Jason Traff on DAT Freightvine Podcast: Escaping the Digital Filing Cabinet of Logistics

In this episode of Freightvine, Chris Caplice, Chief Scientist at DAT Freight and Analytics and I sit down to dive into the evolution of supply chain management, the role of technology, and how Shipwell is revolutionizing the landscape with innovative tools and strategies. From global supply chain nightmares to AI-powered TMS solutions–we cover the gamut.

Find out more by watching the complete episode on YouTube or listen on Spotify, or Apple Podcasts to hear the full conversation. Prefer to read? Transcript below.

{{divider}}

Chris Caplice, Chief Scientist, DAT Freight and Analytics

Hi, Jason. Welcome to the Freightvine Podcast.

Jason Traff, President & Co-founder, Shipwell

Thanks so much, Chris. Excited to be here.

Chris 

Yeah, you're in much nicer weather than I am. I'm up here in New Hampshire, and it's sunny, but we still have three feet of snow on the ground. And you're in sunny Austin, where it's beautiful.

Jason 

Right. It's going to hit 85 today, which is a bit warmer for February than I would have chosen, but it feels like I'm bragging when I tell it to you.

Chris 

Yeah, pretty much, pretty much. But that's okay. So, I wanted to have you on because DAT and Shipwell are partners and doing some great stuff, but I wanted to find out more about Shipwell. We'll talk more about that, but let's start with talking about you. How did you find yourself in your career journey to where you are now at Shipwell? Give us a quick synopsis of that.

Jason 

Yeah, sure. My background is in transportation, finance, and entrepreneurship. Probably most relevant is before Shipwell, I was running a company based in Shenzhen, China, called Copycat Paintings. We employed a thousand artists and made reproduction artwork that we shipped across the world.

If you went to the Van Gogh Museum in Amsterdam, you might have seen one of our little paintings in the gift shop. We made that painting in Shenzhen and shipped it from Shenzhen to Hong Kong to Rotterdam to Amsterdam so you could buy it. While the art side of the business was incredible and inspiring, the logistics side was a complete nightmare.

I was essentially running a global supply chain with zero visibility, all manual processes, phone tag across time zones, and just crossing my fingers that things would get through customs. I realized very quickly that the physical movement of goods was just this massive, stressful black box for me.

Chris 

Let me ask you a question, Jason. Were you mainly shipping to individuals, to museums, to stores? What was the mix?

Jason 

We started with individuals, but we found that, as cool as it is to have your face on the Mona Lisa, we actually did lots of commercial deals. A lot of them were for new hotels, restaurants, and museum deals. It moved more to a business-to-business model.

Chris 

I would think that licensing and copyright would come into effect. Was that an issue?

Jason 

For most art that's over, I believe, 75 years old, it becomes public domain. So, you can actually reproduce them. For most of the deals we did, they either had the licenses themselves or it was just custom modern, landscape-type artwork.

Chris 

Got it. So that's pretty far away from trucking.

Jason 

Well, global supply chains have very similar themes. What really led to Shipwell was I connected with my co-founder, Greg, who's our current CEO, at MIT. We were swapping war stories. He had just left McKinsey, where he was doing Fortune 100 supply chain consulting. He was telling me about all these companies with essentially unlimited budgets struggling with the exact same problems I had.

It was mind-blowing for me because I was a relatively small business, and I just assumed with an unlimited budget, you wouldn't be using 20-year-old legacy software. What we found was that there was room for innovation in creating what is essentially a modern TMS.

Chris 

So, mostly what you did in Copycat was global, I assume? International? Mostly air, some ocean?

Jason Traff

That's right. A lot of air.

Chris 

And then you're at MIT, you're talking about this stuff. What made you decide to found Shipwell?

Jason 

We were looking at the current state of software. This was about 2016. We were all bright-eyed and bushy-tailed. We saw these common themes—giant companies struggling with processes. Meanwhile, as a consumer, Amazon was doing two-day shipping.

Most consumers were familiar with what a vertically integrated supply chain could do and the level of visibility you could achieve. Shipments show up on the day they’re supposed to, and you might even get stop-by-stop or turn-by-turn updates. Meanwhile, massive companies were struggling with what were essentially SMB problems.

We found that most of the software they were using was great in the 90s, but it was essentially just a digital filing cabinet. It didn’t do any work for you; it just recorded the work humans did and served as a reference point.

This was where we felt that, on the back of e-commerce and IoT—much less what would become COVID and evolving geopolitical volatility—this was the time people were going to start caring about what shipping both costs and means for a brand and customer experience.

Chris 

So, you’re in grad school, you’re at Sloan. What was the step for this? Usually, when you start a company or software, you need an anchor client. How did you go from talking about this over beers at Sloan to actually founding something and getting your first paying customer?

Jason 

We originally started Shipwell with the idea of servicing enterprise. What we quickly found was that you just turn into a consulting business. We realized we were going to have to build a global, modular, modern TMS from the bottom up.

To do that, we started narrowly in LTL with a focus on e-commerce, which allowed us to land some VSMB businesses. We then built out Truckload to support those VSMB businesses, which allowed us into SMB. This snowballed until today, where Shipwell is in 12 different modes of transportation with a global footprint.

