Supply Chain Unlocked

Ep. 6 - Fixing Freight Rates with Data Ownership | With Sam Tibbs

• Dr. Matthew Waller

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 26:53

What happens when a Navy nuclear power alum and finance PhD turns his focus to trucking? Sam Tibbs joins us to make a bold case: freight doesn’t have a tech problem, it has an incentive problem. He traces the unlikely route from teaching and data science to building the tools that exposed lane-level profit and drove the most-used variables in Sonar, then explains why today’s rate products miss the mark. If your pricing relies on paperwork extracted after delivery and guesses that fill in missing details, you’re driving with fogged windows.

We dig into the real bottleneck: timely, detailed data at the moment of booking. Sam argues that brokers create the signal, yet pay to buy back their own data days later, stripped of nuance like stops, hazmat, or lead time. His answer is refreshingly practical and familiar to anyone in finance: build a neutral, broker-owned data layer, much like how MasterCard aligned competing banks, where contributors are rewarded for high-quality inputs. With ownership and governance in place, the incentives flip, participation rises, and the rate product can finally show distributions that move with the market instead of stale averages.

Along the way, we contrast “great data plus a good model” with “good data plus a great model,” explore why LLMs can’t rescue missing truth, and outline where competitors should cooperate versus where they should battle. The near-term payoff is a cleaner rate signal and fewer pricing mismatches; the long-term vision is brokers turning a cost center into a profit center while shippers and carriers benefit from sharper, faster decisions. If you care about freight pricing, data quality, or building markets that actually work, this conversation maps a credible path forward.

If this resonates, follow the show, share it with a colleague, and leave a quick review so more folks in freight can find it. What's your take: should brokers own the data layer or keep buying it back?

SPEAKER_00

I have with me today Sam Tibbs, who is the co-founder, CEO, and CPO of TroKi Data Corp. He has a PhD in finance from uh the University of Tennessee. He also has a uh CFA, which is um very challenging to get. Um, and um he uh he has an undergraduate degree in nuclear engineering technology, um, and he has a lot of experience in uh transportation logistics, um, but he also has lots of experience and knowledge of early stage companies and technology companies in general. But uh welcome, Sam. Thanks for joining us today.

SPEAKER_01

Pleasure, thanks for having me. And uh yeah, you uh some of your references take me way back to uh I guess I'm showing my age son, but uh yeah, back to my Navy dating when I was in the Navy Nuclear Power Program. So yeah, it's a pleasure to be here. I look forward to chatting and uh thanks for your time in advance.

SPEAKER_00

So before we talk about Trochi, would you tell me a little bit about how did you get involved in trucking and logistics, especially since you know you came from nuclear technology to a PhD in finance or CFA? Um I'm sure that helps you in in business in general, but how did you make that transition?

Early Wins In Data And Pricing

SPEAKER_01

Right, great question. So I'll start and say, you know, uh as a fellow academic, I mean, uh I've taught lots of classes, and I'm sure you can relate that. Students would be very uncertain what their major was. And I took remind them, you know, you don't know for a while, you know, and and I can joke, I worked at a skating ring, an aircraft carrier, coffee shop. I taught, and then I came, you know, into trucking. So it's manchers and the trucking wasn't what I was expecting. Well I mean, it was a kind of a forced thing in that I had to move back to the Chattanooga, Tennessee area to take care of some family emergencies. I had uh been living in UAE for six years. I wanted to travel a bunch, and that's how I met like Andy Fred Rample and American University of Charter, Great University. And I had to come back quickly, and you can't just get a professional job. So I should have known in advance, since I was turned returned to the Chattanooga area, that transportation trucking would be the most likely outcome. Uh, but for the first 14 months, I was determined because I had a CFA charter, that it's a leading practitioner uh designation for finance. I was determined I'm I'll just get a finance job, things will be fine, take care of family issues, live in Chattanooga, no worries. Uh well, it took me 12 months to have my first interview. Um, and that only came after I learned that data science was this new fancy thing, and it was supposedly the sexist job in America. And I I literally had filled out hundreds of applications and cover letters on the finance side, but I I changed my resume to say data scientist based on the academic research I've done in the past and the modeling. And I immediately had two interviews. One was with Freight Wave before they had their Series A, and also with Covenant Transport, and I started coving to transport about a month or so later. And so I decided at that point in time I would never be unemployed again or do whatever I could, and I'm going to learn trucking as much as possible. And so that started uh that process. And it's, you know, I joke. I mean, it's not my joke. I say that if I'd known how interesting supply chain logistics was, I would have chosen to start there for the finance.

