Scott DeGroot: 0:29

I often talk about what I see a lot is dashboard theatre, a lot of great visibility, but it's not turning into action that optimises the next hour, or two, or three of flow.

Tom Raftery: 0:50

Supply chains have never had more data. But if that data doesn't change what happens on the warehouse floor in the next hour, what exactly is it doing for you? Good morning, good afternoon, or good evening, wherever you are in the world. Welcome to episode 135 of Resilient Supply Chain stories and strategies that keep business moving. I'm your host, Tom Raftery. In this special Resilient Supply Chain round table in conjunction with The Supply Chainer. I'm joined by Scott DeGroot, Executive Director of the Global Supply Chain Institute at the University of Tennessee and Keith Le Londe, VP of Systems at Path Guide Technologies. Between them, they bring the enterprise view and the warehouse floor reality. And that gap matters because operations run in seconds, while planning systems often work on an entirely different clock. We're gonna look at how to move from visibility to action, better decisions, faster responses, and fewer expensive surprises when reality refuses to follow the plan. Let's get into it. Gents, could you do a quick intro yourselves? Maybe Scott, you go first.

Scott DeGroot: 2:06

Yeah. Yes, I'm happy to do that. And Tom, thank you Keith. Good to see you. My name is Scott DeGroot, as you said. I'm the Executive Director of the Global Supply Chain Institute at the University of Tennessee, part of the Supply Chain Department, where I also teach graduate and undergraduate students, supply chain logistics, planning, and leadership. Prior to this, I spent 40 years in industry, 38 years at Kimberly Clark Corporation. So think about things like Kleenex and Huggies, where I was in charge of global planning and logistics for the company.

Keith La Londe: 2:38

My name is Keith La Londe. Tom, Scott, it's pleasure to be here with you two today. I am Vice President of Systems at PathGuide Technologies. Here at PathGuide. We specialise in warehousing and distribution software solutions to help distributors run the most efficient warehouses possible. And just excited to be here.

Tom Raftery: 3:00

Fantastic. And so from the enterprise side, Scott, where does decision making most often break down between planning and execution?

Scott DeGroot: 3:11

Well, it's a great question and we have a lot of experience with over a hundred different companies that we work with and that we're exposed to. That many times what's happening on the shop floor inside of the physical distribution centre or, fulfilment centre and what's happening inside of the planning system, those time horizons aren't really in sync in a way that provides the best operational answer because in operations, we have a clock that's measured by seconds. Trucks are flowing in and out, plus or minus a few minutes we're picking and packing inventory. The dynamic nature of that oftentimes is too much for an enterprise planning system to be able to ingest. And therefore it does not do that. And many times we have human beings that have to intercede between the direction they believe they're getting from the ERP or the advanced planning system and what's actually happening on the floor, today, right now, in this minute, and then the next few hours. And that is the gap. I think that's the gap we're gonna talk about today, which is a failure to execute flawlessly is ultimately gonna drive up cost and drive down service. And, then we all pay.

Tom Raftery: 4:27

Yeah. Yeah. And Keith from the warehouse side, where does the system view most often diverge from operational reality?

Keith La Londe: 4:35

Well, just to piggyback on what Scott had said there, absolutely what happens in the four walls of the warehouse or the distribution centre is completely disjoined often from what is happening in the enterprise system itself. And so when it comes to process execution, the ability to work with that data, leverage that data, and then drive the most, accurate processes to meet requirements that's where a specialised warehouse management system is going to shine on that. So, the disconnect there is, the physical requirements for moving equipment and moving personnel and inventory and stuff like that, that just exceeds what most ERP enterprise systems has the ability to do. They, tend to be really, really strong on the accounting side but, when it comes to the actual movement of material they lack integrations, they lack process options and so forth. And so, that's where we have found our niche as well.

Tom Raftery: 5:36

Okay, so Scott, where does better visibility still fail to produce a better decision?

