Chris Caplice:
0:35
Here's the thing. So you talk about these tariffs and things going on, if you talk to a supply chain professional, they'll say, Well, it's just another Tuesday, to be honest. So there's stuff happening, right? And so, but look at what we did the last five years, right? That we shifted from West Coast to East Coast arrivals for a period of a couple months, things shut down. And, you know, we absorb and I think some of those muscles that we trained are still here. Supply chains thrive on being able to handle these things. I'd like to think of them as shock absorbers. So the supply chain is able to be the shock absorber, right? Of all these disruptions to keep the ride the business operating smoothly.
Blythe Brumleve:
1:18
You spent weeks planning your transportation procurement strategy, but one rejected tender throws it all off for paper manufacturers, the execution phase is where the money's made or lost. Chris Caplice Or Yes, Chris Caplice, from DAT, is here to talk about why your routing guide may be costing you 1000s, and what dynamic tendering can do to fix it. So welcome into another episode of everything is logistics, a podcast for the thinkers in freight. I am your host, Blythe Milligan, and we are proudly presented by SPI logistics. And for this episode, we are presented by DAT as well. So Chris, I've been listening to a bunch of your podcasts, getting, you know, prepped for this call, and so I'm super excited to have this conversation with you. So welcome to the show.
Unknown:
1:59
Thanks, Blythe, I'm glad to be here. I'm sorry you have to listen to some of my podcast, but not as nearly as professional as you. Oh, well, thank you
Blythe Brumleve:
2:07
so much. That's No No. It wasn't. It wasn't. It was actually very not, actually, but it was very insightful, especially one of the episodes that I was listening to that the the host, I think the conversation came from a show on from easy post, and she said that you were the most popular episode of 2024 and that's why she invited you back for for 2025
Unknown:
2:29
so that was a fun conversation. They're good, good people, absolutely. Now,
Blythe Brumleve:
2:33
now for folks who may not be familiar with your work, can you explain a little bit of your background and your roles of what you do for a DAT and also MIT? Yeah,
Unknown:
2:41
sure, sure. So I'm was actually a civil engineer in undergrad and went in the Army, came out, got a master's in transportation, and decided I didn't want to work in for the government building roads, so I came and got a PhD up here at MIT and focused in on the use of roads, mainly truck transportation. Spent about eight years in software development, developing tools to help shippers procure transportation. And that kind of naturally morphed into working with then chainalytics That became and was acquired by d a t almost five years ago, pretty much to the month right in the pandemic, a great time to be acquired right as a pandemic kicks off. And so what I do at d a t is I'm working on the shipper side. Because this, if some of your listeners don't know about DAT, it's the largest freight analytics company out there. We work with shippers, carriers, brokers, and we're the main source of information on pricing and other data capacity, data load to truck ratio, to help all the players, all three stakeholders in the industry, do a better job making the market more efficient and more effective. So they brought chainalytics in, which is a part that I was worked with, and to kind of finish off the trifecta. So working on the shipper side, in addition to that, I've been up at MIT for a little over 20 years, running part of the Center for Transportation and Logistics. We work with companies do research, and we have educational programs. I created something called freight lab up here, which looks at pretty much the same thing I do with D, A, T, how can shippers cares and brokers work together better? And so the stuff I do academically with MIT, and the stuff I do in practice with Da t really merged. The Venn diagram is pretty full there, but that's kind of how I my path to the freight transportation market. And I think you just celebrated what 50 years a 50 year anniversary for for the Center for Transportation and Logistics, is that correct?
Chris Caplice:
4:33
Yes. So CTL started it 52 years now, I think it started when everything was regulated. So the deregulation, the motor carrier act of 1980 the staggers act, all those things before that happened, people didn't know what would happen. And so there was a lot of analysis and studying going on to say, Okay, what's going to happen when we deregulate? And, you know, so is, that's really the kernel of where it started. But since then, we've grown. We have about 150 researchers staff here, about 80 to 100 students a year. We work with over 50 to 100 companies a year. We really try to come up with innovation and drive them into practice. That's, that's kind of what we do up here.
Unknown:
5:14
And do you find that, especially with, you know, since 2020 you kind of brought it up, that supply chain kind of took center stage, at least in the media aspect, is what kind of it felt like to me? Has that sort of been seen in in student interaction? Or student enrollment with your role.
Chris Caplice:
5:32
I think the biggest thing is that we don't have to explain what supply chain is at Thanksgiving anymore to relatives, because everyone knows what it is. It's the thing that stopped me from getting x during the pandemic, but the thing that people don't realize, or they don't, if they don't, they don't think about it as much, the supply chains never stopped. We never no one starved. Stuff came in. Toilet paper. The main problem there. People were hoarding. But stuff kept coming. The supply chains never shut down. I think people have awareness of that and appreciation of that, especially post pandemic, we're seeing that there, there was a huge surge of people coming into the industry, and I think that's still happening. When I was going into college in the 80s, it didn't exist. The phrase didn't exist. There's logistics wasn't even really much of a practice. So it's really changed since then. Most people my age came into the industry different ways, but now we're seeing more people coming in. They do a supply chain undergrad, or logistics undergrad, and then they continue on. So that is changing, but I think the awareness of what supply chains are has definitely increased post pandemic. Yeah,
Unknown:
6:37
I definitely get a lot of those questions around the holidays of what's going on at the ports. And I'm like, I don't even know what a port is. My grandmother's friends are, you know, talking to me about this. And so, yeah, I definitely echo that experience where, you know, used to have to explain it. Now they kind of understand it a little bit more, but not, you know, there's still some education on my end that I have to do it, at least when it comes to family and now,
Chris Caplice:
7:00
personally responsible now for the supply chain breaking, because it's, it's kind of like oxygen. You don't notice it until it's gone. That's what, that's what people have recognized. So supply chain was a big issue when it was hard to get stuff now, when it works smoothly, you should be invisible, right? And that's, that's where we kind of are now in this, this stage of the of the market. Yes,
Unknown:
7:20
and you, actually, you recently penned an article moving into sort of the, I guess, the the paper manufacturer focus for for this conversation, because I really want to get a ground level of what is currently being experienced by these manufacturers, and then how they're addressing, you know, sort of global supply chain issues, geopolitical issues, obviously, the T word tariffs, which, you know, we've been learning about ad nauseam over the last couple of months. But when you recently penned an article talking about when tendering loads don't go chasing waterfalls, which I now have that song stuck in my head. It's been stuck in my head all day. But you open the article with transportation. Procurement has two phases, or moments of truth, a strategic phase where shippers develop a long term plan and select a group of carriers to handle the freight for that period, or and an execution phase, where loads are actually tendered. Walk me through that first phase of what you're talking about, what are sort of the table stakes for today?