SPEAKER_00

Well, that's that's great. Many times combining things that aren't typically combined uh generates innovation. And speaking of innovations, and don't be shy, tell us a little bit about some of your successes, a few of the things you've built.

SPEAKER_01

Yeah, yeah, I'm happy to. So uh when I started coming to transport, uh it was, you know, I dove in, I'm fully immersed, and and and I guess I could admit this now, but I mean I was contract then, so I couldn't, I wasn't allowed more eight hours, but I made sure I would be the first one to make coffee every morning and I'd I'd leave seven or at night, and I had access to tons of data detail from all the carrier transactions and movements that the government had. But then I learned that they didn't actually, on a lane level, know what the profitability by lane was as it was run. And so I created the what they call the, we called the profit analyzer, which allowed, you know, effectively real time or at after the load was done, because you still have to control for all the wait times and everything, um, what the profitability was by lane. Happy to say they used that for many years. I think they finally just stopped using it about a year or so ago. And then at I'm gonna transport, I'm sorry, at freightways that was directed data science. Uh so from raw variable, so from raw data to variable, I created uh 106 of the top 250 variables uh just by myself. And if you throw in the junior data scientists I worked with, about 150 of the top 250, and just myself alone was responsible for 74% of the clicks on sonar for about two years. Uh, so the benefit I had at Covenant was seeing all that raw data. So when I when I got to freight waves, I started that raw data again and created variables that made sense to explain trucking. Uh cloud trucks, I I created their, you know, built the pricing system where the second I was the second data guy. I loved it because when I got there, they already had 10 engineers sending data, but they only had one data person of the second data guy. Uh here fill out the pricing, profitability, uh, heat map, and started some with their um their RFP push to contracts. And I'll go to most recently at Triad Financial. They wanted to build a rate product. Like, I love it. We didn't quite agree on sorting the data, but I built a rate for it product for them. And uh it was a wonderful learning experience, and and then life happened. Uh and uh I'll pause there before I step into TroCai, but some of the successes and it's uh it's been a very fun ride.

The Pivot To Trochi’s Mission

SPEAKER_00

That's great. And uh before I I'm gonna have the editors cut out what I'm saying right now, and then we'll we'll come back to it. Um, back to the interview. But I just thought of something I should have thought of earlier. Um, if you want visibility to companies in Northwest Arkansas at all, and and even uh other things, you there's um there's a new initiative going on in Bentonville called Arcade, A-R-K-A-D-E. It's funded by some financiers out of I think Sweden and New York, and some even local people. But you might want to look into it. It's it wouldn't cost you anything, um, but it would be a way to potentially get some visibility. Just a thought. So we can I'll talk you, I'll send you something about it. Wonderful, one second. Okay, back to the interview. Okay. So, Sam, what really motivated you to start Trochi specifically?