Scott DeGroot: 5:45

I think ultimately in total logistics cost makes up eight to 9% of most of the global GDP. And a lot of that sits in the transportation flow. Transportation is always the most expensive part of logistics, almost always. But ultimately, we see a, it manifests itself. These added costs or poor service manifest themselves in delays on the transport side with the lorries and the trucks with underutilised cubic vans moving in and out, wasted space on the warehouse floor in staging areas and in pick areas, congestion in aisles, because we have either AGVs or human beings moving in and out of aisles in a way that's not optimal from a flow standpoint. And at the centre of all that, we have the customer, the consumer, someone who is paying the bill, right? And they're very interested and very demanding nowadays. They wanna know down to the, hour, when is their product gonna arrive at their doorstep or to their store. And so in that sense, we have this visibility is amazing. And that's only the first step though, until you take action on what you're seeing and prescribing what is the better next step in what to pick, what to pack, what to stage, what to load, how to load it, when to dispatch it. Until we're doing that, we're not really eliminating that waste, we're not really eliminating those costs. And so I often talk about what I see a lot is dashboard theatre, a lot of great visibility, but it's not turning into action that optimises the next hour, or two, or three of flow.

Tom Raftery: 7:24

All right, and Keith, then, what would be the earliest sign that a warehouse or fulfilment problem is developing?

Keith La Londe: 7:33

Well, when you begin looking at workflow and in what areas, what zones of the warehouse, what types of orders are being fulfilled a lot of distributors over the, past number of years, the model for them is changing with the explosion of e-commerce fulfilment. And so a traditional distributor might have, had the, the classic hub and spoke. You've got a large distribution centre, you're picking and packing transfers to your, stores and so forth. And, the whole dynamic of handling small parcel and things like that, that's really changed, the execution out on the warehouse floor. And so it's not enough to just know you've got a bunch of picks and orders, order lines that need to go out, but now looking at those, classifying them as to what types of orders are going out so that they can be handled because they need, separate processes to handle those as efficiently and, as accurately as possible. So, like Scott, mentioned you've got this, dashboard of stuff. Well that's great if you can see that you have challenges like, small parcels or overnights are, backing up and, the two o'clock carrier deadline is approaching. You've gotta get those. But can you actually, get people switched and have the system switch people automatically to make sure that those are getting fulfilled timely. Those are two, two different things.

Tom Raftery: 8:57

Alright. Okay. So Scott, how can a seemingly small operational mismatch become a wider inventory, logistics, or customer service problem?

Scott DeGroot: 9:07

One of the things that we all try to avoid is gridlock of the distribution centre because ultimately we have things in the wrong area. We have orders that go unfulfilled. We have labour that's underutilised, and we have trucks waiting empty to move. And so what our work needs to do is to make sure that we are understanding in a dynamic way near real time, what am I doing in the next few hours? What orders need to be picked? When do they need to be packed, staged, and ready to load? When are these trucks arriving? And then how do I match up my, labour force or my automation to the pick zones in the right area? And then, optimise that work as it changes throughout the day. New orders come in, trucks show up late labour doesn't arrive. All of these things have to be constantly adjusted for the realities of what is gonna happen in the next hour to five hours. And so if we're not constantly looking ahead to reoptimise the next five hours, we're constantly chasing our tails and trying to play catch up. And in this world where, as Keith said, the consumer has their phone, they're tracking the delivery, they wanna know, and why is it late? I had the inventory in my wrong space in my warehouse. Really? That's the problem. That's, that's not sufficient. So these are the problems that we get to solve.

Tom Raftery: 10:28

And to both of you, which supply chain metric do leaders trust more than they should?

Keith La Londe: 10:35

Oh, that, that is a great question. Which, can you ask that question again?

Tom Raftery: 10:40

Sure. I mean, we're talking about dashboards and I'm just curious, is there a particular supply chain metric that's showing up that, supply chain leaders go, oh, that's fantastic, whereas in fact that isn't, so what? What metric do they trust more than they should?

Keith La Londe: 10:53

What's interesting from my perspective. So Scott probably has a better answer on that, but still to this day, too many distributors are trying to decide what metrics they, they need to look at. If you look at like work studies and, and so forth looking at the, top KPIs, benchmarks that people are tracking, dock to stock, gives that measurement of how efficient inbound operations are. Still, we continue to, to encounter distributors who, know they need to begin looking at those metrics and they're just, they're not quite there yet. And so they're looking for a solution that can give them visibility to that. So, Scott, you, have more to add onto that.

