I was asking a question Can you just explain how the SPACs work, because we've not talked about that once on the show, so I think it would be helpful to yeah.
Speaker 2:I'm no expert, uh, but in then Lehman's term, a public company, um, that is essentially just a shell, uh, that that has just nothing but money in it, acquires a private company. Now that private company becomes public, um, in Lehman, that that's a little bit how it is and it's a it's a faster way to get to to be a public company while raising capital at the same time. And, um, it was like the big. There's a big wave, um, during that time, at 21, 22, there's a big wave of, um, a lot of company going through SPAC. We look at a lot of numbers. Lots of bankers told us that was the right time we should go. It takes some time, it's not instant. You don't say yes and two weeks later now you're public. So we did all the prep work. We did all the work and then market had turned very, very fast and we had to make a decision go out or stay. When was this? I?
Speaker 2:want to say, early 22, early 21? I'm losing track of time. I'm very bad with dates.
Speaker 1:I think 22 was the year that everyone made a ton of money yes, that must have been dead.
Speaker 2:23 is when it um, it's easy because I'm going it was 22 um, we announced that we were not going public october uh 11, 2022.
Speaker 1:I mean google yeah, I think it had started to turn because typically brokers make the most money as the market's on its way down.
Speaker 1:Yeah, right so 2022, the market had started coming down. If you were a good broker that had a lot of contractual freight and had taken good carrier customers in the previous years, when the market was doing the other stuff going up, then you probably kept your contract rates at the elevated level they were and the rates to the carrier were going down every day or every month, which meant there was a lot more profit going into into the broker's hands at that point.
Speaker 2:um, but the the market share was was dissipating rather quickly yeah, and that's exactly what happened, and therefore we had to make a really, really tough decision to not go public, and today, in hindsight, it's the best, one of the best decisions we've ever made. We would have gone crushed um, and we would not have been able to do a lot of the stuff that we did, but can I ask what was the?
Speaker 1:what was the interest in going public versus, you know, raising traditionally with private equity?
Speaker 2:I did so it was. It's a. It's a really nice way to get liquidity into the company, into giving money back to our shareholders, giving um and and more, and it was. It was supposed to be a really uh nice exit for everyone, um, but that was the.
Speaker 1:It was really a financial decision, got it so market goes the other direction and you guys realize can't spat. Yeah, so how do you? What's the pivot plan?
Speaker 2:yeah, so when you got became ceo yeah, six months after that or something like that. So, uh, lily, uh decided to to step down um, just because she decided to step down, um, I I supported her decision although I was sad, but supported her decision and um, we talked about it as a board and I just decided to put my hat in the ring on them. I would like the position lily supported it. Um, we did a nice transition, or like um taught me a bunch of stuff and that was it when I was I was ceo. Drew was also extremely supportive um, and he obviously he's still in the business.
Speaker 1:Um, I mean, yeah, that had to be really hard. Um, if I just put myself in the mental state of everybody, because I know of other companies that were trying to sell at the peak and, um, nobody knew it was the peak, they just happened to be near a deal. And I know, I know a company that nearly got a deal done for three billion and like it all fell apart overnight and all these people had these expectations that they were getting paid, and then no, and and no idea when. So so you're kind of taking over as ceo. You're now going to be the face. You're now, you know, the man accountable for the business, and you probably have a team of people who are looking at you like, hey, man, you know, okay, the spac's gone, the money that we were expecting to go on business is going down. Because it's going down for everybody. What was that like for you to step into the CEO role for your first time? Uh, in these conditions, I'll give you what.
Speaker 2:Even better, my very first week is when SVB crashed and we had all our money there. I was like, wait, are you going to say that, like my first payroll, I'm going to miss my first payroll? Like, is that, is that a joke? Um, so it would find like everything, everything was fine. Um, we, we try, we worked it out, um with with people. But, um, yeah, uh, trauma fire, but it's harsh, like I. I think it's just, it's moving fast as it's a business. I know, right, I it wasn't like I was going into a new job, new business, it's I. I I had well, I hope it's weird to say it out loud but I had the confidence of our internal team. I had the support of the board. I have the support of our investors, um and um, one one of them actually, like the news is like we're not. We're gonna give you at least one year so you get some time to like mess things up before it gets better. Like, oh, that's nice what?
Speaker 1:what was your message to the team coming into that role? How big was the team at the time? I want to say 250 people or 300, something like that. How much of that was probably like engineers versus operations.
