Good afternoon. My name is Dan Moore. I'm the Senior Transportation Analyst here at Baird. Fortunate to have a good chunk of the executive leadership of C. H. Robinson with us today. It's a pleasure to have both of you here. Chuck, you as well. Appreciate your time. Appreciate the opportunity to ask you questions.
Happy to be here.
Yeah, happy to be here. Thanks for having us.
How's everything going?
Everything's going well. We're pretty happy with the team and what they're doing. I think before we get started, I do want to acknowledge, because we certainly have veterans in Robinson, want to really pay respect to our veterans and what's going on on Veterans Day. To all the veterans out there, we just want to pass that on.
I don't think we had any slides that we were going through. We were just going to launch into questions.
That's it.
You know, I'd like to maybe start off with a state of right type question. You know, the market's been a very dynamic one
Yeah
really, for the better part of the year. Kind of the absence of seasonality earlier in the year, front loading of inventories, tariff policy, government shutdown. Not to be lost on anyone, there certainly there's a very powerful idiosyncratic story taking place at C. H. Robinson. If you could frame just the state of right and how you see things today, it might be a good place to start.
Yeah, thanks, Dan. As you know, we do not spend a lot of time on the macros, but per the question, just first of all, truckload lower for longer. That is what we kind of see there. From an ocean perspective, you kind of called it out. There is dislocation happening, peaks where it is not supposed to be. We kind of see that happen. There is some uncertainty, of course. That is about kind of what we see on that. We try to design our system to say, listen, cannot control the macros, let us do what we can control and build that system, which I am sure we are going to talk about today. I do not know if you had anything else.
Nope, well said. I do not think there has been much of a marked change in the macro environments for either truckload or forwarding. I think Dave summarized it well.
I know this may sound a little repetitive, but there's still a lot of investors who probably don't fully appreciate the scale of the transformation that's taking place at C. H. Robinson. AI and automation have played a very pivotal role, the cost out piece, the ability to procure capacity more effectively, identify price more effectively, the opportunity to take that model and apply it to global forwarding. If you could just talk a little bit about the transformation that has occurred and what remains to be done.
Yeah, let me start, and then Damon, you can add in too, because you're a big part of this as well. Listen, Dan, you and I, we've talked before on this going back, spent a little bit over almost two and a half years since taking the seat here at Robinson. As we came in, we said back then, I needed to do a diagnosis of the company. And we did that. We did do it along the lines of what we call the four P's, like people, product, process, and portfolio. Just went deep in this company along that grain of those four P's. Came out of that diagnosis, and I use that terminology because it's kind of a lean terminology. Being somewhat of a lean practitioner, that's a big part of our transformation. You have to diagnose a problem before you can treat it.
We wanted to do that scan of the organization to figure out what was going on. What did we find? A number of different things that as we started the treatment, we said, hey, this company, which has really good bones overall, I mean, as a company that started the 3PL market, was used to winning way back in the day in driving that. It kind of started a lot of other companies in the industry. Robinson has a history of success. It certainly ran across some headwinds when it came to digital insurance, price transparency, and a number of other things that what the industry would have had to deal with, and certainly Robinson as well. It kind of lost its swagger. It lost its winning attitude.
As I came in, what I saw were headwinds on margins, a number of other things, activist investor that was in the company, and just a number of other things, unhealthy type of business environment. As we decided from a treatment perspective, this company was in need of an operating model. That operating model was based on lean principles. We started it in January of 2024. What was the intent? The intent was to shift this company to get back on rhythm and get into a cadence, drive innovation, drive speed, and drive problem solving throughout the organization. Some of the things that we saw is that we were a culture that admired problems. Now we do not admire problems. We solve problems. Part of this, I could talk very frank because that is part of who we are. We are pretty vulnerable at Robinson.
