Vikrant Host

Good morning, good afternoon, good evening, everyone, depending on what time of the day and which place in the world you're listening this from. And thank you for tuning in again. Welcome to Execution at Scale. As you know, this podcast exists because most conversations happen over coffee or a beer or a wine and not in a conference room or up on the stage. And today, again, I have Dr. Jim Tompkins with me to complete the three series episode. And I'm starting to think maybe we'll add another three given our conversations never stop. But welcome, Jim. Jim, as you know, is chairman of Tompkins Ventures and recently just founded Tompkins Logistics as well, which is doing amazing, which we'll talk a little bit about in this as well. So the first episode took the long view. The second one was the 50 years, 15 businesses, what we talked about. This one is right now, like what is happening in Jim's life right now. And what is happening in logistics in parallel, because Jim is all about supply chain and logistics. And Jim believes the moment calls for a different approach entirely. So we will hear from Jim today. Welcome back, Jim.

Jim Tompkins Guest

Thank you, good to be here.

Vikrant Host

Absolutely. So we'll go dive in straight into questions, I guess, since it's a three part series. So my first question to you is like, we have talked a lot about supply chain, logistics, technology, automation, and the challenges leaders face today around like, you know, the couple episodes we did. Now, how is the typical supply chain leader holding up under this pressure?

Jim Tompkins Guest

Well, not not real well. Most supply chain leaders are exhausted. They're confused. They're not able to keep up. They don't have the correct tools that they need to do their job well. And they're frustrated that they keep working harder and harder. And they can't keep up with VUCA, volatility, uncertainty, complexity and ambiguity. And they're dealing with nonstop disruptions. And these disruptions, you know, they might be, you know, work from home, they might be tariffs, they might be transportation costs, it might be labor. They're dealing with a rising customer expectations. They're dealing with technology complexity. And the problems they have are growing faster than they're able to absorb. And they so they feel like they're kind of in quicksand and they're sinking. And the quicksand is getting up around the shoulders. And so they're not feeling real good now. They're actually trying to run networks that require real time decisioning across thousands of variables. But they're doing it with spreadsheets and dashboards and meetings. And so the pressure that they're in, it's not a relationship to their talent. These are talented, good professionals. I'm not talking about the lousy operators. I'm talking about the good operators. They're sitting in a situation where what they have is a model of doing business that is no longer relevant.

Vikrant Host

All the things around like Virgin IC is like a little bit smaller, maybe like, you know, because you are looking at the overall, like, you know, supply chain worldwide, right from the view you said that, I said more like close to the 3PLs. And it's pretty much ever like, effort is never the missing piece. It's more about, as you said, acceptance, finding the new technology, having the resources. So I totally agree with that. So that kind of like, makes me wonder and think that what do you think supply chain executives are so troubled right now, like what it is?

Jim Tompkins Guest

Well, they're accountable for things that they can't control. And so that that's where the frustration comes in. They're responsible for cost. They're responsible for service, for resilience, for labor, for transportation, for inventory, for customer experience, but they don't own the assets. They don't have the assets. They don't have the data. They don't have the technology stack. And so they're orchestrating a network without having an orchestration system. And you go, oh my goodness, you can't do that. And so each month that goes by, they find more volatility, more skews, more channels, more constraints, more customer provinces.

Vikrant Host

Now the next thing is that, okay, like, you know, these are some of the problems which we discussed. Now you talk to so many people every day, listen to so many people like me every day who come to you with their problems or solutions. So how can, how can executives deal with being this overwhelmed? You know, and that is a personal question for me too, because I feel overwhelmed, like for sure. And that proves that in a two hours of sleep and 10 cups of coffee a day. So how to deal with this over?

Jim Tompkins Guest

Well, there's, there's two things that executives can do to start getting control of the cyclone they're living in. First thing, and this is something I've learned over the years. This isn't something I've just learned in the last six months or last year. This is something that I've learned by watching small companies and large companies that have struggled when they're overwhelmed. One of the things that I think we need to do is we need to ask ourselves the question, are we the ones that should be doing this? Is this really a problem that we're positioned to handle or is this a core competency? Is this something that we should work on or should we pass on this opportunity and do something different? Most successful companies are brilliant at innovation and product development, customer experience, brand building and growth, but logistics? Well, logistics is extremely important and it's essential for success, but is it a differentiator? Should logistics be a differentiator for our company or is this something that we should say, logistics is not core for us. And so what we want to do is we want to outsource it. And I think more and more companies are reaching that conclusion because they simply can't, at the CEO level, have a distraction like logistics pulling the whole company down. So the best thing is if you can't handle it, give it away.