Chris 

Okay. So, you started with LTL because LTL is complicated at times for pricing and things like that, right? You're combining freight, you've got classification and rules that don't make sense, all the tariffs and everything.

Jason 

Yeah, and part of the naivete is I thought that LTL would be easier than truckload because it just felt like it was more direct. There were APIs we could call, rates we could get, prices we could display in front of customers, and tracking data we could pull. The ELD mandate was coming for trucking but hadn’t quite happened yet. So, LTL felt like the right place. I think if I was doing it again, I might have started with truckload first. But, you know, that’s the benefit of hindsight.

Chris

Yeah, but then, you know, the funny thing is truckload is very crowded with truckload solutions because it’s seemingly simple—point to point. But I’m trying to think, in 2016, Uber Freight hadn’t launched yet. They launched in 2017 around there, but there was a whole cry for "Uber for freight."

Jason 

Yeah, there was definitely a wave there of digital brokers, this idea of commoditizing the experience. And so much of it was about dissolving the inefficiency of the discovery process. If everyone has perfect information in a level playing field, that really changes what it means to transact freight of any mode.

Chris 

So, what made you decide not to be a broker and to go straight pure software? Because a lot of the companies were like a managed transportation service.

Jason 

We actually had a small brokerage for a while. It really grew out of trying to please SMB and mid-market customers. There was one weekend where both Greg and I were tracking freight for a customer that had a carrier fall off, and they were in a panic. We were so small, we were trying to help. Then we looked at each other and said, "Pretty sure this is a different business." So, we had a small sidecar brokerage for a while, and that really helped in the early days to connect carriers and form the process.

The MIT motto, Mens et Manus—mind and hand—is about action learning. It’s about not just building off theory. So, this idea of eating our own dog food made a lot of sense. But as we started scaling up into mid-market customers, they didn’t really have a need for a brokerage. They get enough of that exposure already. It started to deviate from our core mission, which was to build something that connects the world and provides a modern supply chain experience. So, we actually sold off our brokerage a few years ago.

Chris 

So, the idea was to sell mainly to shippers—people who own the beneficial freight. What about selling to brokers themselves? Is this a broker-based solution?

Jason 

We do have a few broker customers. Brokers make up probably about 5% of our customer base. But what we always felt was that the problems that could be fixed were on the shipper end. Specifically, there was this theory about how Amazon provides an amazing experience on the customer side. Even if you look at the largest e-commerce giants in the world—Target, Walmart, Home Depot—that experience falls off pretty quickly.

As you get further and further away from that, we felt like we could continue seeing that slide. But then we thought about what the root problem was, and it was because so many shippers either don’t have tools or they use a dozen different tools. They really just didn’t have good options.

Our feeling was if we could provide them with something that would help get them up to parity and build that modern experience, you could end up in a future where you don’t have just one company that provides an Amazon Prime-like experience, but 500, 5,000, or 500,000. What would that mean for the world in terms of efficiency?

Chris 

And so, initially, you said you started for the small-medium size and then went to mid-market, and then the next tier is kind of enterprise. How do you define those segments?

Jason 

Sure. Mid-market was always our ultimate goal. Our feeling was that enterprise—the Fortune 1000—they all have a TMS. They bought it in the 90s, and it’s very hard to move them off of an SAP TM or Oracle TM. But mid-market was never going to do the three- to five-year implementations or the three- to five-million-dollar costs.

It was never the right solution for them. Even though mid-market was relatively smaller than enterprise, these are customers that still move $20 to $50 million a year in freight on average. There’s a real ROI behind that.

For Shipwell, by count, mid-market customers in that band make up the vast majority of our logos. But we have customers as small as $5 million a year in freight, and we also have publicly traded customers. I think we’ll touch something like half a trillion dollars of public market cap through our publicly traded companies this year, even though that’s a relatively small count of our overall logos.

Chris 

In my experience with software, it’s been many years. It was client-server back then, before software as a service and all that. There was a big difference selling to enterprise, mid-market, and small-medium businesses. A lot of times, the nice thing about enterprise is they have processes, and they’re kind of the same, and you can move from one to the other.

What’s your experience? What is the difference, in your perspective, for dealing with a small company—a $5 million, $10 million, or $2 million freight spend—versus a $50 million mid-market company?

Jason 

I think the biggest switch probably comes in that $5 to $10 million a year freight spend. When you’re smaller, you can bear the manual processes. It’s a lot of spreadsheets and whiteboards and grinding really hard, but you can conceptually cover it all. At some point, it will outscale you. You’ll realize that you have a thousand shipments in transit, you’re not sure what’s happening with them, and you’re never able to get ahead.

That’s really the forcing function we see where companies say, "I really need something that’s not just a digital filing cabinet, but something that’s helping me be proactive, empowering my workflows, bringing real-time visibility, and allowing me to have more constructive conversations with my carriers."

That’s where the ROI really starts to shine through. Most of these customers, even at that level—between $10 and $100 million—don’t have a TMS. They might have used a TMS in a previous role, but 85% of mid-market companies don’t have a TMS still. Our view is this is still the forefront of a digital evolution.