Speaker 2:Engineers must have been at the time. Give or take 100, operation give or take 100, and then sales GNA and all that stuff. Another 100 give or take.
Speaker 1:So 250 people now under your watch. What was your message to them? As okay, the SPAC didn't work, we're not doing the SPAC. What's our plan moving forward? What was your message and how did you navigate it?
Speaker 2:Lily had the beauty and the gift of giving the message that we're not doing spec um, but that. So that was actually really quite annoying um, because we weren't quite allowed to tell our team until we posted it in online, because it's all public matters and you have to be careful what you say, so that that was awful. That was like that was at making the decision to not say it to people until you can put like do the proper press release and all that. That was. That was the worst time possibly ever.
Speaker 1:Um it, it felt just shitty I hated, hated, when you had to keep things like that from your team because we sold to a public company, so we couldn't. I was telling people that we were selling but I was like I can't tell you to whom and how, but like we'll let you know as soon as we can tell you it's the worst.
Speaker 2:Um, it's absolutely. And they asked, like what's going on? And that was the worst. Um, so, so anyway. So I didn't have to say we're not spacking all this. So I took over like about six months later, give or take, and and, uh, that the team was overly welcoming, um, so I that I just basically said let's fucking go and. And we went right, that's, that's I.
Speaker 1:I when you and yeah, well, I was just to say at this point were you already planning to pivot, or were we still going to be a really great broker that was driven by the best automated technology in the industry and we could get to that billion dollar number and like full steam ahead?
Speaker 2:yeah, it was a little bit of a mixed message, because the thesis was whoever is going to survive this crisis is going to get all the freight when you're, when you're on the other side. Whoever is going to survive this crisis is going to get all the freight when you're, when you're on the other side. You just have to survive. So it wasn't like let's fully go. It's like, hey, how are we surviving? Let's, let's think through that. That means like. It means a different mindset, different approach, like we get to like, be like very, very cash conscious of, like where we spend, very, very conscious what freight we get when we get it, how we get it A lot more of a data-driven and conscious approach than a sell and grow and do all this. And I like looking at data. I'm okay at it. So it was a lot of doing that, which I liked a lot. Um, so I was um, it was surprisingly not, um, as daunting as one would think.
Speaker 1:Yeah, so the message was less like grow, grow, grow is more like let's batten down the hatches and let's keep the house in order. Yeah, and let's, let's, let's with, with, with, let's survive the storm yeah, that was the message.
Speaker 2:And and where we can, where we can be more gritty, let's go there. Where we can like be smarter, let's go there. Um, that, that was the message, and we did that for about six months before I got it. Looks like we're pitting so, but at the time I didn't know that was going to happen.
Speaker 1:Um and what, what drove? Well, for one, nobody realized it was going to be the storm of the century that was going to last. I mean, arguably, maybe it's still ongoing, depends who you ask. Um, but I am curious um, when, when was it? When you, when did you decide to pivot, and what drove that decision?
Speaker 2:So it was really an organic conversation with NFI. We were talking to the board. What are we doing? We need to raise more money if we're gonna keep, if we're gonna keep going with the same strategy, and what? What? What's the point? Where are we going? We had this um or like long meeting. I remember sometime, uh, over the summer. I left, sorry, before the summer, like april, um, and we're just thinking like, what are we doing? Do we want to get more money? And just literally, the goal of this round is stay steady, don't grow, do nothing, just stay steady so we can see it on the other side. Or do we do something different? And our board was extremely supportive. We keep doing. They knew our tech really well and they're just like it's too bad, the best asset we have. We can't really sell um and they keep pushing us like can we sell it? Can we figure out how we sell it? So we start doing a little bit of like compliance tool and try to go sell it to broker and I didn't work.
Speaker 1:So when you said. So. When you say sell it, you don't mean like get rid of it, you mean commercialize it yeah, license it okay, and brokers won't buy tech from brokers.
Speaker 2:I don't want to open the pandarus box and talk about uber, uh, but but my eye strong thesis from back then from today, broker is not going to give another broker their data. They shouldn't, they're not going to and they're not going to buy tech from competition. I think that's crazy.
Speaker 1:You're like I'm not going to open the Pandora's box, but let me make all the arguments against why that doesn't work. We don't have to open the box.
Speaker 2:Anyway, I'm sure we can have a whole episode on that.
Speaker 1:I agree with you. I agree with you. But it's interesting because you know you look at your brokerage and you probably have great people and feel good about your customers, but it's not that differentiated from other brokerages. And then you look at tech that you've built, that you know is is strong, and there's a lot of automation features that you don't think other brokerages have, and you're like well, this is the differentiator, how do we get this to market? And the answer is we have to get rid of the brokerage.