Vikrant Host

I could not agree more, you know, it makes me remind a couple of things. One is like, you know, which all of us know the Nokia story, right? Like Nokia was the best big giant and I was actually at a client site last month and I saw this poster there, which they had like saying, don't be Nokia, right? So it's about adapting. As we were just talking, we are living in an ever-changing world and what I feel is that we really don't understand or have the time to understand what we need to do because we don't have time. Like, you know, as you said, like if CEO or the like C-suit management is busy with that firefighting, trying to fix, add more people, do this, get more software. Oh, things are not working. Let's change our WMS or let's change our OMS or whatever that may be.

Jim Tompkins Guest

Well, you know, the fact of the matter is 3PLs and 4PLs were invented to solve this problem and there's a lot of good 3PLs and there's a lot of good 4PLs where when the world was linear, predictable, and able to be managed by humans, 3PLs and 4PLs did fine. In fact, 3PLs still can manage their node and do a very, very good job. 4PLs can manage the transportation they're responsible for, the warehousing they're responsible for, and do a good job. The problem is that the traditional model is insufficient because number one, the current 3PL and 4PL models are provider-based models. The decisions are made to optimize that 3PL or that transportation provider. We don't want to optimize the provider-biased link. We want to optimize the customer's supply chain. It's not directed at the right problem.

Vikrant Host

To your point, I was just thinking this recently that a lot of 3PLs who I work with, they're kind of doing 4PL work without even signing up for it. Think about handling exceptions. Think about a container comes in, but it has to sit on the door because inspection is not done, vendor is not cleared. There are so many different ways. It comes in as a scope creep for the most part because these things were not predicted because as you were talking about VUCA earlier, and where this affects even the provider is that they never predicted it, scoped it, and in most cases, guess what, they don't charge for it. The end customer is like, okay, it was handled, although it was a messy way of handling, but I didn't pay anything, maybe, or it was done. What I'm curious is that we all know 3PL has been there, 4PL has been there, as you said. Is this what you see as a new category, or is it more like 4PL done properly?

Jim Tompkins Guest

This is a new category. We're not trying to solve the same problem. The complexity of today's supply chain has outgrown the traditional operating model. Autonomous orchestration is the shifting from a human-driven coordination to an AI-driven decisioning across the entire logistics network. You say, yeah, they can solve the problem. Yeah, they solved the problem based on where they were standing and what they saw. They took the best action possible. You and I couldn't have thought of a better thing for them to do. However, they didn't tell anybody, and now we have that problem going to the next node, the next node, the next mode, and it's going to bounce. No human team, none, zero, no matter how talented they are, they can't process the volume of data, the variability, and the real-time constraints required to optimize a network across several thousand nodes. They say, well, no, that used to work fine. Well, yeah, it used to work fine because everything was steady state, and it ran how it ran, how it ran, so there was no changes. So you didn't need autonomous decisioning because there was no decisions to be made. You just did what you did, and it kept going.

Vikrant Host

So if someone is listening to this and who has no idea about autonomous supply chain, for example, like in a layman world, or like, you know, how would you explain it to, let's say, an eighth grader, for example, like in a simple way?

Jim Tompkins Guest

Well, I think it's important that we get some clean words out there. The clean words I like is provider-neutral orchestration. So provider-neutral orchestration. The thought process is so new, I can't remember how to say it if I don't look down on my notes, but I got it written right here in big capital later. So what does provider-neutral orchestration, what does that mean? Well, what it says is every provider is treated equally. It's no more about those tickets to the NFL game. It's no more about the nice Christmas party, okay? What we're going to do is make decisions based on performance, cost, and constraints, not relationships. Okay, now people say, well, relationships are critical in business. Yeah, they are, except when it costs us more and when it provides us constraints that we can't deal with. And so what we want is the customer to be totally in control on establishing what their network is going to do, and we want them to decide that not in real time, but before the things occur, and we have a strategy as to how we're going to make things happen. The system optimized the entire network, not individual silos. So what we do is we use AI.