SPEAKER_01

I mean, during my time at Triumph, it was a wonderful time. Uh, a lot of great people there and everything. And I guess I'll comment every place I worked at had the benefit of very nice uh people and experiences. But at Triumph, they want to build a rate product, uh, they want to build it off the factoring document because that's a large port revenue, and they they knew there'd be synergies on extraction of that paperwork, and then they can match it to the audit data set. On the paperwork side, though, the issue is two main issues. One is that it comes after the load is delivered. So, on a timeless perspective, it's not ideal uh for a rate product. The other issue is that doing paper extraction, even though you can get total 95% extraction rates, 95% extraction rates is misleading in the sense that it doesn't mean you got 19 out of 20 rate cons or purchase orders correct. It means 19 out of 20 items. So if you, you know, you had three rate cons, it might be you got 100% of one, 100% one, 85% of another. So it created restricted the data. But during that time of triumph, I clearly saw and clearly understood that there were two big complaints people had about the rate products. One was the quality of the data, and the other related to the cost of it. And the cost of it really confused me in a sense. I mean, I knew the expense before, but I thought of it differently. And I came to think of it like oil and gas. If I'm sitting on a piece of property and have mineral rights and there's oil underneath that ground, I actually directly benefit from. But in the case of rape products, that oil effectively gets sent to the refinery for nothing. And then the producer of that oil actually has to buy their data back when it's aggregated with someone else with other pieces of data. So to me, that model didn't make sense. So unfortunately, Tribe canceled the rate product I was working on, let me go, and I was free to compete in whatever way I wanted to. So I decided this is the right path. So I spoke to some of my friends, and uh we decided to build the product.

Data Problem, Not A Tech Problem

SPEAKER_00

You've said this wasn't really a tech problem, but more of a data problem. What made that clear to you?

SPEAKER_01

If you think of history throughout time with marketplaces, it's more about getting all the people in one spot. It doesn't necessarily need to be tech for that marketplace, that data exchange or the product and monetary exchange to happen. But in the case of let's take uh Convoy or any other attempt at building a marketplace, you know, the the tech's extremely savvy, extremely highly qualified. But even the case with Convoi, they would talk about siloed data when they built in the broker's a broker's type system. So it's really a function of not ability to identify what the best rate is, it's the ability to have an access to view it. So if you have more data in one place, it greatly benefits the process of calculating the rate of doing a marketplace or any other data providing as opposed to some fancy AI that may you know try to fill parts in. Or said differently, I view it as great data with a good model always beats good data with a great model. Maybe sometime in the future, 20 years from now, that'll change. But until this case, every LLM depends on the data providers. That's still the case. It's going to be the case for a long time going forward. So again, the data is the driving force behind whatever rate product or marketplace or whatever the data products you want.

Why High-Quality Truckload Data Is Scarce

SPEAKER_00

So why is truckload data, high-quality truckload data, so hard to get in one place?

SPEAKER_01

Well, that's a great question. Uh and essentially, and and and it's not just truckload data, but it in this case is specifically related to truckload data, and that the systems that are in place that are the structures that are built are again data shared, and the the customer then buys the data back in the aggregated form. But in that case, the creator of the data, in this case, I'm referring to brokerages for the most the spot truckload market, they're not rewarded for the data they share. So there's very little incentive to make it timely or highly detailed. So because of that, you you share what you have to share in order to participate in the current market that, for example, truckload rate. But if they they say, well, we want better, we want best, whatever other detail, it's like, well, I get nothing out of this. I'm gonna do the bare minimum. But in the case of the ownership model, if you realize that if brokers were rewarded for the data they created, then they're much more highly incentivized to share that better data. Now, why is this better data matter? It matters, it doesn't just create a better product, but it also increases great efficiency. So let's say you're a CPG carrier and you're moving freight around all over the place and you don't have good information for contacting or setting the best rate or contacting the best provider to make the most efficient movement of that freight. Well, in that case, that costs you more money. Across the board, there's a huge deadweight loss that exists for shippers that brokers bear the cost of and carriers bear the cost of because of the inefficiency of data. So if you have better data sources because the data providers incentivize, which is why TroCau is going to be 80% owned by brokers that provide data, then the entire system works better and everyone across the board wins.

SPEAKER_00

So, what incentives were missing that kept data providers from fully participating?