Chris Caplice:
8:18
Yes, I stole that phrase from AJ laughley, the former CEO of PNG, the two moments of truth when he talked about it. The two moments of truth for a CPG company is on the aisle, someone grabbing it, and the second is, is it good when you use it? The two moments for transportation, as opposed to most other commodities or other services or products that are bought, there's also two moments. There's a moment where an RFP, an annual bid, is done, and I'll talk about that in a second. But that just is the promise of volume and capacity if it materializes. The second moment of truth is when they an actual load materializes, you have to assign it to a carrier. So two moments of truth. And historically, since probably mid 1990s once deregulation kind of sank in, the first moment of truth was done by an annual request for proposal or bid. Some companies don't like the word bid, but it really is. It's where the shipper says, Okay, here's what I expect the volume to be on every one of my lanes, which could be five to 20,000 lanes, five to five digit zips. And they give it to the carriers, and the carriers come back with a rate on the lanes they're interested in. The shipper does some advanced optimization, finding the best mix or assignment or allocation of carriers to lanes. This is where I did my dissertation on back in the mid 90s, and then they assign it out to the carriers, and that feeds into what's known as a routing guide. And there's been a ton of work in Optimation, in operations research and different techniques to come up with the perfect plan, right? To come up with the ideal plan, use scenarios. You can favor incumbency. You can restrict it so that your minority carriers get a certain amount of business. Your carrier base isn't so big, so it's very customizable, and you can have 1000s of scenarios that you run. But at the end of the day, all you're doing is creating the plan if the volume actually materializes. So that's the first moment of truth.
Unknown:
11:52
And then the second moment of truth is that that actual execution phase. Can you walk us through that? Yeah,
Chris Caplice:
11:58
yeah. So, so the the first phase ends, essentially when the winners of the auction. Right on each lane are fed into the routing guide. And the routing guide sits within the Transportation Management System, or TMS, and what it's there, it's just a catalog, right? So when a load comes in, it's going from Chicago to Des Moines, it goes in and says, Okay, which, which carries has Chicago to Des Moines, and they it goes straight through. And if it's a good TMS, it'll tender automatically to that primary carrier. And if that, the weird thing about truckload because I'm mainly talking truckload transportation here, is that the contracts that were set up in that annual RFP, they're binding in price, right? If I if the carrier hauls the load, they have to honor the price that they committed to in the bid. However, the shipper is not committing that the volume will ever materialize, even if they said, you know, five a week. They're not guaranteeing that. And the carrier is not guaranteeing that they will, 100% accept every load, just because of the dynamic nature. They might not have a truck in that area. They might for some other reason, it might be that the shipper is giving more than they than they said they would to the care, and they just don't have the capacity, so you've got this loose contract. And so the load comes in, goes to the routing guide, gets assigned to the primary carrier. They reject, if they take it, that's fine. That's great. And it goes and that can be depending on the market, 70% of the loads get accepted by the primary 80% even if they reject it, then it goes through what's called a routing guide waterfall, and the waterfall is simply made up of the carriers who bid in the RFP who didn't win, right? And so their rates are going to be higher than the winning rate. And so you go to the first alternate and offer it to them, and if they take it, great. If not, go to the third, the fourth, the fifth that's the traditional way that a routing guide works. So that second moment of the truth is actually finding a care to match to that load right now, not the plan of future business that's done in the first moment of truth. Does that
Unknown:
13:59
make sense? It does. And you actually help me understand that that aspect, because I, you know, I come from an asset based three PL but I was just an executive assistant, and so I never actually booked the freight myself, so that that is hell. I've heard these phrases for years, but now to fully kind of grasp of that entire process, I appreciate that, that breakdown, that's the MIT education, I think, coming out of you. Now, I would imagine that you don't want, you know, the majority of your shipments to reach that, you know, don't go chasing waterfalls, you know. To go back to the the article title, what are some of, I guess, the with everything that's going on in the world, geopolitical tensions and tariffs and all, I would imagine that for a lot of these shipments, are they falling within that 80% or are they falling within that 20%
Chris Caplice:
14:46
it's a good question. And so, you know, so the two moments of truth, right, the RFP and then the tender, the RFPs, are a pain. No one likes them, right? And so they take a lot of work from the shipper. The carriers never given enough time the data no one believes because you're asking the shipper to forecast for 52 weeks volume on each lane. It's ridiculous. However, it tends to work. So how do you can't it's hard to change something that tends to work. 90% of most truckload moves go under contract. So it kind of works until it doesn't. And so what typically happened that waterfall used to work pretty well. And the reason why we did it that way, and most TMS is developed in the 90s, 2000s and 2010s It was simple, right? And so all the all the expertise, all the sophistication, was in the strategic planning that RFP process I talked about with it does the perfect assignment. The execution systems were pretty stupid, right? They just said, Here, match this. It was all about execution speed. Well, now technology has changed. You can have be a little smarter down there, and so you can have a lot more sophistication in real time execution, which is a relatively new capability. So the waterfall method was a great way. It worked most of the times until it didn't. And during the pandemic, it really failed. And failed for two reasons. One is it takes long. It takes too long to do it. You know, every time you go to the next one, you got to give them time to respond or not, and so you can eat up a day, two days, three days, so it gets very stale and your lead time keeps shrinking. The second thing is, they found after the second or third, especially during the pandemic, why bother deeper? No one's going to take it. No one's going to take it. So you need to go out to what's known as the spot market. In the spot market, as opposed to a contract market with RFPs is it's a rate for this lane, this load on this lane at this time, with all the details. When I do an RFP, you have a projection of potential future business for spot market. It's right now. And so it used to be using the spot market. You know, spot was a four letter word to most shippers. You wanted to stay in the contract because you want rate visibility and stability and all that. And so, because of routing. I The waterfall method took so long, and it really wasn't working that well. That's what I meant, chasing waterfalls. And I had to name it that because the song, I mean, come on,