Vikrant Host

So if I'm hearing that correctly, like it means that now we have capability to advance, I guess, like, you know, what we were doing, like maybe on spreadsheets years back, and then multiple softwares. And now, with agentic AI, and all the advancement in technology, we have emerged into a new category altogether. Is that your right statement?

Jim Tompkins Guest

Absolutely. 100%. 100%.

Vikrant Host

And, you know, Jim, like, as you know, that every time we talk, like, you know, and we always talk about like, that technology is not a fix of bad process. So I don't like that's kind of a curveball here, like switching from like, you know, the tech side, but like, how do you see this as an important that like, you know, as you're adding advanced technology layer on top of, let's say, like, if the processes which they already have, what about the data, if the data is not clean? Like, is there any thought process around there, like what you're seeing out there?

Jim Tompkins Guest

Well, so what Tompkins Logistics is, it is the autonomous orchestration layer. And the autonomous orchestration layer has the ability to fix the data. If you put 29 items in a truck, and when you get down the road, you cannot deliver 46. Okay, it doesn't, unless I guess maybe you're shipping rabbits or something. But I mean, you know, you so if it knows what it had before, and it knows where we are now, if there's a difference, there's a single pane of glass that says we all are looking at one version of the truth. And if that truth appears to be untrue, we now know where to go to fix it, because we know is 46. It was 46. It was 46. It was 29. Wait a minute, what happened between here and here is a 46 to 20. Oh, no, no, no, it was 46. But they put two cases back in the wrong spot, and they didn't know it. And so it's going to give you the information to allow you to clean the data. So what Tompkins Logistics does, it's an AI native decisioning engine that sits on top of all of the warehouses, your warehouses, your 3PLs. It sits on top of all of manufacturing, your manufacturing, your outsourced manufacturing. It sits on top of all of your transportation, your private fleet, someone else's final miles, someone else's LTL. And what it does then is it has full knowledge of the entire logistics network. And it does this without having to tear out and replace. And so if you're using a WMS, and this guy's using a different one, and this guy's got a TMS, and this guy's got different, this guy's got a yard manager, this guy's got different, we got all the inventory, we got all the orders, we have everything that's out there as it is, but we're now putting an orchestration layer over the top. Looking at the entire network is the key.

Vikrant Host

And I mean, it generally like, so I think like we have a very clear understanding of what is Tompkins Logistics, what a client should expect from it. Also, like, you know, autonomous orchestration. I think that's like the biggest piece. And also, as we talk about the provider neutral orchestration, along with being autonomous. So it kind of makes me wonder also, one thing is, what about demand planning? Is that somewhat part of this as well? Because we briefly talked about demand planning in our last episode. So I was curious, because we both kind of agreed that that's probably the most difficult piece of the entire puzzle. Because once we know, okay, we have to take 45 rabbits from A to B versus 26. Like, you know, we can figure it out. But what if, what about like the future?

Jim Tompkins Guest

So we definitely are on top of the demand planning that's given. And we in fact, have a whole AI routine that looks at the accuracy of that demand plan. And we can go back to them and say, your demand plan is not good. And here's what it's doing to us. This is the cost that it's creating, you need to improve your demand planning. And it can highlight that. Now, it won't necessarily fix that. But what it will do is say, hey, because your demand plan was so bad, we had to spend an additional $40,000 last month to ship and reship stuff that we thought we were doing it according to your demand plan. But your demand plan wasn't right. Or we've got way too much inventory, and it's filling up warehouses that we need that space to do other things. And so the autonomous orchestration does not, quote, do demand planning, but it evaluates your demand planning, and will help you get it right if you don't have it right now. And so it's very cool.

Vikrant Host

Yeah, absolutely. I mean, I totally agree. I think like making that visible by itself is so powerful. And like, we do something very similar on the WMS side for labor, right? Because as a 3PL, like, you know, they're planning that, okay, like, hey, we need these new people, busy season is coming, so on and so forth. And a lot of it is based on visibility, as you said about lead time. So like, 20 containers are expected to come like today. But what if they didn't, you know, they're coming up next month. Now, generally, from what I've seen is that, as you said, like, you know, someone goes through a list and says, okay, like, what's going wrong there. And this is just one example, you know, and I'm sure people agree on this, they're like, thousands of different, smaller pieces, you know, like, what, what if there was a quality issue? What if what happened, you know? Oh, like, you know, there could be a poor congestion, or we have all seen COVID, like, you know, there's so many different ways to like, go about this. But like, I think the biggest thing I'm hearing from you is that now with AI, poking all these small pieces, and finding that these four pieces might be at fault, even if we don't say that they are the fault, they're the solution, but like, identifying those four pieces have become much more easier. Or I would not say easier, like probably like consumes less time.