Scott DeGroot: 11:33

I'm happy to add to it. I think that's a wonderful answer. And I'll just say that I think many times people create metrics that are self-serving to them. So here's an example. I see a lot of companies order to ship delivery to or line items filled like service. Okay? They ordered a thousand line items and I filled 999 line items. So I have a really high service. You filled it to the dock waiting for pickup. Did it pick up on time? Oh, I don't know. That's not from, well, wait a minute. Who's paying the bill here? And what's the whole intention of the distribution centre in the first place? It's to have things to sell, to create revenue. So this idea that you have a self-serving metric, only the cost that I can control, for example, or only order lines filled, ready for pickup. I think these kind of metrics sell short the real accountability of the supply chain and people in the warehouse.

Tom Raftery: 12:28

Mm, mm-hmm. Okay. And if we, if we talk a little bit about, truth, trust, and ownership. Keith, when the ERP and warehouse systems disagree, how should teams determine which one to trust?

Keith La Londe: 12:41

Well, we believe that you always trust the WMS, the warehouse system itself. That's where all of the granular transactions are taking place. You might have an inbound delivery scheduled. Does it get checked in? Where does it get unloaded in stage two? When did it actually begin getting received items, quantities, whether those are gonna be validated against the purchase order or an advanced ship notice for those distributors that are consuming and, utilising the, advanced ship notices. But, one of the, challenges that we see when implementing a WMS is that practises on the enterprise side need to change. So that the enterprise system can continue to manage the, costing and the inventory totals, but let the WMS handle all of the granular detail of where all those things are. And typically, if there's a discrepancy between those two, you know, if you've got the right tools, you'll be able to identify what happens. More often than not, what we see is that, somebody has gone into the enterprise system, they've done something that they probably shouldn't have done for a variety of reasons, so that they can, earn a higher rebate or commission or something like that. We've literally seen those types of things. And the WMS is the, inventory record. That's the system of record. And we're fortunate in that we stay outta the financial side of that so that the two systems can actually compliment each other.

Tom Raftery: 14:17

All right. Okay. And Scott, how much freedom then should local teams have to override an enterprise plan?

Scott DeGroot: 14:24

Oh, okay. So that's maybe you won't expect my answer. If we're ever in a situation where someone is actually overriding the enterprise backbone system, then I think we've put our people in a very bad situation. Because there ultimately, the enterprise system will likely be almost all cases the, system of record for financial reporting and the revenue accounting system and the order management system. Very few WMS has actually manage the interface with the customer. I don't know if Keith does, my apologies if, if yours does, but most of 'em say no. They take the order in and then they'll dispatch and allocate inventory and labour and dock doors and warehousing resources. And likely the system will also need to interface with the transport management system, the TMS, which is managing probably a spend that's twice as big as the warehouse. And so in that sense, I would not recommend that people override other systems. You have to put it in place, some sort of governance that allows this data to say synchronise in a way so that all of them can make the best decisions and we're not undermining one to the next.

Tom Raftery: 16:00

Alright, so is the idea then of a single source of truth more comforting than useful?

Keith La Londe: 16:10

I think that the single source, and Scott kind of hit on it, is, is you've got the checks and balances. And so, in the example where I mentioned if there's a discrepancy the proper way to correct that for example, inventory inconsistency is just through cycle counts. And so when I refer to, well, the WMS is clearly the system of record because if, an employee has put, 20 widgets on the shelf, we believe that there ought to be 20 there, right? And if for some reason the enterprise system is gonna suggest and believe that there's 21 or there's 19, and you've got this discrepancy, you clearly need to go through and, cycle count it. And so from a process standpoint like the way that we wrote our cycle count module for example, the assumption is is that the employee's gonna go there, there's gonna be 20. And if there's not, then not only are they correcting the WMS, but then the end result is that that inventory adjustment can be posted to the financial system with the appropriate reason codes, so the appropriate ledger entries get created and all of those, things. So it's truly a handshake and not a continual run of an imbalance where, some people believes the ERP. Scott is absolutely right, the financials, the ERP, that's the system of record there. And when it comes to where, where's the granular data at, the WMS can own that successfully.

Scott DeGroot: 17:42

I think that that's right, and I think it's reasonable in today's environment to have best of breed systems running the detail at the granular level of 10,000, a hundred thousand, 300,000 SKUs in a building that might have 1.5 million square feet of space. We want the WMS to do that. And to have some mechanism to bring together the millions of lines of data from customer files, from product files, from transport files, and to find a way to bring them together in a way that we do not have active ongoing misalignment between those data sets and everything today. And we can talk about AI and we can talk about robotics, but that all requires very, very sound master data management. And I think Keith did a good job at describing what's required.