Unknown:
17:14
which I'm still seeing in the back of my head right now. Now, there was another article you penned as well, and the title was, the RFP is dead, long live, the RFP. And that kind of sounds like how we're, you know, we're, we're trying to balance that fine line of the RFP. Is that accurate? Yeah, it's
Chris Caplice:
17:30
a pain. Anyone who's been involved with it. It takes some shippers, will take up to six months to get their data set and everything, give it to the carriers. Carriers get a week to respond back, and then they spend two to three months analyzing, running all these scenarios. So it could be that the rates were submitted and they don't get enacted for six months. So it just, it's so hard to do this, but it works. It tends to work. 90% of all shipments, truckload generally go by contract, as opposed to a dynamic price. But one of the things that's come up from this, it's really kind of interesting is that, you know, using contracts, it's kind of a one size fits all, where you set the contract for every rate. And when I was running bids back in the 90s and early 2000s we'd work with shippers, and they might have one load on a lane last year, and most shippers will take what happened last year, project that to the next year and modify it slightly, you know, as as they need to. But they would keep every lane that had any volume and bid it out, and they would round it up to one a week, right? Because you'd show average weekly volume. And so what would happen is you'd have tremendous number of lanes that get bid out, and unfortunately, a majority of the lanes, not the volume is gets ghosted, which means the shipper creates it, puts it out to bid. Carrier analyzes bids on it, shipper assigns it to them, and then, over the course of a year, nothing happens. No volume goes on. It's called ghosting and so we've been analyzing this with a bunch of different shippers, and it can be up to 50, 60% of the lanes in an RFP. And so as you look deeper, and when you look at the network of a shipper, especially a paper manufacturer, anyone who has heavy transportation costs as a percentage of total costs, they tend to be very dispersed. And so Pareto, the 8020 rule really works. And so 80% of the volume is usually on just 15% of the lanes. The other 20% of the volume is on the 75 or 85% of the lanes. So you have high volume lanes, and then you have a long tail. And the the thing that we've come up with, or suddenly the realization is you don't need a contract for everything. So those ones that are really long tail onesie twosies really not consistent freight, then treat those differently. And that's one of the things I brought up in the waterfalls argument, because most the lanes that fail, like I described and have to go down the waterfall, you can identify ahead of time. They're low volume lanes. And you know, you only do it three times a year. And so, you know, it's going to fail. So why even go through the process of bidding and just treat it dynamically? So you want to set up some kind of other mechanism to assign a carrier to that you might not know the price right away, but to be honest, it was going to go to the spot market anyway, because it was going to fail the routing guide.
Unknown:
20:24
And so for a lot of these, you know, sort of low volume lanes of what you're talking about, I'm curious, is this sort of a new data point that that you guys are analyzing and looking at over the last say, you know, you've been at, you know, MIT, I think, for, you know, 20 years. You just said, how has that data collection evolved? Is this a sort of a new data point for you guys to monitor. I'm curious if there's other data points that maybe have, you know, come up or popped up onto the scene within the last handful of years? Yeah,
Chris Caplice:
20:53
I think everyone's kind of recognized that, you know, there are high volume and low volume lanes. I don't think people looked as much as the impact, because there's a there's a strategy in procurement, which is, you get a contract for everything, right? You because then it's like, you know, Linus is safety blanket. You've got the blanket there. I've got a contract on that lane. So it provides false certainty, and it's there. But if you're a CFO or a chief purchasing officer, that's good. You have a contract and everything, because the last thing you want to do is have spot rates, because then how do you budget for that? So there's been a fear or reluctance to allow dynamic kind of planning. There people try to put certainty on it as much as possible. The problem is it doesn't work. It isn't effective for that. And so there's been more of an acceptance to use spot as a strategy, not for everything, but for a certain segment. And thinking of procurement, more on a portfolio where you've got, like, dedicated lanes with a private fleet, where it's high volume, consistent and balanced, right? You have out and back and things like that. Then you have con. Lanes which are relatively high volume, and we can talk about what that what that means. And then you have the lanes that are very sporadic, inconsistent, and those should be treated differently. And you should do those dynamically. But one of the points is, if people look at volume, but in our analysis, it's actually different, it's consistency matters more than volume. And let me give you a simple example. Everyone has a lane where you might have 100 loads per year, but if it happens in a two week period, that's not a consistent lane, right? You want to have a little bit of volume every week, every month. So what we found is, you'd think those would be correlated, but in our analysis, they're only about a 53 to 55% correlation. So just using volume by itself isn't really a good proxy, because you get some of these wacky lanes, especially if you're an ag, where it'll just during the season, your volume just goes through the roof and then it goes away. So looking at consistency, which is the number of weeks per year that that lane has volume is a real good proxy for whether you should do a contract on it, or whether you should let it go to dynamic freight. Now,
Unknown:
23:12
zeroing in on sort of the paper manufacturers for a moment. Now, there's a lot of talk, especially in the news around obviously manufacturing, bringing manufacturing back. I'm curious what the overall landscape looks like for paper manufacturers. I understand a lot of them are located in the southeast. I'm based in Jacksonville, Florida. We were kind of known for, you know, a few decades about, don't go to that town because it stinks, because we had a, you know, a bunch of paper mills around. Some of those have have shut down, especially in the the downtown area. But still, the southeast is, is very, you know, paper manufacturer heavy. I'm curious, is that sort of, I guess, a victim of offshoring or near shoring or is it primarily staying in the United States for paper manufacturers?