Chris

Well, Jason, would you say the majority of your customers, um you're the first TMS they have? You're the first install?

Jason

Yeah, that's right. Yeah, I mean, usually they might have had ERP customization. They might have been running it decentralized. You know, hey, I've got six locations and each location just manages it themselves. And they just send the invoices into HQ once a month to reconcile. 

But usually this is part of a larger digital sort of innovation initiative they have where, and COVID was a big change for this, where they recognized, hey, this is something that is not only a lot of money, but something that impacts our operations our customer experience our ability to compete you know efficiently in market and so maybe it's time to get good at it and that's that's really the change we've seen.

Chris

Yeah. It seems like when, depending on the market you go for, you have to make a trade-off and the TMS that we were building was for large you know, fortune 1000s, fortune 500s, 100s. 

And so for that, you had to be able to handle all the complexity, all the different little ins and nooks and crannies and stuff, or else they wouldn't want it. It sounds like you guys made a different trade off. Not saying you can't handle that, but it's ease of implementation.

Jason

Thank you.

Chris

Can you talk about how the market you're going to influenced where you put your efforts on, whether it's ease of implementation, you know, first TMS versus handling those complex corner cases?

Jason

Yeah, and it's hard, right? Because you have to be both simple and easy to use, but also powerful. Because I think what we find is that supply chains have a lot of change management that's required. And even companies in the same industry have vastly different, you know, supply chain footprints, different locations, different workflows, maybe different modes of transportation. And so there's always a change management element to it.

But you know we had the benefit because we started in 2016 that we got to use full REST APIs, modern programming languages. you know We're a full cloud native microservice REST API with you know Gen AI chatbots and things like that.

And so we had the benefit of starting when we did. So we could build something that was more nimble, more configurable, and then focus on building out configurations rather than deep customization like you would need for enterprise. And this essentially allowed us to set up a TMS where even today we implement in about a quarter of the time of most of our legacy TMS competitors. And so what this means for mid-market is, you know, whereas the SAPs and Oracles were never a great fit, you know, this isn't this is an initiative we need this year. We want savings this year, not three years from now.

This has essentially become that trade-off where we distinguish ourselves.

Chris

So, what's the gating factor for your implementation? Because as I've read, it's fastest, fastest in market. I mean, you have Gartner recognize you as the easiest to use, easiest implementation, I believe.

So what, what, um, what is the rough gating factor that makes it that doesn't make it not be even faster? What's is it the training? Is it the getting the connections? Is the data?

Jason

Yeah, I'm really proud of how well we do on implementations. And I think it's a really good fit for mid-market companies. There is still so much room for improvement with it, though. And I think that's part of our own internal initiatives, especially as we look at how we use and adopt AI to make it even better and more efficient and actually make it easier for customers.

What we find is if I go back to my example of the company with six locations that's now deciding to centralize, fray, put all their purchasing power in one place, they might find that each location does it differently.

Chris

Yeah.

Jason

They have different workflows. They keep track of addresses differently. And so even though they might have used a TMS before, they probably underestimate how difficult it is to get all the carriers in one place, set up all their addresses, pull all their SKUs together.

There's usually an ERP integration involved where we're pulling in orders to consolidate those in the shipments. And so there's an IT component as well that's going to be going through its integration efforts. And so it just takes lots of coordination and I think gentle handholding and then sort of like keeping an eye focus on the bigger picture.

Because at the end of this, what we find is that most companies end up getting a 10 to 20x ROI in their first year. And a lot of that could be reductions in freight costs. It could be increases in efficiency, centralized purchasing power.

You know, we have so many stories where companies have underestimated the ROI they'll get. But then in their first year of going live, what they find is it improves, you know, their net margin is an entire company by several percentage points.

And then for some companies, that is a very meaningful difference.

Chris

Yeah, that makes sense. I just want to mark that it took us 16 minutes to mention AI. That's a new record for this. But I wanted to go down because you raised a point. The thing that I found working with smaller companies, the smaller the company, the more peculiar they are.

And it takes, you know, they've got their own things. You mentioned having like 60 seasons, they all do stuff differently. They store their addresses differently. um So do you, I'm just curious, do you adapt to that or do you force them to adhere to a single standard? How do you approach that, cause there's two different ways you can go.

Jason

Yeah, and it really depends on the personality of the company. You know, we generally see companies coming through in two different flavors.

Chris

Okay.

Jason

One is usually a more traditional, maybe family run hundred year old business that's going through a digital innovation to stay modern and outcompete competitors. We also see very young, fast growing companies that, you know, really see the need for a TMS because they just can't keep up.

But there's always a change management element. For us, we try to lean very heavily on best practices. You know we're always mindful that there might be a reason why a company needs to do something a certain way.

Chris

Right?

Jason

But it's really very unlikely that any TMS is a perfect fit out of the box for for a company because there's just not enough standardization in supply chain today.