Speaker 2:Yeah, and frankly I didn't quite see it In my mind. It's like maybe we sell the whole company and they can decide what to do. Maybe we I wasn't sure, um and I had the dinner with um, um, uh, a set up dinner was.
Speaker 1:Dave Jeffersonian dinner, as Dave called it.
Speaker 2:Exactly that dinner, um and Dave was talking about how, like he has this, this view of how TMS are all wrong and what they should be doing. That's what we do. We do. That that's how we do it. And literally after the dinner we hung out for five minutes, talked and we just sort of came to the conclusion that you should buy our brokerage, I should buy your brokerage. And I thought he was messing with me. I thought he was absolutely messing with me, but he's like no, I, I'm, I can only meet at 6 am tomorrow morning. Can you meet me? Like, yeah, sure, um, it was at 6 am. He was ready to go. We talked for like an hour and they moved. I have nothing bad to say about the NFI team. Those teams can be, those deals can be grueling and you can't see the other side anymore. But I have nothing. This was great. It was a really, really interesting process.
Speaker 1:Two couple thoughts. One you know this will be my 40. 42, 43, maybe the 45th episode I've done. Dave is maybe the easiest person to talk to and also one of the most energetic. He's got a great energy to him. But it's cool that I heard the others. I've had him on the show and he kind of told the other side of the story. I just want to call that out. But the other thing I'm curious about is talk to me about the due diligence process. How long was it?
Speaker 2:it was four or five weeks. Um, they went super fast. His team knows how to buy brokerage. They've done it many times like this is what we're looking for. This is what we're doing. Uh, dave and uh his um, cto did, uh, it did a walkthrough of the tech. They liked it. I mean, knock on wood, I think they liked it, given where we are. So, and yeah, deal happened really really seamlessly.
Speaker 1:And how hard was it? So this is interesting because I'm recording this the same just this morning. This episode will come out in two weeks. Interesting because I'm recording this the same just this morning. This episode will come out in two weeks, but I'm recording this the same morning. I just recorded my making of a merger episode with two companies that are merging that are announcing what will be about two weeks from prior to this being live. But they also had their due diligence process only be four to five weeks and it's crazy to me to hear that for now, two of these deals like that and one of the reasons that they said they were able to go so quickly is because these guys had experience, public company experience and their data room. They already had all the audited financials needed. Did you guys have all of that prepared or did you guys have to kind of go crazy getting everything ready and into a data room for the team?
Speaker 2:We just almost backed not that long ago. So we had everything in order. We had everything more buttoned up that they should be for the company we were more buttoned up than they should be for the company we were and the finance team that we had and that we have everyone is extremely, extremely qualified to do this right. They was NFI knows what they're doing. They knew exactly what question to ask and knew exactly, and on our side, everything was buttoned up as we just went through like a process of way harder um than than this was. So we it just perfect timing that makes sense.
Speaker 1:That makes sense, so, so they. So nfi buys the brokerage operations and agrees to license the technology as your first customer, or how did that? How did that all they?
Speaker 2:they, they license, uh, the whole thing, right, because they still so they, they kept the, they bought the people, um, all of our brokers and all that stuff and it's, and, and the customer and the carriers that came with it. So they bought that part of the business that today, and from the day of the acquisition and today and probably for the future, um is powered by all of our tech. So they're licensing that part.
Speaker 1:And is that white labeled then for them? So like is there now an NFI app that their carriers log onto to book loads or it's still the transfix app.
Speaker 2:It's still our transfix app. Um, there are are loads, or it's still the transfix app. It's still our transfix app. It's still our transfix app, you just?
Speaker 1:see the NFI loads in there and so where does that leave the transfix in your team moving forward? How many people are on the team and what is the new vision for the company and what you're building?
Speaker 2:So now we're down to about 50, give or take, and it's funny. So we came out I'll take a step back in order to answer the question Like, we came out of this and we had so many different tools and product and we thought, okay, well, if we come out and say we have a TMS, the sales cycle is going to take years. It's just going to take a very, very long time. We're trying to convince you who have you broker, abc, who is starting to invest into your own team to build your own TMS, or you've had Mac Cloud for 10 years. We're trying to tell you who are we to do that? So we thought, fine, we're not going to do that. We're going to have modules and each of these modules is going to be connected to your TMS and completely our automation around tendering, accept or rejecting shipments. We'll have an API so that you can stay in your TMS and we'll automate that for you, or pricing or all of these stuff. And we came out and we started pitching all of these different solutions and there's one thing that was very clear In this current market, people aren't as focused on automation if it doesn't make them more money because they have bandwidth, because they have people. They're not looking for automation, they're looking for increase and the best way to do that is pricing. It's price better and we realized quickly we priced at Transfig very differently than anybody else.