We put on the table what's broken so that we can fix it. This particular model was one that said, we're no longer going to prosecute the person. We're going to prosecute the business and the problem. That kind of led us really on a scale since then, along with our technology, which was really good, but this operating model allowed our technology to bloom and to open up. We went from kind of machine learning to generative AI and then to the advent of agentic, which we could talk about. Those three things, that technology, that operating model, and then our logisticians who are really, really good, the best in the world, that allowed us to kind of shift this company. Since then, it's been seven quarters, I think, of really good performance within the company. The company has gotten its swagger back.
It has gotten its winning nature back. Anyone right now, if you walk in there, they're not high-fiving or anything. They're still going to work because we've got a lot more work to do at Robinson. We feel really good about what we're building. I don't know if you want to add on that.
I'll just add a couple of comments. I think to your point about what's been underappreciated, I think the operating model has been underappreciated. I think certainly people have gravitated to the technology. And certainly the technology has been a critical element of our success. But the discipline, the rigor, the problem-solving capability that the operating model brings, I think, is underestimated for the folks that follow Robinson. I think there's some that get it, some that don't. Because the operating model makes the technology better. The technology makes the operating model better. It really drives the enterprise results. The other area I'd say, I think, is underappreciated is the sustainability of what we're doing. I think a lot of people question, can you guys continue to do what you're doing? Our answer is absolutely.
The fundamental changes we've made to the company, these aren't hatchet changes. These are very surgical, very purposeful. They have fundamentally changed the way we operate the business. When we get the question of, well, will this sustain when volume comes back to the system, the answer is it'll absolutely sustain because the processes themselves are fundamentally different. To me, those are the two items I'd highlight, Dan. One is just, I think the operating model is underappreciated. It's absolutely a critical element of what we're doing. I think the sustainability of what we're doing is probably also underappreciated. We feel very confident that if we were having this conversation two years from now, it'd be the same conversation on overperformance at Robinson versus our peers in the market.
Dan, the people appreciate it. They love it. They're having fun. And they're inventing because we fail fast, we break things, and we stand up and keep going. One thing that's for sure, in 12 months, we will create something that we don't know about yet. And that'll happen in Robinson. And that's just how the company moves at a speed that's different for this industry as well. And that's why we always say, hey, we were a company that was being disrupted, but then turned that into the disruptor. And we feel good about that. And we've got a lot more in store. And so Damon and I always say the next two years are going to be more exciting than the last two years. We fundamentally believe that, what we got going on.
You know, it occurred to me, Damon, you just mentioned that you would expect the benefits to be sustained going forward. In a stronger volume environment, a stronger longer-term price environment, you could make the argument that they should be expressed in a more substantive fashion because you get the leverage effect of volume and price over what you've done. You should be able to benefit to scale. I want to make sure that we address that and then maybe pivot into agentic and what the opportunity set is to go deeper in the organization, address more complicated opportunities.
No, you hit it. I think it's the operating leverage that we expect to demonstrate when volume returns to the system. If you think about the processes, we've fundamentally changed, many of them on the back of AI. Where the process used to be human-heavy, now it's human-light.
It is a technology-based process. Rather, let's just take requests for quotes. Rather we're getting 600,000 quotes or 6 million quotes, I don't need any more people. The technology is all that needs to scale. To scale the technology, it's tokens. That is why we feel so confident in our ability to roll this model out regardless of market cycle. The strategy we have in place, the fundamental changes we've made to the company, we believe they will be successful in every phase of the market.
We've really looked at this at a quote-to-cash process, as we talked about before. That quote-to-cash process, for anyone that's in this industry, knows that that's somewhat of a physical process. I mean, it's handoffs. It's manual. We've really attacked that with this technology. For that, certainly, as Damon said, those manual tasks, we're eliminating, but we're shifting as well. We're investing in small, medium business. We're going to customer-facing. A lot of our hiring we're investing in is kind of shifting to the right. Some of the operational things, as you know, that's where productivity came 40% since the end of 2022. That's something, being asset-light, that will continue to drive. We committed to single-digit productivity no matter the market, and double-digit at times where we have a wave of technology. We had double digits committed to or single digits next year.
We switched that to double-digit productivity improvements with the things we expect to come out of global forwarding as well. This is just how the company works and how we create.