Chris Caplice:
23:57
Paper manufacturing is predominantly domestic, where we had a lot of exports of waste paper, and that's probably, I honestly don't know how it's being affected right now with the tariffs. I know so for a while there, China was refusing, but most of the domestic because it's such a heavy product, it's being produced domestically. The reason why a lot of mills have gone down is there's been consolidation of the paper packaging industry. They were way overcapacitated For the 80s, 90s and 2000s My brother used to work in the industry, and you've just seen a lot of consolidation, more efficient and effective use of their mills. But the thing that makes paper and packaging so interesting is that it's such a heavy product, and it's relatively low value, right? You're not going to ship a million dollars worth of product in a in a trailer. It's something where the cost of the transportation is very high as a percentage of total cost of goods. And whenever that's the case, then you have to pay especially close attention to your transportation. It just, it just makes sense. It's such a big piece of the pie. If I'm shipping, say, iPhones by truck, right? I wouldn't care as much, right? Because the value of an iPhone, you know, I'd be worried about theft at that point, a truck full of iPhones. But for paper, you you have to be very cognizant of every penny for transportation. That's why it's very important for them to understand where the market's going, where rates are, and what's the best way to secure capacity for the lanes that they have freight. And
Unknown:
25:27
are you seeing those manufacturers sort of lock in those, you know, RFPs early on? Is it this a once a year type, you know, bid process, or is it much more, probably steady freight, I would imagine? Well, it's, it's changed,
Chris Caplice:
25:42
right? It used to be, pre pandemic. Everyone would generally run an annual bid, and then, as things happen, plants close, new customers come online, you run what's known as a mini bid, where it might be okay, these 20 lanes that are for a new customer. You bid them out, and it goes quick. During the pandemic, everything went to hell, right? You have CPG manufacturers suddenly created more manufacturing locations, ship care, excuse me, retailers were requesting, you know, high otif On time in full and so they didn't have enough product to send out right away. So if you talk to some of the CPG manufacturers, they were taking product from the line onto a truck, no intermediate RDC or anything. So there was a lot of shipment of partial loads, so you saw a ton of inefficient shipping just to get the product out, because you'd have such change shifts in demand. And then the production facilities would open, close, and packaging, especially the packaging side of paper, and packaging is really a derived demand, right? They are. Their demand for packaging is contingent on the product being sold. It's not that no one buys packaging. Just to get packaging, you get it for your product like transportation. Transportation is a drive demand. And so you you have to to see what your demand is going to be. You have to forecast your next customers. And your customers, customers demand, which is very hard. So the paper and packaging have it especially hard because one, it's such a high percentage of their costs, and two, it's derived, so it's highly volatile on things that they can't control.
Unknown:
27:19
And I know that some folks, they will follow what's called the cardboard index, where if you know, if you are purchase, if there's certain amount of sales for cardboard, then things are looking good, then demand is high, and then if it kind of falls off of a cliff, then demand is low. And you can kind of, I guess, judge the economy based on, you know, how much cardboard is being shipped out. Have you heard of that index before? I'm sure. Yeah.
Chris Caplice:
27:39
I know. I know of it. I don't follow it as much. I think it, it's something does good indicator. I know that like purchasing fuel, there is an index for how people will buy product and what their prices will be, because as a commodity, you usually have an index to something, and so in trucking, you have the same thing for fuel, where you peg it to what the DOE department of energy releases every Monday, and you kind of share your surcharge there, because for any kind of procurement of a commodity, you always want to separate out the things that are controllable and the things that are not. And for trucking, fuel is uncontrollable. No one controls the price of fuel. It moves on macroeconomic factors for paper and product, pulp is about the same thing. So you look at that pulp index, and that can actually give you a good indicator along those lines. Oh, that's
Unknown:
28:30
interesting. So instead of the cardboard index, we should be watching the pulp. There's there's different
Chris Caplice:
28:34
indexes here. I'd have to defer to my brother, who knows this industry better, but it's the same concept. It's the same concept, something that you cannot control. You index, and you try to share the risk, or balance the risk between buyer and seller. And
Unknown:
28:47
so for a lot of these manufacturers, of what you're talking about, that I would imagine that they would want to, well, probably I would imagine, for a majority manufacturers, they want to keep their freight in a contract system. But are you seeing for paper manufacturers, are you seeing that go more to the spot market, or, you know, some of the waterfall carriers, what like what you mentioned earlier.