Tom Raftery: 18:31

What then is more dangerous? Acting on imperfect data or waiting for certainty?

Scott DeGroot: 18:38

Yeah, I don't know Keith. I dunno. I've run a lot of warehouses. I've seen a lot of warehouses. You can never wait for certainty.'cause certainty is this elusive thing. You have to act within some guardrails, set by the business, or that you set for yourselves. You reconcile issues and every day mistakes are made. Thousands of them in the supply chain. That's our job to find them and to fix them and make the next day better. And so I don't believe you can wait for certainty, but you can never give up on working to get better than you are today.

Keith La Londe: 19:08

I would second that. And so, you need access to the tools to, correct those mistakes or flag those inconsistencies for follow up. So that any of those discrepancies,'cause it'll never be perfect. Certainty, you can't wait for that, right? You've got orders, you need to get'em shipped and so forth, right? But if a picker goes to, a bin location and, everybody believes that the product is there, and for some reason it's not, how easy can you make it for that situation to be notated so that depending on what the circumstance is, if it's a bin that needs replenishment for some reason, you didn't think you needed it replenished, but now you do, you've gotta get that situation taken care of. Well, what you'd really like is the pickers to stay productive. You might want 'em to skip the empty bin for the moment and go to a secondary or, queue it up until the, replenishment is done. Those types of things. So we put a lot of energy into you know, having it, at the fingertips, a keystroke of a scanner or whatnot. The ability to flag that to get that visibility to those situations that need to be corrected quickly.

Tom Raftery: 20:23

And Keith, if we're designing for action, what information should reach a frontline worker immediately?

Keith La Londe: 20:28

Well, the, the frontline worker, they're being directed by the work on the computer, the barcode scanner or the headset that they're wearing, right? And so, they need the information that is critical for that specific task that they are doing. So if you can picture the picker, and maybe they have completed one pick and they're ready to go onto the next one, and maybe they're taking their, cart with their, cartons and they're picking right into the carton itself, right? When that screen comes up, that needs to send them to, whatever the next bin is that they need to be sent to to see the item, maybe see a picture of the item, the description the quantity that they need. If it's like a bulk pick how many orders are they picking, at one time. All that information, you need that as clear as possible on the device so that they can easily find the right item, identify it typically by scanning the barcode, instant feedback that, hey, this is actually the right item, and that it makes it into the right box. And then onto the next pick.

Scott DeGroot: 21:33

I agree. And I think, Keith was describing sort of area of picking and, replenishment of pick zones and pick bays. I think the other thing we need to think about is the physical space and the movement of both robotics, AGVs, and human beings in and around the dock area, in and around the, unloading and loading zones. I know a story of a really large company whose name I won't mention outta respect for them, who installed a very expensive AGV system, and yet all the AGVs and the human beings constantly running into, not really running, but into the same space, slowing down the total productivity of this multimillion dollar investment because they're not thinking about the physical flow in the building at the time. So it's another area to be thinking about. The WMS needs to make sure that the workflow is, connected to the physical space in addition to just the inventory availability.

Tom Raftery: 22:26

I'm curious, all this optimisation we're talking about, how much of that goes back to some meeting at an Amazon boardroom someday where they said, you know what we should do next day deliveries.

Scott DeGroot: 22:41

Well, I, I would never underestimate seriously the impact that someone at Amazon made some years ago. And we can talk about Jeff Bezos, but we should also talk about people like Dave Clark, who built it out. The capability for the consumer to be driving the pace of, I'm showing my phone for a reason because the consumer drives everything. The consumer, myself, my wife this morning ordered some things. It's gonna be delivered at our house between seven and 10 this morning. It delivered. I saw it. So this is the expectation. And so this is what the supply chain and logisticians, warehouse operators need to build for. We're not gonna undo it, I don't think in the short term.

Tom Raftery: 23:24

Yeah,

Keith La Londe: 23:25

I, completely agree. We see exactly the same thing. As we're helping our customers stay as competitive as possible with the Amazons, with the Walmarts for fulfilment. And also being able to, emphasise whatever their specific value add might be

Tom Raftery: 23:46

Alright. Coming back to designing for action, coming back from that tangent that I went off on there for a second, Scott, which, which decisions should be standardised across the organisation and which should remain local?