Maybe that's an opportunity to touch on agentic. You talk about global forwarding. Global forwarding is not as standardized as truck brokerage. You're dealing with different currencies. You're dealing with duties, tariffs, different countries, different policies, different procedures. A lot to process. Can you talk about the ability to leverage automation into global forwarding?
Yeah, we can. We both will. First, start off by saying that business has done a really nice job. Going back to the importance of the operating model, you're talking about a business last year that grew every quarter, but also dropped its expenses. That was just on the backs of the physical operating model. That business goes through the same process that NAST goes through on our review set. It has just fundamentally gotten better from a discipline perspective. Now we've leaned in heavy with our technology stack on NAST, typically on generative AI. As you said, the difference between the two, you're talking about on-system data. In truckload, it moves really, really quickly, especially at our scale. This was a perfect application for generative AI.
That is why you're seeing a lot of the rewards that we're seeing within NAST. Global forwarding, as you say, you go from China to North Carolina, there is a lot of sausage making that has to happen behind the scenes to do that. What's included in that is data that's off-system versus on-system. This is where the advent of agentic technology works. A lot more reasoning. It works with data off-system, and it learns. If you think about it, it takes several days for you to put together a quote, potentially, to go from a load from China to, say, North Carolina. It takes several days. There is a reason for that. You start doing this technology, and you can get down into hours to do that, which is why we feel really good about where we're going with this technology.
By the way, it's not just for global forwarding. That agentic technology will also come back and help NAST as well as it comes back.
Yeah, great clarification because we get the question a lot. Is your NAST business in the later innings of technology deployment? The answer is absolutely not. We're in the early innings of our technology deployment at NAST. As Dave mentioned, the entire universe of opportunity is that quote-to-cash cycle. I mean, if you think about that, that's hundreds of sub-processes that are target-rich for automation. And we've only touched a fraction of those for our NAST business today. So we're still in the very early innings for NAST. On global forwarding, we call it inning one because up until this point, most of our technology has been over-indexed to NAST. That was done intentionally. That's where we could prove out proof of concept. We could get scale. We could get benefit quicker. Now we're to the point where we're going to move that indexing over to global forwarding.
As Dave mentioned, we think agentic is perfect for global forwarding because just kind of the quick assessment of gen versus agentic. For gen AI, it works really well when you're dealing with simple processes. Now, the volume can be extreme. As Dave mentioned, it can be hundreds of thousands of processes or hundreds of thousands of transactions, millions of transactions, but they're simple. One-to-one, simple processes, high volume. Gen works perfect for that. It also works perfect for that if your data is on-system. For us, our book of record is Navisphere. The majority of our data in the NAST business is on Navisphere. Therefore, target-perfect environment for gen AI. You move over to global forwarding where not all of our data is on-system. A lot of our data is in, call it SharePoint sites, spreadsheets, and maybe in customer websites, less structured than NAST.
A portion of our data is off-system. Gen AI can struggle with that. Agentic AI has higher levels of reasoning. It can look at multiple data sources, whether beyond-system or off-system, and produce a much more efficient product than gen AI can. As Dave mentioned, the physical attributes of global forwarding where you may have a many-to-many relationship or a many-to-one or one-to-many, it is just more complicated on the global forwarding side than the NAST side. Again, agentic works really well because that requires a higher degree of reasoning that gen AI does not have the great capability of doing. We are really excited about it. If you think about all the benefits we have realized on the NAST side of the business due to the technology, and now we are starting to index that over to global forwarding.
We do believe that'll be over-indexed to the second half of 2026. There's like a 12-18 month cycle time when you start, go from concept to full-scale operation on the AI technology. We think we'll certainly be seeing benefits of agentic and global forwarding for the first half of 2026, but it'll be over-indexed to the second half of 2026.
Dan, just to put a point on this, on how we started this conversation, I can't emphasize more the importance of how we're running the company from that operating model. Everything Damon just said, how we got to agentic AI, that really came out of our operating model. That came out of our reviews in which we are saying, hey, we have to reach a certain point. In our discovery, our five whys, as we're going deep, our technology team said, hey, where we are, we can't get there. We think the only way to get there is this new technology in agentic AI. That started a very fast process in which we all looked at that, made a high judgment call. Next thing you know, we're building these agents as we're moving forward. That typically would not have happened. That's not easy to do.