Chris Caplice:
29:06
I think if people had their druthers, they would go contract for everything if they could. But it really depends on one of the hallmarks of the truckload industry is that it's not stable. Because you have over 200,000 carriers. And I don't think we've talked about this, one of the statistics is that 96% or 94% of the truckload carriers, truckloads, point to point, not the LTL guys, the point to point full truckload. They 96% have less than 20 trucks. So you say, Oh, they're really small. What people don't do is look at the other way. 4% of all the trucks have about two thirds of all the truck capacity. So the market is, yes, it has a long tail, but it's pretty consolidated at one end. And so because you have such a long tail, and the barrier of entry and exit is very low for truckload, all you need is a CDL, a truck and a smartphone at this point. So by because these carriers can enter and exit at at will, pretty much, and some of them will enter the market, and then the market gets soft for them. They'll park it and they'll go work construction, or they'll go fishing, or they'll do something else, but the capacity can come right back in. So because of that, the truckload market, it works in cycles, and we saw the biggest cycle was the pandemic cycle that really took off in q2 of 2020 and it ended in q2 of 2022 so two years we had these ridiculously high rates where spot was higher than contract, and that's when every shipper wanted to go to contract, right? Because he didn't want to pay those spot prices. But then, since the summer of 2022 it dropped, and it's only now recovering. In december 2024 it finally approached the the levels that it was pre pandemic in the first quarter of 2020 so that's been three years it took. So two years up in a cycle, three years down, crawling back. And since then, what's actually been happening? We've been flat. So when you ask to do, do shippers prefer contract or spot? The answer is, it depends. If the market is tight, which means con spot rates are higher than contract, then yes, they want contract. When spot drip dips below contract, which it has been for last three years, then they'll say, Well, maybe spots not so bad, right? And if you're a heavy like a paper and product packaging manufacturer, you're that's a big chunk of change, so you'll tend to use the Spot more. And so what, what you see now is the relationship between the shippers and the carriers. Do you honor the contract? You know, on which side we did a paper up here, gosh, about four years ago, and it posed the question, are carriers elephants or goldfish? And so the idea is, and this was pre pandemic, so it's looking to the last cycle before the pandemic. Was a tight market, 2017 2018 and it crashed in 20. 19, right? And we posed the question, if a shipper treated you poorly during a soft market, how did you treat them during a tight market? Are you an elephant, which you remember those who did and you treat them accordingly, or are you a goldfish who has no memory, right? And so what are your thoughts? What do you what do you think are carriers elephants
Unknown:
32:25
or goldfish, depends on how the money is. I guess they are
Chris Caplice:
32:29
total goldfish total, because they can't afford not to be. And so we're in the middle of doing this process, the same project, but for the other side, our shippers, elephants or goldfish, my colleague, Dr Angie acacella, is working on this, and we'll probably have something over the summer, but our hypothesis is shippers are total elephants. They remember, if a carrier screwed them over in their opinion, or didn't provide capacity, they're not going to get the business in the bid. So it's kind of an interesting dynamic. And this kind of goes to, you know, what do shippers prefer? Depends where the market is right, where I stand depends on where I sit. So, but we do see in better shippers, right? That it's not like everyone bottom feeds goes trough to trough. So what you do if you do an RFP, and let's say, you say, I think there's going to be five loads a week on this lane, and if there are five loads a week, you go contract. But if there's a six load, right? You send that to spot, you don't give it back to that contract carrier. So we're seeing better shippers will adhere to that contract. And the same thing happens in a tight market. If you have those five loads and a six load comes, the carrier won't say, Oh, you send that to spot. They will, you know, they will try to capture it as spot, because they will cover what they promised, but above that, they'll treat differently. So I think both sides are starting to do that and but it really depends on the state of the market. Where it is
Unknown:
33:57
is that where dynamic tendering kind of comes into play?
Chris Caplice:
34:01
I think so. The thought is, if you have that, you're 10,000 lanes, and only, you know, 500 of them, maybe 1000 of them, say, 2000 of them have volume of more than, say, one a week. Then those other ones, why bother set up a contract for them if they're going to fail anyway. And instead, you have to set up some kind of mechanism to get capacity in real time. And that's what we call a dynamic tendering. And there's many different ways to do that. One is you could go to a load board, which brokers do more than shippers, but you can also set up a small set of carriers, whether they're asset based or non asset. It depends on the mix you want and you want to be able to give them a shipment, a tender, and have them come back with a price they'll guarantee, and then you solve it that way. So yes, it's dynamic in that the price will change, but you've controlled the market. You're going to you're not going to the entire market, because you want to validate the carrier. You go to a set core. And so those set of core carriers also have committed and they probably have business elsewhere in your network. So these are, like your cream the crop carriers, mainly non asset brokers, are really good at this, because they can control things a little better. And so you just have to have some kind of mechanism that, in real time, will assign a carrier to that. And again, it's a closed group of 567, carriers, but the price will be dependent on the situation. At that time, other shippers will use the DA T rate, which we can develop a market rate for whichever day, for whatever location to location origin, destination. And then you can say, well, the carrier will do d a t rate plus 7% whatever. So there's different mechanisms you can have to make a dynamic matching that's not relying on the routing guy that came from an RFP 18 months ago. So, and it makes sense for those low volume lanes, high volume lanes, it works like a charm. Don't even, don't even bother contracts work. Everybody wants stability, and that gives stability to both parties, shippers and
Unknown:
36:10
carriers. Now, does that also apply to sort of the, I guess, the real time pricing, or do maybe more API integrations? Do those play a role in both of those different strategies?
Chris Caplice:
36:23
Yes. So API is just so if you look at the history of connections, you know that you've got the phone, right, that people connect and email, but the predominant method that shippers and carriers communicate for this is EDI Electronic Data Interchange. And it's bulky. It's pain to set up, but once it's set up, it's, it just works like clockwork. It's just works. So for high volume lanes, it makes sense. You know, you set up, you have a pain right up front, but then it just flows. Once it's set up. For the low volume lanes, the cost of setup is not, isn't worth the amount of volume going through. So you want to do something different. APIs are. Good way of providing custom information, and it's system to system. So it really ties to that, and it makes it a lot easier to communicate. You don't have to go through the formal EDI structures. So that works pretty well. What the thing though, is there is no standard. I mean, the joke with EDI is supposed to standardize everything, and the joke is that it works because everyone has their own standard. There is no standard EDI, everyone has some kind of the field in the 64th spaces means something different. So the question is, do you want to find a standard, enforce a standard? Or what we're seeing now is more firms are using AI generative, AI to say, You know what, however you react to me, however you send back to his email or something else, my layer will take that and that unstructured data and turn into structured data. So there's two ways to approach communication. One is agree upon a standard which is hard or move the the intelligence to receiving and being able to take any kind of input and turn it into something that I can use. So we're seeing a growth of both of those over the last couple years,
Unknown:
38:14
and I imagine that that's a perfect use case for da T's IQ product, where they can use all of those different benchmarks and apply it to any given situation. Do I have that assumption Correct?