Scott DeGroot: 23:58

The standard decisions across an enterprise depend in, in a region, let's just say a geography, either the United States or the Northeast or Chicagoland, but you know, within a particular geography that's relevant to where the revenue is occurring. We need to have some standard approach to customer prioritisation, to value prioritisation in the moment which orders, what items need to get priority when there's a constraint. We need to bring some standard approach to the idea of safety. Probably should have started with that, but there's nothing in my belief that in warehouse or an order that should ever compromise the safety of the worker or of the item or of the consumer. So, I think those kind of things should be standard. I also think that how we address mistake inventory discrepancies or order discrepancies or transport discrepancies. We're not even talking about freight fraud, but we could. Those things need to be standardised. I think human labour needs to be a local decision. I think we need to, obviously local the warehouse operator at the moment needs to understand the dynamic inside of the building in the next few hours. Who gets assigned to what work, who has shown up, who's capable. Those kind of decisions ultimately should be fully in the hands of the local operator, in my opinion. The yard, we're not even talking about the yard, but the yard is a place where a lot of unsafe things can happen. So let's make sure that we're managing this interface between drivers and warehouse operators.

Tom Raftery: 25:28

And if we try and talk a little bit about, what fails in practise, Keith, what sensible, let's say, during a systems workshop, but fails on a busy warehouse floor?

Keith La Londe: 25:42

Well, it is interesting. Scott talked about, the introduction of warehouse automation into warehouses. And often what can fail is when there is a lack of consideration for how an investment might actually impact the system's in place. For example you've got a functioning warehouse and a lot of our distributors, multiple warehouses, but none of them are identical, right? If you're Amazon, you're building a lot of, identical, buildings, but they're each unique. And so what might work at one site might not necessarily work exactly the same way in another site. And so the last thing that, a distributor would wanna do is, come back from ProMat or something like that, see all this wonderful equipment and decide that they're gonna buy a conveyor system. Well that's wonderful and it's gonna move the tote from this zone to that zone. How's it gonna impact the processes? What are the data requirements? And does the WMS will that easily support it or, or not? And we've literally seen examples of, people deep into a project where, they failed to realise that, hey, how's this going to impact the system? And so, that would be a classic example from my perspective.

Tom Raftery: 27:03

Okay. And, and Scott, if, if I ask you same thing, you know what looks sensible in a planning meeting, but fails when commercial and operational incentives collide?

Scott DeGroot: 27:12

Well, I agree with Keith, and that's the number one thing, the failure to understand the interface between automation and human beings. So, but I'll pick another one to build on it, and that is the failure to respond to the dynamic nature of external in outside of the four walls of the building. What is the transportation flow in and out? What is my labour availability? What is the price? How much do I have to pay for an hour of labour versus all the new warehouses that are being built around me? Another provider shows up, maybe one that we just mentioned a moment ago, and they pay a dollar hour more, and all of a sudden you don't have 20% of your labour is gone. I think it's these, kind of things that often don't show up when we're doing a conference room pilot. And probably the one that's most important is the reliability of transport flow in and out of the site because it is all about connection with the transport.

Tom Raftery: 28:06

And what have organisations implemented that created more information but no better execution?

Keith La Londe: 28:16

Well, a, a couple that come to mind. We've had some that began investing in automated storage and retrieval systems, ASRS, and you are increasing the density in the warehouse and then relying on the robots and the shuttles to bring the, product. Clearly you have to have a solid interface between those two systems in order to be able to make that work, to bring the goods to the picker in that way. There's a lot of, data that's available, to be analysed for the storage of that data. AI can be applied to that so that the shuttles internally are continually evaluating how often the WMS is requesting that inventory be brought down and so forth. And so there's a lot of data there that could, be analysed. But kinda like we were first talking when we, first started, you have to be careful about, too much data and trying to analyse it too much. Just for the sake of, analysing it.

Tom Raftery: 29:15

And Scott, where have you changed your mind about real time data or automation?