Jim Tompkins Guest

Yeah, and you know, back to your example, demand planning. We don't have a quote, demand planning algorithm in our software, because you already have that. But what we can do is say to you, I mean, there's, you know, three things basically wrong. Number one, you got the lead times wrong. Number two, you're not properly looking at seasonality. Or number three, you got a real problem with people doing promotions and not informing the system. But it's almost that's there's 95% of the problem. And so what it's going to do, and it's going to go in and look at the lead time and say, I couldn't you haven't updated the lead times in six months, man. I mean, that doesn't work in today's world. You know, you got to make that happen. And then you go, Oh, boy, I guess I better fix that before I get fired over here.

Vikrant Host

I totally agree. Like, we cannot wait six months to update lead time anymore. It's probably every six minutes.

Jim Tompkins Guest

And like, yeah, exactly. Six seconds you have to

Vikrant Host

Yeah, yeah. Which again, may or may not be humanly possible. So I totally might be a little bit exaggeration, but that's what I see maybe in the next four or five years. There are times like everything is so real time and bots are talking to bots.

Jim Tompkins Guest

Yeah. Well, if Hey, if the lead time does you haven't looked at it in six months, and we're looking at we are out of inventory, and we've got $20 million worth of orders for this product. Guess what, I'm going for the airplane, you know, you know, polish up your transportation buzzer, because we're going to kill it. But we're going to get some product in here. And we're going to take that lead time. And we're going to cut that down to about three days, no matter where you are in the world. Yeah, make less margin, but still make the sale.

Vikrant Host

I mean, like, you know, I was talking to like this one of our customers, they sell like puzzles. And it was the same conversation because like around the holiday season, like as you said about seasonality, so they have problem with that, like, you know, but they sold out what they were thinking to sell by end of October, they ended up selling that by end of August. But their current lead time is six months, but based on like, you know, the speciality work they do. So technically, they sold out whatever they predicted for the entire holiday season, because these are puzzles. But now, is that a good thing or a bad thing? In a way, it is a good thing that yeah, like, you know, you sold what you wanted to sell, like now it's sit back on a yacht and relax or whatever. But can, as you said, like, you know, how we can expedite that lead time. That's a very important piece, because like, it's not only about revenue, it's about like, losing on that potential, if I can sell $5 million more for the next two months, and maybe even if I lose $200,000 in it, no problem, but like, still maintain my customer base, and those customers will come next year, I'll get more predictability. I think there's always two sides of a coin, as we all know.

Jim Tompkins Guest

Yes, everything is connected. Network, the whole network, not just one.

Vikrant Host

But I think like, you know, Jim, like that brings me to like, I guess, end of my questions for now. But I'm sure I will have more, which will call for another episode with us. But is there anything, Jim, which you feel that I missed in my questions, or you would want someone listening to this for the last two minutes, a key takeaway from this?

Jim Tompkins Guest

Well, I guess what I would want to emphasize, and I think we've covered it, but when we do this, when we get ourselves really thinking network, and really having the information, and the autonomous decisioning, what does that allow us to do? And what are the outcomes? I think there's three outcomes. Number one, we get better performance. We're going to have better on-time delivery. We're going to have fewer exceptions. We're going to have faster cycle times. And so it's going to perform better. We're going to have better from a cost point of view. We're obviously going to have reduced transportation spend. We're going to have better labor utilization, because we're going to know what we need before we need it. And we're going to have lower inventory costs, because we're going to refine that as we were discussing. And then thirdly, it's going to have an enhanced resilience. We're going to have fewer disruptions. We're going to have faster recovery. And we're going to have predictive rerouting. And that's going to give us the ability to still stay afloat, even though we're taking on water. We've got problems, but let's respond to those in a way that we can make it happen. So I guess in today's world, control towers report problems. Autonomous orchestration prevents them. And that's the big thing we're bringing to the table, is we are now avoiding the problems by getting ahead of it and making sure all nodes, all modes, all shipments are optimized in real time.