Chris Caplice:
38:26
Well, we use AI in a bunch of different ways, and a lot of it is a generation of rates in that. And so we have a bunch of very sophisticated modeling techniques because, I mean, the way that the shippers networks are, most shippers, paper manufacturers, you know, they don't. They all have their plants in weird places. Let's be honest, everyone, all of their networks are snowflakes, every every industry. So the last study I did is a couple years ago. If you stack 3040, 50 shippers on top of each other, their networks, less than 5% of the lanes at a three to three digit level will overlap, very little. It's their snowflakes. Now they those will tend to be the very high volume lanes, right? Atlanta, Chicago, Los Angeles, you know? So those, but then the rest of it, it's very unique. And so you have shippers that are very different from each other. And so then you have all of these different lanes. And so how do you price those? How do you set a benchmark on those? And that's where the modeling comes in. And we use very sophisticated modeling. Some was developed by chainalytics over the last 1520, years. And dat is really our our analytics team has done amazing things, and technology has improved dramatically, using AI to come up with a better rate for that for any situation, even if you don't have if that lane has never been used before, Portland, Oregon to Richmond, Virginia. Who does that? But we know what the price would be, because we can, our model is able to predict that. So that's why you need to have different modeling for that. Because you have so many different lanes, that's where it really comes in handy, the AI, that's super
Unknown:
40:06
interesting. Because I imagine, with all of the different scenario planning, can you, can you actually plan for for what we're currently experiencing right now? We are recording this on april 22 2025 just in case, folks at at 145 you know. PM, just in case, you know, these things change, but with all the geopolitical issues that are going on. How do you how do you find what benchmark works the best?
Chris Caplice:
40:31
Yeah, it's funny, because one of the biggest users of benchmark transportation rates are people or companies that are doing network design. So I want to put a facility here. I don't have something there now. What drives the decision a lot of times is, yeah, it's labor cost, land costs, all that, but it's also transportation. So using these benchmarks, you can actually determine, Okay, what would the rough price be for this if I move my facility out to this location? And so that's one of the big use cases of it. And during the pandemic, it was used heavily because companies like I said, before, CPG manufacturers had to open and close manufacturing facilities almost on a weekly basis, so they were doing network flow analysis weekly instead of annually. And so to do that, you've got to understand where your transportation rates are going to be. So we're finding, as it gets more volatile that they need for benchmarks become even more important because you're doing things that you haven't normally done. If we're a steady state, it's kind of boring. You keep going. I'll read on. But if something new happens right, the Francis Scott Key Bridge goes down, that's going to affect rates right? The West Coast. Remember when it had 100 ships waiting off and everything suddenly come to the East Coast, and we realized that we didn't have the infrastructure to take that many containers off the east coast. So these are things that your benchmarks can really help you with. And
Unknown:
41:55
then you've mentioned a couple times about, you know, using different AI throughout maybe your shipment process, or maybe there are companies that are using AI. Slash, you know, machine learning. I'm curious if you've seen any other maybe interesting use cases for using, maybe, like a large language model or something that shippers are using during their you know, I guess facing complications and facing different issues that are going on within the industry. There's, you know,
Chris Caplice:
42:22
there's a lot of debate of what is AI and what is not AI. And if you go to some of the large conferences for software, AI pretty much means they use math sometimes. But if you really look at it, AI is an aspirational term, and it historically, it's a moving target, and it's usually been the thing that we can't quite do now. So things that we you know, 20 years ago, a lot of machine learning was considered AI, you know, neural nets and everything, but it's it's moved. So I think people have gotten comfortable with the idea, and they say AI slash machine learning, ml and machine learning is simply a method of, whether supervised or unsupervised, of gathering insights and recommendations off of data. And then there's a whole suite of techniques there. Generative AI is another technique that use a large language model that does matching and probabilistic nature of determining what makes sense to follow pattern matching. It really is all about pattern matching. So in my opinion, generative AI, large language models do really well when you have unstructured data, unstructured data is emails, you know, blog posts, things like that, pulling that down and making sense of it. If you have very structured data, it doesn't add as much value, to be honest. So some of the use cases that I've that I've seen that I think are very strong, or one that I described earlier, that enabling you to receive data, information from any form, from from your carriers, brokers or other other partners, and making sense of it, getting unstructured data and making and making it into structure so you can act on it, instead of trying to force them to adhere to a standard, which they never will. Another one is if you have to summarize, say you're have a marketing channel, and you have a bunch of things that you need to understand, and so you're pulling something from manufacturing reports, something from sales, something from all these different sources, generative AI is very good at pulling things together and making sense of it and providing a framework for that. So I see that being very useful as well. Where I don't see it being useful, it to be honest, is in a TMS when you want to know, okay, what carriers are not performing well, or what Lane has too much capacity, questions like that, because the data in a TMS is so structured, you know, the six reports you want to ask. Now, beyond that, you might ask some more interesting things, but generally, I think for structured data, you know what you want to ask generative AI won't add as much to that, but there are other vendors out there that are doing that. So a chat bot, essentially, I think that's fine. That's helpful, but I don't think it's where the real money changes, the transformational change. And
Unknown:
45:06
so it almost sounds like you need to have really those two approaches, where you can contextualize the unstructured data, like your emails, your blog posts, but then you have more structured, firm data that plays probably the primary role, if I'm understanding correctly,
Chris Caplice:
45:21
yes, yeah, yeah. And I mean generator has so many other applications, procurement is a big one, especially if you have a lot of products, a lot of skews, and a lot of times your product Master is not in good shape and generative. AI can be very good at that to helping determine and better identifying things. So we're finding more and more use cases for it. I'll give you another one that we're doing up here at MIT. There's a classic problem called the vehicle routing problem. And it's just like it sounds right? You have a bunch of loads for pickup or delivery, and how do you route your driver based on the capacity, the cost, all those different things? And this has been studied since the 17th century with Leibniz, right? It's been out there forever. And so there's a bunch of or type techniques to try to minimize that, the traveling salesman problem, all those kind of things. Well, one of the challenges that you find is that drivers don't always follow the path because they know things that the model doesn't. For example, they might be delivering, and they know you can't park this time of day, or that street is overly congested at this time, so they will deviate from that. And so we did a project with Amazon trying to understand, well, can I use generative AI, large language models to come up with more viable, driver friendly routes, not minimum cost? Because what they found is the plan, just like in truckload procurement didn't match the execution for many good reasons. And so what you can do is you can look at all the good routes that were done, and you have to train on a lot of data, and you start understanding what makes a good route, what makes a good sequencing. And so you can use generative AI to help you with vehicle routing. So. This was really hard for me to understand, but you think about it, we all know generative AI from like our cell phone, when we start typing a message in and it recommends the next word. Well, think of a route as a sentence, and every stop as a word, and so all it's doing is saying, Okay, here's my recommendation based on the probability of successful ones that you should link together into a chain. So it's very interesting early days, but it has shown to perform as well as some of these very sophisticated mathematical programming kind of approaches. So there's a ton of areas that we're looking for, Gen AI. We're still on the hype cycle, right, but we're getting there, and we're, I think it's going to find its areas where it really adds value.