So typically what we've done is we've built Shipwell to the best standards and best practices, but there's always, there's always a, you know, back in my day, the red button was red. It wasn't green. And so, you know, there's always a little bit of that change management that a company needs to be willing to accept because the goal is not to have all the red buttons be green. The goal is to save $3 million dollars in transportation spend, gain back countless hours of, you know, insanity and manual work so that you can, you know, keep pace and distinguish yourselves as a business.

Chris

If you had to look at the ROI and put it into buckets, just roughly, I know it depends company by company, customer by customer, how much you think it is time savings, labor savings, rate reduction, process simplification? What buckets do you put this in when you convince for the ROI?

Jason

Yeah. And so we work with our customers for this because there's about a dozen different levers that we'll look at. And each company has different levers that are probably more applicable to them based on either where they are or their industry.

Chris

Sure.

Jason

And so we see it against, you know, planning, procurement, execution, settlement. And so it could be everything from, you know, we see a lot of these companies just going back to my earlier example of the six locations.

They're buying individually. You know, they're not pulling their capacity together. There's probably a person that's in charge of, you know, we had a large meal kit company that was a customer and they had their individual buyers. You know, the person that buys the meats and cheeses and lobsters also booking the freight.

And what you find is that your guy that's great at buying lobsters probably isn't great at both booking lobster freight and tracking lobster.

Chris

Right. Right.

Jason

And so by pulling all these together, you know, there's usually an efficiency just from economies of scale. There's also more efficient consolidation as you start looking at the mix between truckload and LTL.

Chris

Yep. So, so this mode conversion a big one for that too, where people might be doing a bunch of LTL and thinking about combining that.

Jason

Yeah, absolutely. I mean, I think it depends a little bit on the business. You know, some companies, their freight's not so conducive to one motor or the other.

Chris

Yeah.

Jason

But for lots of other ones, you know, they're dealing with multi-leg, multi-stop, you know, multi-skew, multi-shipper sort of customer orders. And, you know, it's very difficult for a planner to keep all that in mind, set it a week in advance, hope it doesn't change in that week.

And so having a system that really combines these things efficiently, um it can add to a real savings.

Chris

So when you're selling to a company and they've never had a TMS before, do you find yourself selling against other TMSs or against staying with Excel? Is it ever the fight where we're doing just fine?

Jason

I think we see both. Yeah. You know, ah we used to see that a lot. We used to have this persona that was the no change Charles. It was someone that had been in logistics for 20, 30 years.

Chris

Right, right.

Jason

They're retiring in 10 years. They don't need to rock the boat. They're doing just fine. But largely since COVID, I think that persona disappeared. You know, I think everyone realized how brittle supply chains are.

Chris

Okay.

Jason

And I think this is the idea that we as an industry are moving from like the back room to the boardroom, I think is the analogy I hear a lot.

Chris

Yeah, yeah. Right.

Jason

You know, people are familiar that supply chains are a key risk or opportunity for a business. And so while we might be the first large purchase a transportation department has ever made, they usually put lots of care and thought into that.

And so we do see other TMS’ in the mix. You know, we distinguish ourselves very well in those comparisons. But, um usually the choice is not to stay and do nothing. There's usually a larger drive, which I think is necessary because it's not an easy thing to stand up.

Chris

Right. Yeah. Cause you've got to, if you're fighting those internal battles, hopefully the company's already fought that battle and decided to move forward.

Jason

Yeah. And then on the other side of the coin, this is where it's really important as you get stood up to show the ah ROI, how many shipments have been, you know, tracked, booked automatically, how are rates comparing not only against last year, but against benchmarks.

Chris

Cause that's almost.

Jason

Like, I think we see this where, you know, most supply chain leaders have never had these tools before. And so making sure they can appear, you know, hardworking, intelligent, informed, proactive as they're showcased for the first time, like that's one of our core missions.

Chris

So one of the things we talked about earlier, you know, we like everyone to do this and that's what we designed to. But then you find most companies don't follow the happy path. They do something a little different, and yeah the deviations. 

And so I always think about this when it's in urban planning, whenever there's a snowfall. You see the streets that get plowed and the paths, but then there's the desired paths where people walk, even though it's not plowed.

Same thing for grass. You'll see the path going across the field because it's the shortest path. It's the same thing. So for deviations, do you track or how do you handle when someone hacks the system or deviates from it?

Do you, because they're not, that, knee-jerk reaction sometimes is to build a fence and don't let them do it versus try to understand why. Do you ever look at that for the deviations, why they deviate, why they go bespoke?

Jason

Do you mean in terms of like a customer's supply chain requires customizations to use a TMS or that they're using the TMS in innovative ways as a workaround for something?

Chris

They're trying to get the TMS to do something that it wasn't necessarily designed to do. And they're finding a way around it to get it done.

Jason

Oh, we love that. I mean, I think that the TMS really sits, you know, as sort of like that center hub in like a hub and spoke model of supply chains. You know, we're connected usually to an ERP, warehouse management, yard management, a number of carriers, maybe customer orders as well.

And so, you know, we love it when people think more creatively about how that gets adapted to their own supply chain footprint. And so one of the things that we really designed Shipwell around was being configurable for that.

So we have an OpenREST API.