Scott DeGroot: 29:23

Yeah, I, I will say, and anybody who knows me will know that I I'd like to believe and have believed in the past that with the advancement of mass data ingestion engines that you can really build an advanced planning system that can see everything so that when I'm deciding on production or procurement or demand allocations or inventory flow between sites and buildings and consumers, that I could take into account all of the possible scenarios and actually have, in my mind, in operational and practise a real time view into the entire world, okay? Even though that's might be possible still in the future. And I don't ever discount human ingenuity. I also have learned that near real time in many cases is enough. And what I really need to know to run an efficient warehouse operation is what's happening, the next 24 hours and through the next week as I do labour planning. Then what I need to do, inventory and demand planning, I can look out 10 days that I really then need to segment the data in things that are operationally important in the moment. So the fact that I have 135 inbound trucks coming to my building today, that's good, but I really only need to know which ones are coming in the next 12 hours and what's on those trucks, and how am I gonna allocate unloading capacity and storage capacity for that stock. That's what I need to know right now, and that's the data that I need. Otherwise I'll just drown in an ocean, right? Drowned in the ocean.

Tom Raftery: 31:05

And, and, how can leaders tell whether decision making improved rather than reporting merely becoming more sophisticated?

Scott DeGroot: 31:14

Yeah, I do think the metrics help here, and not just backward looking metrics, but operational metrics that are building. I mean, obviously you're gonna be measuring some level of service orders filled or cases shipped, or OTIF maybe. Is that improving? Do you see that coinciding improving in revenue? You're gonna see distribution costs per hour, or picks per minute, or transportation dwell time. These kind of operational metrics should be improving. If not, then maybe you did not take the right action with the data. Certainly, you know, you need to be measuring the, performance of the warehouse against some sort of standard that you're improving upon. And if you have a hypothesis that if I instal this WMS, my inventory accuracy will improve by X, did it or didn't it? My productivity of warehouse worker picks and pack moves per hour. Whatever you measure, did that improve by the 20% I expected? If not let's reallocate, and let's change our hypothesis. And so I do think you can look at the operational metrics for improvement. And if you don't see it, then maybe you did the wrong thing.

Tom Raftery: 32:21

Alright, and Keith, what changes on the warehouse floor when people genuinely trust the system?

Keith La Londe: 32:28

Yeah, the main thing that we see is that productivity across the workforce improves. A classic example of that might be somebody who is transitioning to their, their first WMS, right? It is imperative that data in, the WMS be so that as employees, walk to the bin to do their pick and walk to the next one and, and so forth, that when it is accurate the people fully embrace the system. They begin following it, following the recommendations and believing the, reports that are generated out of that to be able to make smarter business decisions. That accuracy just allows people to really perform better and have an increased higher morale out in the warehouse, because the employees are no longer firefighting all day long because by and large things are where they're supposed to be. And what this also allows is those workers who really just want to come in and, be recognised for their contributions. with a strong WMS and accurate data and accurate reporting setting to those KPIs, people's contributions to the organisation they're, better represented, right? Because I can just come in and we have a classic example of somebody who implemented employee metrics and, and scorecards for the first time. This particular distributor was not quite 24 7, but 22 hours a day. They had an active workforce on the warehouse floor around 110, 120 employees all throughout the day. And way back when they implemented and they began looking at this, in a workforce that large, if somebody called in sick or somebody quit, they would just pick up the phone, call the temp agency. This was when you could actually get labour and stuff like that. When they implemented metrics to begin looking at who's actually doing the work, one of the things that stuck out was first off their best guy, always running around on the fork truck actually didn't turn out to do very much, right? Always looked busy, but transactionally didn't actually do very much. But what was more of, a shock and an eyeopener to them is that when they began looking at the data, some of the individuals that, you know, hey, if they don't show up next week, no big deal. We'll just hire somebody else. Right? Many of those employees turned out to be some of their highest productive workers. They just weren't the ones that stood around the water cooler socialising, right? They just wanted to come in, clock in, just let me do my job. I'll clock out, go home. And they turned out to be some of the highest productive workers. So the whole morale side of it, recognising people for what they do, believability in the system so that, hey, I don't have to sit and wonder if it's telling me to go to bin, ABC 123 for this item. And I, I happen to know because I put it away yesterday. It's over there. They just believe the system and, and, and follow it. And so it's really a win-win for organisations.