Unknown:
47:50
Yeah, I think you're right. And I think you hit the nail on the head when it's, you know, certain tools are great for certain things. It's knowing which ones to use at the at the right moment that's going to have the maximal impact that you want to see. And speaking of which, you know, for paper manufacturers, we've talked a lot about, you know, all of the different tools that you could and strategies that you could implement, what are sort of the, I guess, the table stakes right now for dealing with all of these unexpected costs and how to maybe potentially save money in the future.
Chris Caplice:
48:23
So I So, I think this, here's the thing. So you talk about these tariffs and things going on, if you talk to a supply chain professional, they'll say, Well, it's just another Tuesday, to be honest. So there's stuff happening, right? And so, but look at what we did last five years, right? That we shifted from West Coast to East Coast arrivals for a period of a couple months, things shut down. And, you know, we absorbed and I think some of those muscles that we trained are still here. Supply chains thrive on being able to handle these things. I like to think of them as shock absorbers. So the supply chain is able to be the shock absorber, right of all these disruptions to keep the ride, the business operating smoothly. So if you think about, you know, tariffs come in one country. There's someone in procurement thinking about, how do I dual source? How do I move production over? So you're really well equipped, especially in supply chain over the last five years, to be really responsive to these changes. So I think the tariffs, you know, yeah, do I agree with the policy as a discussion for another day, but what a shipper can do about it, and especially paper and well, you know, they're going to figure that out. This is what they do. This is why they get paid, what they do. And trucking, to be honest, is very fungible. You can move things around, as opposed to, you're not building new rail track, right? You're not building an intermodal yard. So the ability for truckload to be so responsive makes it very little easier. It's not easy, but you can adjust as much more flexible a tool. And so I think in some, yeah, we're seeing some big churn and all this stuff. And it gets, you know, the press is on it. Now the press is suddenly, they're not. They're experts in supply chain. Now they're experts in global trade. And you're getting all these, you know, the Dire Straits of stuff, but in the real world, you're dealing with it. You deal with this every day. And this is just one of those situations that's finally rising to the level of public awareness. There's a whole bunch of stuff going on that people aren't aware of. But now this is just we're hyper aware of it.
Unknown:
50:30
And I imagine we kind of started this conversation talking about how, you know, our relatives and things like that, know about, you know, the supply chain, and know about, you know, just common logistics practices. I'm curious, though, during a phase like this of what we're experiencing where, you know, the the media hype cycle is all eyes on supply chain and logistics is the the seat at the table still there for, you know, the the chief supply chain officer, the chief procurement officer. Are they still getting that regular seat at the executive table for for large scale manufacturers? I imagine they are. Yeah, it
Chris Caplice:
51:05
has not changed. It has not changed. In fact, I'm looking back at your background, and I see you have Levinson's book, the box on your stocks there, and that was one of those obscure books that came out, what about 15 years ago? And he has a sequel that came out with it that looked at it more. But I think back then that was, it was kind of obscure, but now it's, it's at the table. It got, I think it definitely got its table, a seat at the table, rather, during the pandemic, and it stayed there. It hasn't changed, because now it's just the opposite. One of the big things that we do at Da T is try to arm transportation executives with the arguments, to go to the CFO and see CPO and CEO the whole C suite to understand why things are different. Truckload trucking, transportation is different from most other commodities, and because what happens is in a soft market like we've had for the last two to three years, if you run a bid, you're going to lower the rates, the new rate differential, which means the rate contract rates coming in compared to the rates leaving you. Will be lower. So you'll, quote, save money. If a CFO sees that one year, they'll expect the same thing next year and the next year and the next year. And we have to train people say, No, you're just it'll come back to bite you, right? It's just the market moving. It's not like you're doing a amazing things to the markets in your favor. And the same thing with the idea that if I have a certain amount of volume, if I double it, if I increase that I'll get a lower rate. And that's false too. There are no economies of scale to purchasing truckload transportation. It's all about economies of scope. And scope just means a balance. Do I have loads going here? Do I have loads coming back? The more unbalanced I have, the higher my rates are going to be. It's just a natural part of the economics of truckload transportation, but being able to explain those things to someone who's not in the space is another thing that we do at Da t try to educate so you know that the C suite understands the challenges that you're facing. And so for
Unknown:
53:06
those procurement workers, Chief supply chain officers, you know, all of those different roles that serve, let's just go back to the paper manufacturer, example. What are some of those, I guess, insights and tips that you would arm those people so they can go back to their executive team and essentially explain, you know, what is the current market conditions? This is what our costs are expected to be, but it might fluctuate at somewhat, you know, in the near future. What would those little tidbits be? It's,
Chris Caplice:
53:33
it's, it's a tough thing. So we'd like to say that, you know, the time to fix your roof is when it's not raining, but it's a temptation. And so we were advising people through the end of 2024 that we expect rates, truckload rates, in the in North America, to rise two to 4% over the course of the year. And so if you're running a bid, you know you can easily squeeze it and go the lowest cost, or you can give some back. Like, every time you like want to favor incumbents, you're giving some money back. And so we're advising them to do that, because that we sense that the market is tightening now with these tariffs, it's pushing that tightening off. We if you look at the rates, you know they go like this, 2017, 1820, 1920, 20 goes back down in 2022, but then it's been flat. It's not rising, it's not dropping. It's been an in what we call a trough for last 12 to 18 months. So when a system is at equilibrium, where it's kind of balanced, any kind of sharp reaction will cause a big jump, one way or the other. And what we're trying to figure out is, well, what will cause rates to suddenly increase? It can either be increased demand, dramatic increases in