It shows the power of why you have that discipline operating model because it drives discovery and innovation. That's super important for where we're going.
There's another theme here that I do not think gets talked about enough. It is this notion that there is a consolidation story, both in terms of domestic freight brokerage as well as global forwarding. Brokerage is a highly fragmented business. Most brokers do not have, no broker has your data set. Very few, if any, have your tech stack or anything remotely close to it. The ability to leverage those things into a market that cannot respond the way you can to develop the tools, the cost out, the efficiencies, it presents the opportunity to consolidate in a way that these tools as technology, I mean, it forged you that opportunity in a way that never has been present in the past. I think that is in front of us. I think what has been occurring in your business is what you have delivered is very much a C. H. story.
What's in front of you also, though, is very much, I think, a consolidation story. Could you speak to your views around that, whether or not you think that's a prescient theme, whether or not there's substance to that?
Yeah, I think certainly any marketplace where you've got the fragmentation and the long tail like we have in brokerage, I think the thesis is always consolidation makes sense. I would argue that consolidation is happening now. You have brokers leaving the system every day. They may not be the brokers that make the headlines, but there's tens of thousands of brokers, and they're exiting the system every day. Some of that consolidation on, say, the tail of the industry is already happening.
Now, to your point, and again, we try to articulate this as best we can, the way we're using AI and the way we're deploying AI, we believe is quite different than most companies. We're not buying our solutions off the shelf. We're not using third-party vendors. We're not using consulting companies to help us integrate it. We have 450 engineers that are C. H. Robinson engineers. They've grown up in the company. They built Navisphere. Now they're building our AI agents for very bespoke, customized solutions for C. H. Robinson. Why is that a benefit? Benefit number one is it's a custom solution for C. H. Robinson. It's not a generic solution that we have to make fit our business model. It's a solution that we come up with an idea to solve a problem and generate a return.
We build an agent custom fit to that solution set. That is benefit number one. Benefit number two is we built Navisphere. Our transportation ERP system is already custom to Robinson. We control the code. We do not have to rely on a third-party vendor to let us augment that ERP. Therefore, our ability to deploy this technology, I would say, is substantially faster than anybody that is using third-party capability where they are having to negotiate timelines. We do not have to do that. We own the code. We can put AI agents on top of Navisphere. We control that timeline. Lastly, I would say, is our cost advantage because we create our own tech. We create our own agents. Therefore, once we have created an agent, the marginal cost of managing that agent is close to zero. It is just tokens. We are not paying by the drink.
We're not paying a third-party vendor for fees. I mean, it is a very attractive cost curve. To your point, I think AI for smaller brokers could actually become cost prohibitive. They're not going to be able to develop their own tech. They're going to have to feel like they're going to buy it off the shelf to be competitive. I actually think in most cases in that scenario, it's not going to drive efficiency. It's just going to add cost to their already thin margins. I think there is an opportunity as we go through this journey where AI may drive a further consolidation in the industry just because those that have command of the tech that can drive that leverage and that competitive advantage. Those that don't have it, I think they become part of that consolidation. That's right.
Dan, I think you and this audience would also appreciate it's also speed. It's also these engineers know freight. They grew up with it. Knowing freight and growing up with it is a big difference than I'm going to learn freight and do it. Time is money in this industry. We feel like that's a competitive advantage.
Right. I have one more question, then I'll turn it over to the audience to the extent that they have any. Growth versus margin improvement.
Yeah, both.
High quality opportunities. You have these choices in front of you. You guys have realized a tremendous amount of margin improvement. It affords you the ability to take your, I do not mean to say eye off that, but to hold that in a steady state and really begin to leverage what you have done to begin to take share and go after growth in a way that you maybe have not in previous cycles, which I think for better or for worse has been a struggle for C. H. through the last cycle and maybe even the last two cycles. I think you are very well positioned to demonstrate a different growth trajectory this next cycle.