Speaker 2:It was done by machine learning models. It was done by data scientists, but also by brokers and our data science team bookload. They talk to drivers. They understand what it takes to negotiate. They hash it out every week with our brokerage team. The model was wrong here. The model was right here. The model six months ago predicted this rate and now it's this rate. What happened? We have a unique experience data science team that has built these models and now to give you one quick example of what I mean by like, it's very different. A lot of the brokers we're talking to are trying to understand where the market is going. With all due respect, it matters to some extent, but not really. What's more important is to understand that when it's a tight market, you buy 150% above the market and then you need to understand the market. It's more important to understand what your cost is going to be than what the market is going to be, and we base all our decisions based on our cost, not the market.
Speaker 1:And how can you know? Like, history doesn't repeat necessarily. I get that it rhymes as the saying or whatever the saying is, but historically, when the market tightens on Chicago to Dallas this is a made up example the rate goes up by 150 percent. For you this time it might go up by 200. So, like, how can you have the level of confidence that you do in the pricing, given that I don't know the future is is yet to be told. Does that sound weird to say it like?
Speaker 2:that no, no, but they're like, so they're. The easy answer is really really complex data science behind us. I'll give you some results before I go into the how a little bit. And there's a couple of reasons why we can be this accurate compared to the rest of the market. There's two ways to understand how. How model works and their accuracy. And I'm just going to speak in super high level terms because the just to make sure we have the time. Otherwise I can talk about this for hours.
Speaker 2:But you've got like your MPEs and your MAPE. The MPE is sort of like the average error. You take all your errors, you average them out. That's your percent of errors. The MAPE you can think of it as like the highest volatility, the worst case scenario, like when it's bad, like how bad does it get? We had a model that we put in December of 2023 that priced every single month of 2024. And in October our MP was negative 1%, meaning we were, as an average, under by 1%, and our MAIP is 8%. In the worst case scenario, when we thought a shipment was going to be $1,000, it ended up being $1,080.
Speaker 2:Those results just don't lie and there's a lot of reason why we can get that level of granularity. And the first one usually stems from geography, on how people see the world versus how truck drivers sees the world and where capacity come from. A simple example is Boston to Houston versus Springfield to Houston. If I'm telling you this, you're telling me it's the same capacity. Capacity comes from the same spot. They're 40 miles apart, 50 miles apart. Right, you're going to grab capacity from that same area.
Speaker 2:In every single tool out there. There are going to be different markets, there's a different zip, different cities, different everything, and so once you can understand geography differently and you can understand where capacity comes from, regardless of where the pickup is, and understand that like they're going to, uh, like a great freight market or a terrible freight market, or understand all the idiosyncrasies of a lane, of a shippers, of a warehouse, um, of all this, and you realize that the market is only a small factor to what comes into that rate. And when you can do that, and then just all of the features that comes into our model, you end up with a fairly accurate rate. Now, if there's another COVID, that will be interesting, we'll have to react and we'll have to figure it out. But outside of a global pandemic, those models are a little bit scary accurate.
Speaker 1:Interesting. So how would you describe, how would you fill in the blank in this sentence? Today, Transfix is a blank company.
Speaker 2:Pricing solutions company. We'll take care of everything pricing related. If you want, if you want automated spot, if you want, uh, rfps, like we'll, we'll come in, we'll ingest your data, will be an extension of your data science team. It will feel like we're part of your team because our model is going to be your model. There's only going to be one of those. It's going to be unique to you because it's going to understand how you react when there's a short lead time. How is your team structured? How is your carrier network? What is your shipper network? How good are you at scheduling appointments? All of that is going to go into the model. It's going to end up being your model of your baby, which is going to help it grow. That was weird. That was a weird analogy.
Speaker 1:No, I think it sounded good. I heard it your model, your baby. We're going to help it grow. That could be a commercial.
Speaker 2:I don't like it.
Speaker 1:Maybe my marketing jews go listen to this like, oh no, what did you say? I thought it sounded good. Um, okay, so pricing on pricing solutions, pricing automation solutions and will this be sold marketed strictly to brokers, to assets as well? Does it make sense for an asset carrier to use this for pricing their own business?