Scott DeGroot: 35:39

Keith, do you mind Tom, if I build on Keith's answer because I'd like to, and that is absolutely, I make, most people would put in the WMS if they don't have one or improve the WMS to ultimately lower labour costs and the operating cost of the dc It's efficiency, against some sort of engineered standard. Coming along with that is there should be a substantial improvement of quality. Load quality, product quality because the WMS is the truth and is directing the right people to the right bins, to the right truck and so forth. So load quality and totality. And the, and another thing will be safety on the dock floor because you don't have the person who's just driving around looking busy, but not really doing anything. Maybe not even following the standards of good work practises. And so when you see safety and quality improve and cost go down, all of those things are generally people would be happy for all three of those outcomes.

Tom Raftery: 36:34

And we've skirted around AI a little bit without getting into it in any kind of depth. So let's dig into it a little now. So for both of you, where can AI improve operational decisions today without creating unacceptable risk?

Keith La Londe: 36:52

Well, from our perspective, AI will be able to help give easier access to the vast data that is being captured by the WMS to make operational decisions. Like, looking at trends. Typically on a, Tuesday in the afternoon, zone three tends to get heavy, so don't wait to put people in it and so forth. So, with the WMS, you have this vast amount of data. Looking at it from a forecasting perspective we rely on the, enterprise system to manage all of the buying and stuff like that. So that we just simply have the visibility of, what containers are coming, what trucks are coming, what we believe may happen. AI can handle the likelihood that, that would handle. But from our perspective, helping in the organisation of the warehouse, the slotting, the reslotting and, and so forth. There's a lot of data there and AI will help be able to trudge through it and make it easier for people to, from a voice perspective or, or typical Google search, help me optimise my warehouse and not have to run a lot of classic report wizards. Next, next, next. What would you like to do? It'll be able to handle the, translation between those two.

Scott DeGroot: 38:10

I agree. And well, maybe just to add on I've seen many, several companies now, more than several, 20, 30 companies now who are building a bridge between what they might model in some sort of quote unquote digital twin view of their warehouse flows into operational recommendations that can be reoptimised. Now maybe you can do this too much, but almost on, a daily basis. What demand is attached to what DCs, which orders for what items are flowing to what queues. And then inventory zoning, and warehouse density and workload optimisation is all being done, in quote unquote near real time for the next three days. Your warehouse in Holland Street, in South Chicago, you're gonna have these orders going to these places. And this is how you should zone your warehouse. And putting that right into the operational WMS. And then operational WMS, then reporting back through AI feeds and agentic agents reporting back to this was wrong, this was right. Next time, let's reoptimise these transport flows so this inventory doesn't get stuck in trailers in the yard. These are examples of what people are doing already to using AI, or Agentic AI to bridge the gap between the operational reality and the ivory tower ERP systems.

Tom Raftery: 39:29

Alright, time now folks for the lightning round. So this is where I ask you short questions and you give short one sentence answers. So ready for this?

Scott DeGroot: 39:42

I'm ready.

Tom Raftery: 39:42

All right, Scott. Speed or certainty when disruption hits?

Scott DeGroot: 39:48

Speed.

Tom Raftery: 39:50

Speed. Okay. Keith Central rules or local judgement? Keith La Londe: Local judgement Very good, Scott, more data or cleaner ownership?

Scott DeGroot: 40:03

Cleaner ownership.

Tom Raftery: 40:06

Okay, Keith, what should never be automated?

Keith La Londe: 40:10

Personnel related decisions.

Tom Raftery: 40:13

Very good. Scott, what metric do leaders trust too much?

Scott DeGroot: 40:20

Their, their service metric, they don't, fully understand it.

Tom Raftery: 40:24

And Keith, final one, visibility without action is what?

Keith La Londe: 40:31

I'm trying to think of the word, visibility to a situation and failure to take action on it? Poor management.

Tom Raftery: 40:39

Okay, Alright, we're, we're, we're wrapping up now. So for, let's say, look, looking into the future a little bit, over the next three years, which supply chain decisions will change most, do you think to both of you? Scott, first maybe.

Scott DeGroot: 40:57

Yeah, I, I think the transactional activity that is so often consuming, a good portion of our administrative staff, I think that should be going away because of the, amazing capability of agentic AI to do things like freight bill matching, order zoning, data cleansing, transportation bidding, labour scheduling, labour scheduling. How much time do warehouse managers do labour scheduling? Right? AI is already doing that. A lot of companies that I know are using it in a way that just freeing up huge amounts of time for their staff. And I, I think these transactional activities should go away, quite honestly, because they're not really intellectually fulfilling. They just take up time And allowing us as advanced human beings to, to work on more strategic problems or to think about more important questions.