SPEAKER_01

So the main incentive is initially when you know when Truffalo rate products started, there wasn't there wasn't a requirement or an incentive. I believe the first national account for DAT with Alan Lund, for example. In that case, they just needed a rate product, and it wasn't high requirement or anything to make us it, you didn't need a great rate product at that time in order to have something that was substantially better than nothing. Since then, you know, obviously, amount of data and detail and the handling of it, the uses of it that can take place through, you know, the great increases in computing power, et cetera, means that better data is much more valuable. And so based on initially how it's structured and and brokers didn't have ownership of it, it didn't work out well for creating a better system with better data being applied.

SPEAKER_00

What tends to go wrong when third parties try to extract value from data that they didn't create?

SPEAKER_01

Not to say you can't have a good rate product. Uh not to say you can't have good things with it, but if again, if you know, if you look at certain data providers, their their rate may be quoted three or seven days old. Uh some of the rates may be filled in based on estimates uh from algorithms or models that do a pretty good job as long as the freight rates are consistent in the areas and the lanes in which they're comparing to. But it it's it's a very imperfect system that just you know unnecessarily it creates improper pricing and inefficiency. So if you are available, if the actual true data is available and the tech exists to put it together in a timely fashion, which is certainly the case these days, then those other inefficiencies that exist that should be replaced. And the people that create the data should be the ones they're rewarded for. And that substantial, you know, important relationship, you know, it's means everything is moving forward more efficiently.

MasterCard As The Mental Model

SPEAKER_00

Yeah, at some point you realize that trochi resembled MasterCard's early structure. And if you wouldn't mind explaining what you mean by that, but also um how did that hit you or when did it hit you?

SPEAKER_01

So, you know, for so I've been I've been chasing this data for over a year, year and several months. And uh if you heard that, someone's hunting nearby. I live in the middle of the woods. I don't know if you hear that. I guess like business. So someone apparently my guess is just shot something over a deer or something. I shouldn't have even said anything about it. But I wasn't sure how well you could hear it. But uh, now I apologize.

SPEAKER_00

No problem.

Convergent Evolution And Governance

SPEAKER_01

So I realize it was MasterCard, you know, only about I'm gonna say literally a month and a half ago, but I've been doing this for close to a year and a half. Now, initially, I should say I'm reading Brad Jacobs now, how to, you know, his title, How to Make a Few Billion Dollars. It's a very interesting, catchy title. But there's a lot of good content there. One, he makes a comment about beating yourself up. I can say a month ago, when I was listening to some other podcasts and et cetera, and outside the weeds of of trucking that I'm usually always buried in, I was listening to a podcast on Visa and it clicked. And I'm like, wow, what we built is is actually, you know, I dug in more and MasterCard started in a slightly different structure, but what we're building is very, very similar to MasterCard. And I'm I'm like, you know, literally like, how could I not notice this? I mean, MasterCard's a financial transaction business. I'm a finance PhD. You know, I spent, you know, I don't know, much more time than I should have going, ah, ridiculous. But then then what I realized is that, you know, I take this, it's really quite reassuring to say that independently of understanding how my how Micro I almost said Microsoft, how MasterCard was built, we build the same type system by saying, well, what are the proper incentives to have people companies work together that are competing? What kind of governance structure is needed to say that we'll cooperate on this level and we'll compete on everything else? And we independently build that, and then we came to the conclusion afterwards, wow, this is like like MasterCard. So uh, so that's reassuring. Uh, and I passed the annoyance of of you know being myself overt, but uh it's really quite a nice situation that MasterCard they decided that they would have their infrastructure and and standardize certain things so they could have scale and operability across the US backward a day when banks were much more localized. Uh whereas Troki is saying, hey, you guys have this joint data layer that you need to share, come to terms with what you want to share and what businesses, what parts of those uh data uh products you wish to have to reduce products, the first one being the truckload rate, then make other choices after that, and then have great say in how the businesses run. So that the banks start in MasterCard and the brokers are starting Trochi, very similar systems. So even though very different industries, it's the same outcome. And I I credit it to economics win.

SPEAKER_00

You've described that as convergent evolution. What does that mean in the business context?