Speaker 2:we. We're probably going to try one day, just not today. It is a different business. Um, shippers, on the other hand, is a lot more interesting. Uh, we're we're targeting some large shippers that, like, are doing rfps and helping them understand what is the cost of their network. What happened if the real interesting question for a shipper is you managed to negotiate with this broker and beat him down, great. And now it's six months in. We're going to tell you they're going to ask for a reprice. They just are. This is where this thing is going, this is where the market is going. They're going to ask you for a reprice and this is the reprice they're going to ask for.
Speaker 2:This is how we worked with shippers. So, when we saw that the market, the rates, were going up, we often came in before the RFP. We would try to get a meeting before the RFP season. We would run their network from the previous RFP. The RFPs don get a meeting before the RFP season. We would run their network from the previous RFP. The RFPs don't change that much. They change a little bit, but year over year they don't change that much. We take the previous year, which show you. Here's the rate If we price at where we should be, to not renegotiate with you for the whole year.
Speaker 2:We're not going to be competitive. We're not going to be competitive. We're not going to win any freight because everybody's going to bid right here. So we're going to bid right here with everybody so that we can win the freight and you can. If we're wrong, you don't have to pay for it. But if we're right in six months, we won't be the first to come renegotiate, but we won't be the last. And this is what we're going to ask. Um, and that that conversation usually went really really well, um, and that's so that we can offer that now for the shippers on the other side yeah, I mean that's interesting.
Speaker 1:Um, do you see any risk or conflict in trying to cater to both shippers and brokers, or they being concerned there, or anything I like?
Speaker 2:I don't think so. We, we do it purely on your data, right? So like, even if we do it on two, two brokers for the exact same lane, they're going to be very different price because it's how good are you at servicing that lane?
Speaker 1:is there no sense in offering to aggregate the data and offer to the brokers like hey, you know there's 50 of you in the pool. If you guys want to contribute, you get better. Theoretically, you get better pricing at that point or more accurate, or is that not make sense?
Speaker 2:No, we actually. So we have a partnership with both the ATN, sonar to get an understanding of the market. They already do the aggregations. They do, they have do the aggregations. They have a lot more data than we can aggregate with 50 shippers. And so, though, when you work with us, your data is your data. We don't mingle it. And the problem what happens when you mingle this data and then you get an index, and then I have to understand how you price against that index, well, your data, it influence that index quite a bit. Unless I have significant, unless you have a lot, a lot of data, when you distill down to lane level, 50 broker is not going to be that much, and so it's going to be highly influenced. So the model is going to show back to like oh yeah, you're buying at the market, because my vision of the market is your data anyway. So we're not looking to do that.
Speaker 1:I'm curious you said that you have partnerships with DAT and Sonar, because initially my guess would have been that this is a competitive product to those tools.
Speaker 2:And both of these tools are there to tell you what the market is. But we still need to have an understanding of what the market is. We just don't want to build that understanding for ourselves by doing what you just said and aggregated all those indexes together, all those data points, to create an index together. We believe that, whether you like it or not, we can debate that another time. But carriers look at the DAT. Carriers look at Sonar. That is going to influence your negotiation. You have to know what those tools say, Because it's going to influence the negotiation. You're going to hear on the phone from a dispatcher well, I see that at $1,200. You're going to hear that. But if our tool doesn't reflect an understanding of that, it's not going to work.
Speaker 1:And I'm curious. I mean, are you guys in the you know market cycle prediction game these days like can you tell me when the market's gonna has them as the freight recession? Over, I can do you feel like you have a confident ability to answer that question?
Speaker 2:as much as I don't see into the futures. However, in December 2023, when I told you we ran that model, we also ran it to understand. Like hey, go as far out in the future as you can and tell us when the market turns. That model said June, july 2025. Today, I still believe that's accurate.
Speaker 1:Today, I still believe that's accurate, um, but I think the definition of market turning might be different for your definition.
Speaker 2:Yeah, it might be a little bit different. I don't think we've had pre-covid. The cycles were long and steady, right going up, then going down, but they were long and steady and they weren't like crazy spiked away. I think the combination of COVID and tech has made the cycle go much shorter, much faster. Now, coming in 2025, I think that what I mean by turning is that less trucking companies will be running at rates that they can't run at forever. The market will finally stabilize in ways that truck drivers won't lose money driving, which, frankly, is heartbreaking in the first place. So that, to me, is what I mean by market turning.