demand, or contraction of capacity. We've seen the carrier base contract over the last two years, but there's still 100,000 of those small carriers out there who might have parked their truck, but if the rate comes back, they'll come right back in so you're trying to understand where where do we stand right now? And so we're advising people, get your roof in order, get your routing guide in order. And, you know, get take, be willing to take a one, two, 3% price increase. And so now we're seeing that push off a little bit with tariffs, because tariffs will not increase demand. We don't see that happening, maybe, if it works, according to what the administration is saying, all this manufacturing suddenly moves back to the US, maybe, but that will be years in the making, and assuming that the the policies are stable, and all signs are pointing that they're not stable, they're changing at the you know, enacted On Monday, removed on Wednesday. So in this era of uncertainty, it's really hard. And so my recommendation that we always recommend to shippers, work with your core carriers, make sure you have your core capacity in place that you need, and make sure that you have a stable relationship for the spot or those, those low volume, low consistency lanes, create a connection with brokers, three non asset based cares that can provide ramped up capacity when needed and keep them in tow as well. So just prepare for the surge up when it does happen, because it will happen. It will happen at some point. I don't think it'll be like the pandemic, but it's not going to continue to go down. If you look back 15 years, a compound annual growth rate in rates for Drive and truckload is about 2.4% and so if you continue that out, that's kind of sustainability. If you talk to a carrier, it's tough at a two point whatever percent growth, because their insurance went way up more than that, wages went up more than that. Tires are more than that. So we're that, that weird equilibrium. I wouldn't expect rates to go down any further, but we're, we're waiting to see if what would cause rates to start increasing beyond the trend of 2.3 to 3%
Unknown:
57:03
Yeah, I think a. Lot of folks are holding their breath, hoping that we've already reached the bottom, or maybe we're currently at the bottom, as far as rates are concerned. But I think it's sort of the the commonality phrase of it depends. It depends on a lot of things.
Chris Caplice:
57:15
That's why this markets. I mean, the industry is so interesting. I've been in this for what, 30 years now, and it's fascinating because you've got so many players, and it's such a big it's, it's the biggest niche market that we have. It's about $500 billion and it's weird. It's, it's, I used to give a presentation called truckload. Keep truckload weird, um, because it has different behaviors. It's a drive demand economies of scope, hundreds of 1000s of players. No one sets the price. No one's a price maker or or sets the stage. Everyone's a taker. So it has all these interesting components to it. And it's been the play child of economics professors and or for decades, because it's fun to play with. You can do all these interesting things. It's highly competitive, and it's strangely deceptively complex. It's, it seems simple, but it's very complex.
Unknown:
58:07
Well, I think that's a perfect place to it's sort of a mic drop moment. But one last question here, is there anything that you feel is important to mention that we haven't already talked
Chris Caplice:
58:17
about? No, I think you've hit most everything. I mean, it's it's such a big industry, and there's a lot of everyone is a little bit different. Every ship is a snowflake. Every broker has a different angle, every carrier has a different area of expertise. But what makes it interesting is really finding the puzzle piece to fit them all together. Where does it make sense to use which different asset, which different type of relationship? And so, like I said, it's been a very interesting industry over the last 20 years, and even more so in the last five years has technology has changed. I talked to the President of oida, the owner operator, independent drivers Association, and he was describing trucking in the 70s, and he said, you pull into like Chicago, you stop it at the local truck station, and figure out where the address is, and see if anyone there knows where the heck you're going. And you look on a map now carriers, an owner operator, has the market in their pocket on a smartphone. I mean, with D 81 all those other things, load boards, you can see the whole market. So one of the things that's contributing to this more stability, equilibrium I've been talking about is technology. Carriers are much better informed and have much it's much more frictionless than it has ever been before. So is that leading to less volatility? We'll see. We'll see, but it's an interesting time to be in this industry. So, Blythe saint, thank you for having me on today. Absolutely,
Unknown:
59:37
knowledge is power, and you are definitely sharing a lot of those gems throughout this entire conversation. So So Chris, where can folks follow you? Follow more of your work, all that good stuff. Just
Chris Caplice:
59:47
go to D, A T. I'm the Chief Scientist there. So we do a freight vine podcast every two weeks. Just Google freight vine and I'll pop up. Or you can do capital at MIT, and you can find me there. Amazing.
Unknown:
59:59
This was I wrote down a post it note of all of the great timestamps of really good quotes during this conversation. So it's quite full on my post it notes. So thank you again. Chris, Great. Glad to be here. I hope you enjoyed this episode of everything is logistics, a podcast for the thinkers in freight, telling the stories behind how your favorite stuff and people get from point A to B, subscribe to the show, sign up for our newsletter and follow our socials over at everything is logistics.com and in addition to the podcast, I also wanted to let you all know about another company I operate, and that's digital dispatch, where we help you build a better website. Now, a lot of the times, we hand this task of building a new website or refreshing a current one off to a co worker's child, a neighbor down the street or a stranger around the world, where you probably spend more time explaining the freight industry than it takes to actually build the dang website. Well, that doesn't happen at Digital dispatch. We've been building online since 2009 but we're also early adopters of AI automation and other website tactics that help your company to be a central place, to pull in all of your social media posts, recruit new employees and give potential customers a glimpse into how you operate your business. Our new website builds start as low as$1,500 along with ongoing website management, maintenance and updates starting at $90 a month, plus some bonus freight marketing and sales content similar to what you hear on the podcast. You can watch a quick explainer video over on digitaldispatch.io, just check out the pricing page once you arrive, and you can see how we can build your digital ecosystem on a strong foundation. Until then, I hope you enjoyed this episode. I'll see you all real soon and go jags. You.