SPEAKER_01

So if I think of you know, sharks versus dolphins, is is the first example that comes to mind, and that they have very similar characteristics and shapes. Uh dolphins, of course, are mammals, sharks, of course, fish, but they have very similar characteristics because of the environment they were in. And so we're in an economic system, both cases, MasterCard and herself at different time periods with starting, but still very similar economic systems, and that those shapes just to look quite similar. Uh I would like to think that, you know, since dolphins are smarter and they work together better than sharks, I prefer us to we'll call us the uh the dolphin portion and say MasterCard are the sharks, but I I do think that's a just an example of how things under similar conditions come to the same point. So um yeah, it's fortunate that uh that that's the case.

Where To Compete And Cooperate

SPEAKER_00

How do you decide where competitors should compete and where they should cooperate? So in this case, I'll I'll do two parts.

SPEAKER_01

Uh first, I'll make I'll make a I'll make a academic joke and say that hey, you and I I'm sure love having super smart students in our classroom, but there are times I couldn't answer the question. As an early professor, I used this. Later on, I didn't use it. If I had a question I was asked, I would say, well, I could answer that, but then it'd be testable. And then no one would answer. You know, no one would comment on it. But uh instead, what I do is later look up and find the answer once I'm matured and then come back to the class with it. Uh in this case, my easy answer is well, brokers own 80% of the business. If brokers want a particular product, uh then by all means they can develop it, guide it, guide strategy, who competes, where on what. My guiding point is that I need to have a very strong tech system. Our CTO was the one that led the software team that built FreightWave Sonar. Uh, we have a stack stood up, we have several engineers that we work with. So we have the new right tech structure in place. Now we work on that the governance structure in place, and that governance structure says, all right, let's very carefully delineate what's what data we're going to share, what that data looks like. And then the next step is what data is that data used for. And so the first step is a truckload rate product. Everybody already uses it. Uh, there's effectively zero product fit market fit risk. And so the next step is we make sure we have the system in place. If they wish to build a marketplace, uh DAT and Triumph Exchange are both building marketplaces in between them over the last year, they spent at least half a billion dollars to build that. And they're doing that with inferior data. Uh brokers create knowledge data. If they wish to have the same thing, my goal is to have the system in place. They can do that. If they say decide on that data layer, that's not a product they wish to compete on, that's fine. There's other opportunities as well. So, but the brokers make the final answer. I think I have some right answers, but I I'm not the know-it-all. Uh, I'll make suggestions and then and then they decide.

SPEAKER_00

Why is it important that Trochi not disintermediate or compete with brokers?

What Changes First For Brokers

SPEAKER_01

Um we said so that brokers weren't 100% our partners. We want them on equal playing field. Uh when I first started this and was pursuing VC money, VC said, okay, we love it, blah, blah, had some great contacts. Some I still talk to today. They said we need to have different voting class shares. If you're gonna give away so much of your ownership in exchange for data rights, then we don't we don't feel comfortable, you know, so quickly giving up our ability to manage the company and guide the company. Um, okay, we can have dual class voting shares. Berkshire Hathaway has it, others have it. Uh MasterCard actually now has it as well, uh post-IPO. Um as referred to the earlier times of what he cut it. And but I talked to a few brokers, and brokers said, you know, hey, you know, if if you really want to be our partners, then just ask us for some capital. Just make it reasonable, make it fair. We know that your goal is to convert our truck load data cost center to a profit center. And if you want to do that though, don't don't chase the outside money, just work with us. And and by that process, that means that brokers own more, they have greater input. And I'm not doing anything that would change that position. I firmly believe long-term that the intermediaries that exist in the market, the brokers, that that will continue to expand. I think the low barrier to entry to carriers is continue to be a larger portion of the poor higher market. And so we see brokers as the long-term winners, regardless, based on that situation. But if they can also control their data and create different, different and superior products, then it's a win-win. So it only makes us real sense for us to profit or to partner with them so that we all profit together. And to do that, we can't have anything in place that doesn't have us 100% a lot. Because the main thing is the data is not our tech, it's not what we propose or the profits we believe we can make for them. It's we're not, they're going to trust us. So we have a governance structure in place that means we're not going to intermediate them, we're not going to harm their business, and we show that they're the guiding force, then that's the right framework for this to exist.