Speaker 2:I think today you're seeing signs of it starting to happen, because some rates are getting inflating and that could be a result of a lot of macro factors. Again, there's another Pandora's box. Who knows what's happening with tariff and all this stuff. We'll see. We'll see what happened. The spike today is like going going to stick. I'm not sure, but what I hope is that by June July 25, you get stabilities on the on the supply side, and that should drive the rest. Give me an understanding.
Speaker 1:Only time will tell. Give me an understanding, only time will tell. Give me, give me an idea of what you think the future of brokerage looks like in, in, you know, as as a guy who's an engineer at heart and who understands this tech at a level that I think most of my audience probably is is not there, myself included um, how do you see the role of ai coming into the space and, you know, evolving or influencing brokerage? You know? What big changes do you see potentially happening? Give me your kind of perspective there uh, this is.
Speaker 2:if you ask me this a week ago, I'd tell you like AI is just a fab and it's just not. It's not gonna go very far. And my grandfather always said, like only idiots don't change their mind. And I had a test done on me with a bot calling me and negotiating as if I was a truck driver or dispatcher negotiating with me a rate, and that bot was good. That bot was real good. Um, and so now I have all my uh, all my billions uh, shaken up a little bit. Um, I I think there's a huge discrepancy where the market should be and wants to be and the tech they use today, and that makes it really hard to understand where everything is going.
Speaker 2:We got one side of the market talking about AI bots that are calling with incredibly complicated tech and things like that, and on the other side, you've got Broeger and Shipper using EDI. Edi was invented in the 60s, I think. I think cassette tape, which I'm not sure a lot of people will know what that is these days, was invented after EDI. Hey, so that's a tech that is used today and we're talking about, like, where is AI going to take us? So I hope that EDI stop existing very soon for the sake of everybody. But it makes it hard to know where are people going to invest and where they're not going to invest, when their TMS can't plug in to all this new tech and when it requires the providers of this new tech to figure out how to integrate with those TMS. That that makes innovation very challenging.
Speaker 1:Right, yeah, just, I think the old school TMSs are in trouble, the ones that you know, the on-prem um as400 type green screen that stuff is. You know, I wouldn't want to be someone messing around in that world.
Speaker 2:Yeah, today yeah, I so, yeah, I totally agree. Um, and so where it's going? I hope it's going to a place where ai can get keep impairing people, can keep, uh, can keep, frankly, bringing a stable life, more stable lifetime to, to carriers, like I don't know. I tend to be very focused on the carriers because I have immense empathy for what they do Immense, immense empathy. It's by far the hardest Out of all of our jobs within supply chain. I don't think there's two, there's another one that compares into how difficult that is. I hope at least they'll be focused on helping demo. That's a naive hope and I'm not sure I'm right.
Speaker 1:I guess we'll see. What do you think the future looks like for transfix?
Speaker 2:I don't know exactly how long we're talking about, because these pricing solutions, I think we're going to be selling them for quite a bit. Our thesis has always been we have all these tools, they're still available, they're basically built and so, if we get in with pricing, when are we evolving? Like I think there'll be an evolution towards bringing our carrier app into this world and then bringing our automation to this world, and I don't know when it is, but we it's not because we've pivoted that are like um aspirations have gone down. By any means, I still very much intend to try to make an impact, to get people lives in our industry better in one way or another, to have a mark. That's what drew and I like set up to do right. We want um, we want to change the way this, this industry, works, and this is just a different path towards it. Uh, obviously for the better.
Speaker 1:Um, well, final question what's the most valuable lesson you've learned in your 10 years in freight?
Speaker 2:you know, I should have, like I've listened to enough of your podcast to know you were going to ask that yet I don't know if I ask that all the time, this one is just going to naturally float into that lesson I don't know if it's a lesson there's just like people, um, doing right by people, by your team, is probably the number one thing that matters, uh, because when you need them to be here and they are there, there is nothing better, there's nothing that is going to help a company grow better. There is nothing that is going to be the people on your team, and making sure you're here for them so that they're here for you, is something you know. Right, I didn't say anything earth shattering, but there's a difference between knowing it and then just feeling it is through, through the ups and the down of a con that a company can have. Um, that to me is, um, this probably, uh, the ceiling versus the knowing.
Speaker 1:Well said, we're going to end with that. So to our listeners, I hope you enjoyed the transfix story, jonathan's story and his role in building this business, helping it pivot and become a whole new thing. And that's all we got. We'll see you in two weeks.