And so, you know, we have our user conference here in two months and we're going to have customers showcase some of the incredible things that they've built inside their yeah ERP to bring, you know, some of that TMS functionality and surface that visibility, maybe in front of like a salesperson, you know, that has a customer calling in asking where an order is.

So rather than logging into the TMS, they see it right in front of their dashboard.

Chris

No, that makes sense. And having a user conference like that is so nice because they learn what we found at DAT. It's so important to have them talk to each other because they're doing it every day and they're learning different things. Let me ask you a question about just the tool, because as you mentioned, transportation has a pretty big footprint.

And both in terms of what systems they touch, but also in terms of the type of decisions being made, strategic planning versus tactical execution.

And so I assume you touch upon all aspects of this to some degree. How much, but would you, if you had to classify yourself more as an execution system or as a planning system, where would you put that mix?

Jason

Yeah, I will say that the lines are blurring for us.

Like, I feel like a TMS is core has always been execution, but you know, I think for a lot of software development that came through the seventies, eighties, nineties and beyond, you know, it was much more likely to just have a rigid sort of definition of where something started and stopped.

Chris

Absolutely.

Jason

Whereas I think what we have found is that there's often a lot of value into blurring those edges. So Shipwell does touch on some elements of planning because we feel like they're native to the TMS and being able to know where orders are and how shipments are set up and being able to have that be more proactive and more real time is beneficial.

Likewise, on the tail end, you know, we are built on top of a full business intelligence layer. And so being able to pull analytics in real time, have it exported to a data lake like ah like a Snowflake,

Jason

You know, we try to give people full control over that as well. And so, you know, the core is always execution.

Chris

Right.

Jason

Can you book freight efficiently? Is it priced well? Is it tracking? Is it settling? Are the invoices getting paid? All that. But I think what we find is that the periphery has lots of value as well.

Chris

Yeah, and it seems like, you know, when you talk about execution, it's all about speed, right? You don't spend hours on a transaction. On planning, you're willing to spend a little extra time because if you do something wrong there, it has a longer shelf life.

Chris

And so, yeah, it's and different tools, different people and how they manage that. What about procurement itself? Are you involved, in like, the RFP or the mini bid space that feeds into the routing guide that I'm sure is part of Shipwell?

Jason

We do. So we have an even standalone RFP tool that both works for native TMS users, but can also be used as a standalone.

And I think this has been part of you know how we've sort of built a philosophy around what rate negotiation looks like. And I think we've really moved past this era where you know you had a certain percentage that was spot and you had a certain percentage that was contract and you ran your ah RFP once a year and then you revisit it a year from then.

And I think what we're seeing is that these things are much more dynamic. you know ah for most of our customers that are using our RFP tool, they'll run multiple RFPs, you know the idea of a mini bid.

Chris

The mini bits. Yeah, yeah, yeah, sure. Okay, so um let's go back to AI now. So it's 2016. We know what AI is, but no one's using it really, except maybe internal at Google. um But then, you know, when did the shoe fall?

2022, 2023? Guys say, oh my gosh, this changes everything? When did you wake up to that? Because I assume you didn't have AI built in 2016.

Jason

Mm-hmm.

Chris

Yeah.

Jason

We didn’t. And, you know, I think part of the vision of what we wanted was enabled by AI. But, you know, we've had AI in the platform for the last three years, really with generative AI and chatbots.

Chris

Okay. Sure. Yep.

Jason

And we've had chatbots on our REST API so that, you know, the IT teams doing the integrations could have a Q&A session. But, you know, just for us, we're usually the youngest company in most, ah you know, competitive processes we're in. We're one of the youngest in Gartner's Magic Quadrant for TMS. And so I think we always felt, um an obligation to explore frontier technology because, you know, we just didn't start 30 years ago. And so as part, we don't, right?

Chris

You don't have the legacy. You don't have the copper wires. Yeah, yeah, yeah.

Jason

And I think those those those companies have their own, you know, their own perks and they have long feature lists and they've done tons of integrations. And I think that's wonderful. For us, you know, our differentiator was speed to value and then keeping our customers competitive with what the vision of what a supply chain should get to.

Chris

So if I look at, oh, go ahead, go ahead.

Jason

And so for us, you know, no please, please.

Chris

Okay. So, if I look at Shipwell pre 2023, So the first seven years and now the next three years, how different would I find this product?

Jason

Yeah.

Chris

That's the introduction of AI is pretty much generative AI, GPT launched.

Jason

Yeah, I mean, the product has changed and evolved so much from where it started. You know, I think we've always had a very this is what happens when you have a CEO who is a legitimate genius and holds multiple master's degrees from MIT with a background of machine learning and all the fun stuff.

Chris

Yeah.

Jason

But I think what's been interesting is, you know, in the 10 years that Shipwell has been around, the first few years I can categorize as feature parity, you know, building a solid execution platform.

Chris

Sure. Mm-hmm.

Jason

And then really sort of the next three years was really around the ELD mandate. And this was really the idea of like the visibility era, you know, ah getting getting real time data from your supply chain so you can react to it. And ideally, layer in, you know, what like what we did was algorithms around traffic and weather so you can try to get proactive even.