SPEAKER_00

If Trochi works as intended, what changes first for brokers? And what do you see as success five or 10 years from now?

Closing Thanks And Conference Plug

SPEAKER_01

Well, that's that's some great questions. Uh on this, I'll make my assumption that some of the things I'm supporting or promoting uh are accepted by brokers. Um ultimately it's going to be again their choice. They own 80% of the business. Uh, the first thing is we get a better rate product. The current situation with rate products is all typically or almost always, the details aren't shared. So if you have a truckload rate that is a multi-stop and a hazmat or liftgate or steam or escort or something, that detail may not be in the quote. So you have the quote and you have the pickup and destination, uh, but you don't have those other details. So it's very difficult to actually understand what the proper standardized price is for particular globe. So we'll give better pricing information. And because of the detail we get, and we get it at booking date, um, we can show the full distribution. And Jason Miller's talked about this at our couple of conferences, where it's not just knowing the 25th and 70 percentile on the average, but if you can see the distribution, how the distribution shifts, you can have much better insight to how the market's changing more quickly. Uh, so better insight for that. I know AVRL wants to be, they've talked about being the uh McKenzie of pricing. Uh, every good McKinsey or consultant can better perform their job if they have better tools. And so the better pricing detail, the better information as far as actual trucks in the outer market, lead time, the ability to drop loads that are picked up too quick, but I mean it's picked up, let's say, or are set up, priced 12 hours before pickup. Um, those can be dropped out, but all the additional information, and if you're not in truckload, I apologize. But the summary is all the additional data detail we're able to provide because brokers are now willing to share it because they're rewarded for it, means a much better system in place.

SPEAKER_00

That's great. Sam, this is really impressive. Now I now that I've heard the full story, I understand more clearly how your finance expertise and your logistics trucking expertise really came together in this, because I think you need that kind of level of expertise. And of course, having a PhD in finance in a CFA, you have all kinds of knowledge and of finance that very few people in trucking do. Very few. Uh even CFOs, a lot of times, of trucking companies, I mean, they may have a lot of understanding of finance, but a lot of times they don't know as much of the information about the, for example, rate structures and how rates change and supply and demand. And um so but thank you so much for joining me today. It's been really interesting.

SPEAKER_01

My pleasure. And if I don't I'll make a quick plug in that we've had two conferences already where Jason Miller leads the conference and uh he gives us great insights to the markets, et cetera. And we're gonna have another one probably in Chattanooga, then later in Chicago. So just follow me if you want. Sam's hands on LinkedIn and you'll get more information. And uh I appreciate your time. It's been a pleasure.

SPEAKER_00

Okay, let me, it's still uploading, so let's give it a couple of minutes. Or look uh it's close to it's at 99%.

SPEAKER_01

So sorry, I didn't think I mentioned about that future conferencing before. So I hope to plug it in. Yeah, no, of course.

SPEAKER_00

Absolutely. In fact, I'd I'd be happy to post about anything like that you want. Well, I'm glad that this now I understand when we talked the first time I had a a general idea, but now I understand better. Slick idea you have. And I I can see how having worked at Freight Waves gave you insights to this opportunity.

SPEAKER_01

Right, right, right.

SPEAKER_00

Which is great.

unknown

Yeah.

SPEAKER_01

I'll go further to say that when I first called to an investor the first time he rubbed his temples and said, You want to give away what your company? But you'll appreciate this in that, yes, we're only trying to keep an investor's group bound or trying to keep 20% of the company. But if brokers own 80%, they're gonna cut off data to day T. They're gonna cough sonar. So I as I tell people, I'd rather try to have I was owned 20%.