Talk to us for just a minute about how you plan to manage the balance between those two, recognizing you've achieved essentially 40% margins at this point in NAST, but there's still room to go, not lost on me. Could you just frame that?
Yeah, I'll start. Damon will finish, and then we'll take questions if they have them. Number one, this is our strategy that we started with. We said, hey, we're going to do two things that are critical. One, we're going to take market share, and we're going to expand margins. This is not something new for us. We were building a process and a system to do those two things. I think we're demonstrating that we can do that in a really, really tough environment. There are handoffs that we do every day when it comes to that. Let's hit on your margin percentage. Why don't you just double-click on that and we'll go there.
Yeah. I think all of you are aware. We had our investor day in December of 2024. We put out what was deemed pretty aggressive targets, both on margin and results. We just came out with an update to that investor day target set in our Q3 earnings. You will notice it was not a typo by Chuck. We did not change our margin targets. We left NAST at 40%, and we left global forwarding at 30%. Those are what we have called mid-cycle margin targets. As you just referenced, we are close to those now. The reason we did not update those targets is the optionality. I think historically, there was a misnomer in this industry where you could either grow or expand margins. You could not do both. As Dave mentioned, we set a strategy that said we have to do both, and we are doing both.
Our operating model, our technology, our leadership team, our people, we're doing both, and we have been doing both now for several quarters. There becomes a time where when you're already demonstrating industry-leading profitability, at some point in time, the best decision for earnings is to invest some of that margin back into growth. Once we establish with sustainability those margin targets, 40% for NAST, 30% for global forwarding, we reserve the optionality in any given quarter that we may grow margin to 42%, or we may not. We may take that 200 basis points and go take demonstrable profitable share. Different mindset than we've had historically. There may be quarters where the right decision is to grow margin because the quality of the incremental growth that's available just isn't there. We're not going to chase volume. We're not going to take bad freight.
We believe that's the right equation going forward. Now, over the long haul, we'll do both. If you think about over years, we'll continue to expand operating margins. We'll continue to outgrow the market. We'll continue to grow on an absolute basis. We just want to make sure we don't put ourselves in a box on the quarter-to-quarter basis where we can make the right decision for investors and not suboptimize that decision because we're trying to get to a new margin target.
That was part of the strategy of operating leverage that we always talked about. We will turn it over to you on questions.
Any questions?
Just a quick question as it relates to the vendors of the technology. To the extent that you're really productive and you are expanding margins, are the contract structure where they might try and raise the price to extract some of the value that you're getting, or are these kind of SaaS-type contracts where they just share?
It is a great question. It is actually one of the competitive advantages we love about how we are using AI. If you think about the AI ecosystem, you have got the hyperscalers and the chip guys on the left. You have got the data center guys in the middle. You have got companies that are using AI to drive business results. We think we are one of few that are on that far right that is actually delivering revenue growth and margin expansion by using AI. To your point, we are using LLM models, and we use them all. Therefore, from our perspective, we are in the perfect spot from a cost curve perspective. One of the themes you hear is, the LLMs are already becoming commoditized. That is great for us because we use them all.
The more they become commoditized, the cheaper they become for Robinson, the less costly they are for me to deploy my technology. For us, we actually do not see any cost pressure at all. In fact, we see cost deflation as it relates to the use of those LLMs and the tokens. One stat, over the last 12 months, our use of tokens has gone up over 250%. The cost of those tokens has come down between 25%-30%. We feel like we are in the sweet spot as it relates to the cost of deploying our technology for the business results that we are getting.
That's why we say it's an undervalued industrial AI play because of the results we're getting and because of that question. That's an important question from a cost perspective that many don't realize.
Anybody else have time for one more question? Gentlemen, thank you for being here. Appreciate it.
Thank you. Thank you, Dan.
Yeah, appreciate it. Look forward to catching up with you in the fourth quarter.
You got it.
Sounds good. Thanks.
Happy holidays.