And so this was sort of that middle era of visibility. And I think where we are now is we've transitioned past the visibility era into what we would sort of refer to as the action era.

What I think we found from so many of our customers is that it is wonderful if you can have a dot on a map that shows you where a truck is and then it's late.

Chris

Yeah, yeah.

Jason

But that is no longer a competitive advantage, it just lets you watch the car crash you know in real time you know and the current landscape i think is currently split behind uh like legacy TMS’ that are still those digital filing cabinets with a long list of features that are very customizable. And then sort of like platforms like ours that are much more modern that, you know, are actively taking a stake in agentic AI where we're trying to, you know, grab the steering wheel, know, fix the problem, act as a digital co-worker, not just a co-pilot.

Chris

Yeah, that makes, that makes, it's funny you say co-pilot. But um when I was selling software, maps were like the thing that sold, but no one used them in practice.

Do you see the same thing? There are certain features that the buyer likes to see, what the executive, but the practitioner, the actual user doesn't use it.

Jason

Mm-hmm.

Chris

I don't know. Is that, is that still a dichotomy that you see in the market?

Jason

I think, you know, I think a lot of the ah ROI when we talk with our customers is very contextual. Like if we talk about visibility with the customer, visibility means one thing to the C-level, and it means something totally else to the person doing the tracking and tracing, right?

And so I think we've had to always sort of split that context so that we're providing value to the person using it because you need them to not complain all day about the TMS in what is already a very stressful job.

And I think this is one of the areas where we've actually seen the agentic AI have its biggest impact. is allowing them to take off a lot of the grunt work that is inherently frustrating and low value and focus on more strategic things, you know the planning, the procurement, ah that allow them have more constructive you know conversations and relationships with their carriers.

Chris

Yep, yep.

Jason

And in turn, still create great structured data so that when the C-suite is looking at visibility, you know they can see the benchmarks, they can see it getting better, they can see the ah ROI. And I think you have to you have to span that that full gamut.

Chris

Yeah, no, that makes that makes sense. One of the big changes that we've seen the last five, 10 years is it used to be when I was doing bids for companies in the 90s, 2000s, they put every lane out to bid and you'd have a contract for every lane, even if it shipped one load per year, whatever.

Shippers got smarter now and they're doing more and they're finding better ways to handle random freight, the one-offs. Cause every, every truckload network specifically LTL is different. They have one-off kind of lanes. The majority of lanes are are kind of small, low volumes. And it seems like instead of waiting for the routing guide, you know, to go through the first, second, third, were all the alternatives, there's a quick exit to a spot mechanism. 

And it sounds like you have something like that where you have zero human except interference for this, no human in the loop for doing that quick connection for the distressed or low volume freight. Can you talk a little about that?

Jason

Sure. I mean, I think we all love a world where supply chains work perfectly, right?

Chris

Yeah.

Jason

The problem is, is that that's just never the case, right? There are too many real world factors that interfere with the best laid plans.

Chris

Right.

Jason

And so, you know, I think specifically like ah this, this fits well into the context of like our RFPs, right? You, you, you can run an ah RFP, you can award a contract and then you can start working through the preference of the contract.

Jason

But what happens if those carriers can't take it? What's the failure state of that? And so this is where like the idea of like a waterfall bid, you know, you have a pool of carriers. How do you, how do you coordinate, communicate to sort of find the best preferences, both, you know, for the carriers and for yourselves.

Jason

Sort of if you end up in that sort of contingency scenario. And so just we have so many high volume you know truckload shippers specifically that that's absolutely something that they find value in.

Chris

That makes sense. So there's been some recent press and and this is so funny. The market is so nervous about things. A press release comes out from a company that doesn't have any customers that they can reduce transit time, you know, ridiculous amounts and everything and all the logistics company stocks tank.

What are your thoughts on this? What's your opinion? Is AI going to eat software as a service or where do you see the moats built? What are your thoughts?

Jason

Yeah, I have lots of thoughts. um you know And you mentioned it earlier, right?

Chris

I thought you would.

Jason

Most people still haven't used AI. And I think that's especially true in our industry. Just as an industry, we we technically lag behind other industries for technology like adoption and innovation.

I think we technically beat agriculture and mining, but, you know, we're usually not the front runners like we've never had the luxury of the budgets or the products to choose from. And so we live in a world today where point zero four percent of the world has ever coded with AI.

Just a fractions of fractions of percents have even you know paid for AI or seen the full capabilities or use it in their day-to-day lives. like We're still very early as a global you know sort of statement in the adoption of AI, much less as an industry.

And so I think part of this comes with like a natural skepticism around, like is it real? Does it work? How would I even use it? It's going to have tons of problems. you know I think we're, as an industry, very familiar with all that.

Our view, just we've taken a lot of internal initiatives around AI. Every one of Shipwell's developers will use AI tools, for example.

Chris

Sure.

Jason

Most of our departments, see like our implementations, for example, will use AI tools to try to make it better, faster, easier. Like these are all things we see adding value. And so from that, what we can see is that AI absolutely empowers people. It empowers them differently.

Jason

But everyone that's adopted has the use case of someone that's built something, you know, in a few hours that would have taken a week. before. And being able to intelligently apply that so that you're not creating more problems down the road, I think is a differentiator.

Chris

Yep.

Jason

And so, you know, I think that some companies will struggle because the marginal cost of creating a solution is much more approachable now. um But, you know, we're filming this and in February 2026 and you're right. FedEx and CH Robinson were hit by one announcement into it was hit by another one about accounting software CrowdStrike was hit with another one about cybersecurity.

All these companies, you know, there's a lot of volatility in market still, but essentially you're still going to need the same things you always needed. You need a moat, right? You need network effects. You need to be a system of record. you need to have a brand that people know and recognize.

And essentially, if you have those things, what you're able to do with AI, I think acts as a differentiator and less of a threat.

Chris

Yeah, yeah. and The whole to me, the thought of someone vibe coding a TMS just doesn't make sense.

As soon as something touches money and records and multiple customers and vendors, it is just the data structure is so much more complicated.

Jason

Yeah. Yeah.

Chris

And so I don't see it. But let me let me ask another question kind of related to this. um Do you find that the presence of AI makes it easier to get an implementation to sell or does it make it harder?

And do you find yourself having to explain what is AI, what is ML, what is OR to someone or is that gone? Do people understand that?

Jason

No, no. I've, you know, we have customers that are AI companies. And I think negotiating the AI parts of our or MSA are always interesting. Because I think everyone, even them, are still figuring out what what it all means.

I think we're in a phase where we've passed the peak hype. You can't just say the word AI and expect someone to sign the contract. But, you know, I think there are instances where people see a lot of value and they're innately very curious. You know, like if we look at what we've done with our track and trace agent, you know, we're routinely saving 70 to 80 percent of manual work that used to go into just tracking a shipment, finding the documents, sending emails.

Like there's lots of examples for this in transportation that, you know, will have further applications.

Chris

Yeah, you were just at Manifest, right?

Jason

Yeah.

Chris

So I was there two years ago and suddenly everyone was AI. Everyone was AI for it. And you knew it was just the marketing because they didn't know it. This was 2024. twenty four And so things had just kind of hit.

Has that changed? I wasn't there this year. Is everyone still an AI-based company?

Jason

I think a lot of people have, have added that label. Yeah. And I think what we're going to find is that there's a real differentiator, but between like, what does AI mean to you?

Is that building a graph for you? Is that, you know, writing an email for you or is it creating a whole, you know, digital coworker that you've trained that can work autonomously 24/7 to actually make your life easier?

And that when you wake up in the morning and you go into the office and you can see all the jobs it's done while you were sleeping that you're presented with a moment of joy which is rare in our industry.

Chris

Yeah. So um, I love a report that Anthropic came out with. Gosh, it must have been six, eight months ago, where they looked at, you know, AI, generative AI tends to be used for... One of two ways. It either augments what we're thinking and think about, you know, writes an email, it does something, expands, whatever, or it's about automation and makes something more efficient. And they said recently it was all about augmentation, but there's been more of a shift to automation.

If you think about those two dimensions, where do you find AI having the most value? You can't say both.

Jason

Yeah. It's okay. I won't say both. I think, so this is actually what I wrote my thesis at MIT about, was how technology changes traditional industries.

Chris

Oh, why first, first, wait, wait, wait, wait. You're a Sloanie. Sloanies don't do theses. Why did you do a thesis?

Jason

So, yeah, so ah I dropped out of between my first and second year of my MBA to start another company that went through Y Combinator, which is a startup accelerator in Silicon Valley. I sold that company and then I wanted to come back to MIT. And then they they said they wouldn't give me an honorary degree.

And so they said, you know, we want to plug you back into the matrix. And I managed to negotiate some classwork, but also writing a thesis to cover some of those extra credits.

And so my thesis wasn't specific to transportation, but it was specific to the idea of how does something go from a purely manual process to something that is fully automated? And I think we see a lot of parallels in technology with supply chains.

Right? Because there's a lot of digital trust. There's a lot of context you need. If I was to say, I can fix a thousand shipments for you, but it's going to catastrophically ruin 10, no one would be interested, right? There's so much risk asymmetry that's sort of associated with this. And i it makes sense.

And so I think what we see is that you need a high context system of record like Shipwell, all the structured data, all the context of of what needs to happen and what needs to happen next,

And then you need to very intelligently apply AI in a way that builds digital trust. So even when we talk about AI, you know we will use as little AI as possible. you know You need to have very defined binary if this, then that outcomes. So you have incredibly high predictability and then use AI very sparingly. And what you find is by doing that, you're able to create really good, sustainable, high high ah ROI outcomes that people then are able to build more and more on top of. And so I think in the early adoption phase, augmentation is a very natural step in automation.

And so having AI help you write an email and after it's helped you write a few dozen emails, maybe thousands of emails, you start to trust it. You say, oh, it actually writes an email almost as well as I do, sometimes better.

And then you can start getting to this world of like, would I let it send that email for me without reading it first? And that's that's really the the curve that I think we we see is now starting to like shift in supply chain specifically.

‍Chris

Yeah. That's a, that's a really interesting point because a lot of people thinks that they'll say, if I don't understand what it's doing, then I'm not going to trust it. But then you ask them to explain how a microwave works and they don't but because, because exactly what you said, Jason, you get a level of trust.

Jason

Yeah.

Chris

So it's more about repeated success. That's what builds trust for these things. And I like what you said about the, you know, the, there's a thought that you're gonna have one big robot and it's gonna do everything. And that that seems to be not the path that most companies are taking. 

I was talking with the guys from CHR and they have low agency. You want low agency agents. they They know what they're doing. You don't want them to write poetry as well. You want them to do one task and you put those things together.

Um, and the nice thing about having that also, it helps you the explainability because that is still an issue for a lot of things. Like why exactly was this carrier chosen instead of this, or why was this done?

Chris

And to trace that, is that becoming less of an issue, the explainability portion, or is that still an issue?

Jason

No, I think it's at the core of it.

You know, the way that we built our agents in our AI Studio is, you know, right now we can do track and trace, you know, soon we're going to touch settlement and sort of invoice payment and automation. 

But, all of our agents are no code, highly configurable. You can go into your dashboard. You can configure how you want them to communicate, what hours, you know, all these different settings that you can then set.

And then once you start running the agent, you have a full audit log to see what decisions it made and why it made them. And so, you know, I think that you're always going to be able to need to explain something just because our industry is run by exceptions.

Something weird happened. What happens next? And so you always need to have that at the core.

Chris

Okay, so you're 10 years in, right? You're having your 10 year anniversary, yay! And so what do you see in the next five years? What what is Shipwell gonna be 2031? in twenty thirty one

Jason

I think that's a really hard question to ask a company founder, because I think at no point have I ever thought, you know, about where we would be in such like a defined time period.

Chris

True.

Jason

And all my answers basically involve some version of global, you know, domination and relevance and those things. I think when we think about just the industry we're in and how it's changing.

You know, for the longest time, it wasn't changing much at all.

Chris

True

Jason

And now I think if we think five years out, it could be radically different. You know, we are in one of the largest industries in the world that is still one of the most manual, where so much of the value is derived not by firing people and cutting headcount, but like just the inefficiency.

And this affects both shippers and carriers together. And so for Shipwell, our focus is always just going to be finding customers, making them successful, ideally having them tell other customers about us. But by doing that, we're creating a network of networks.

You know, Shipwell today is connected to one out of every four trucks in North America.

Chris

Yep.

Jason

We'll process 100 million rows of location data. We will do terabytes on terabytes of like machine learning and all the AI. And as we think about what this means, the world is becoming more connected.

And so as we think about whatever happens, volatility, tariffs, people thinking more strategically about where they manufacture or source or procure goods from, all of these are decisions that will become increasingly more real time and not just a question that someone in the boardroom asks that then takes two weeks of scrambling to try to come up with a decent answer for.

And so as that future just has a very clear profit incentive and has all the technology it needs to come true, like our feeling is that supply chains become more real time, become more proactive.

And you get to this world where you're not so much just buying software, but you're essentially hiring a supply chain co-worker that begins to take this load for you. And so I think we're going to see a lot of excitement in the next five years.

Chris

Yeah, and what's interesting is the actual tasks have not changed as far as the truck shows up, the truck gets loaded, the driver drives away, those kinds of things. But you're right, the efficiency of the system has really changed. 

And so, and there's still so much left open for it. I have a running debate with some people. I don't think there's a driver shortage for this. I think it's an efficiency problem because of the wait time and all those things, which is a really, really hard problem to solve because no one owns it.

And it's easy to just make the driver wait because they have no power. So it's I agree with you 100%.

It's an efficiency game on how it goes. But what's interesting, the management of it is changing dramatically. Information, all that stuff. The physical movement. The one thing that has definitely changed, I would think, is in the cab of an owner-operator. They are much more sophisticated than people think.

They have smartphones. They see the market in there on their phone. It's different. If you talk to a trucker from the 70s about truck stop maps, they'd have to look to find where things are.

Jason

Yeah

Chris

So that is. Yeah, I have truck, man.

Jason

Well, this was DAT, right? This was ”Dial a Truck.” This is how it started.

Chris

Jevitz. Yeah, yeah.

Jason

Yeah.

Chris

So we've come a long way. We've come a long way.

Jason

Which is wild because I remember even when we started Shipwell, the idea that most truckers still had a flip phone. They didn't even have an iPhone back then. And so you know in the same way, I think I see it very similarly. Shippers and truckers both need excellent tools.

But the way that MIT teaches negotiation MIT teaches negotiation is that once you have that visibility, everyone has their clear preferences. This is still grounded the rules of physics. Where are you? Where are you trying to get to? Where do you want to go?

Then you can start to have a really constructive conversation, right? Price, performance, preferences. And then this in turn solves a lot of those problems.

‍Chris

Yeah, it's the matching game. Yeah, it's an interesting thing. Well, Jason, I really enjoyed talking with you about this and I look forward to see where Shipwell goes in the next one, two, three, four, five years.

Jason

Likewise. Thanks so much, Chris.

Chris

Thanks everyone. Okay, stay tuned for the Truckload Market Update.