Tom Raftery: 41:50

And Keith, same question.

Keith La Londe: 41:52

What continues to change are the requirements being placed on our customers by their customers. So whether that's the tracking of data if you are a aerospace supplier, right? Whether you're supplying, fasteners or building assemblies or things like that, it is just a continual expanding data necessity track vital attributes of, of inventory where that's gonna be, shelf life heat numbers or things like that. And the changing requirements that the customers have for that. What originally started as just, simple requirements from a labelling perspective is now, cradle to grave lot tracking. Being able to, see what inventory came in, where it went, who did it go to, and using the aerospace example, ultimately what tail number of an aeroplane did something get used in, or a piece of equipment get installed in. And so those requirements have only sped up for what customers are looking for. Access to data, access to APIs to be able to retrieve it themselves. Or ultimately, and, and Scott held up his phone earlier where customers are driving the requirements that, distributors need to, meet. And I think I mentioned, we, help customers try to, do business and stay competitive against, the Amazons because they're the ones that, help shape that customer experience. And our customers too need to be able to, as efficiently as possible, pick, pack, ship, fulfil those orders so that the customers receive that, thank you for your order, your order is shipped, here's your tracking number. And, and their thought is, wow, that was fantastic, because that's what we've all come to expect. And what started as, just a consumer requirement these days it's, across the board, even in, industrial distribution, mid-size distribution, what have you. People, if you're not able to, provide that same level of, customer experience then a person is left wondering, why is it again that I'm choosing this person to do business with versus somewhere else. So you need to be as impressive as possible in that level of service.

Tom Raftery: 44:14

Okay, great. And for people who are listening, if they want to improve their operations tomorrow morning, what's the first thing they do? Scott first and then Keith.

Scott DeGroot: 44:28

I have to understand where their losses are. Where are they losing revenue by not fulfilling the service level commitment. And number two, where are they losing margin by having costs that are higher than they should be, either through transportation, warehousing, or, storage costs. And so, having this sort of zero loss mentality and understanding where things are leaking, both on revenue and on margin is the first thing that a good operator needs to do.

Tom Raftery: 44:54

Keith.

Keith La Londe: 44:54

I would just add to what, Scott said thoroughly understanding the requirements. What is it that we are trying to achieve? And any project, whether it's a complete WMS implementation or the addition of, some warehouse automation, right? Why are we doing this? What are the objectives and what costs are we willing to incur in order to achieve the objective. Because every project is a trade off of how many, process steps do you want? What more steps often is, a more cost in order to do that. And what is the accuracy that you're, getting for it? And if somebody is like, absolutely, we cannot make mistakes but, I'm not gonna double check, orders that, might be an example of that. Or we're not gonna weigh the boxes and, and capture the, the weights to make sure that what we think is in the box is actually in the box and things like that. So there's this trade off. You have to know what you're trying to achieve, and then you need visibility from your systems, from your WMS to say, Hey, these are the bottlenecks that, we are seeing. And if you're gonna add steps or take away steps, how is that going to impact your accuracy or your productivity level? So, you start by making sure that you understand what it is you're trying to achieve.

Tom Raftery: 46:20

Alright, Super gents, that's been really interesting. If people would like to know more about yourselves or any of the things we discussed on the podcast today, where would you have me direct them? Scott, you go first.

Scott DeGroot: 46:33

Yeah, we'd love to see all of you come and join us at the University of Tennessee. You can go to University of Tennessee Global Supply Chain Institute on any browser, and you can find out a lot more about our programme and about me personally. And I would love for you to do that.

Tom Raftery: 46:50

And can people do remote courses there, Scott?

Scott DeGroot: 46:52

They sure can. We, we have a lot of asynchronous adult and executive education for everyone.

Tom Raftery: 47:00

and Keith, people like to know more about you or any of the things we discussed.

Keith La Londe: 47:04

Absolutely. So, I, my profile's available on LinkedIn. That's Keith La Londe at PathGuide Technologies, as well as our corporate website PathGuide.com. And link there to be able to, to reach out to us or even just give us a call and you'll get steered. Tell 'em you're looking for, Keith, and, and they'll track me down.

Tom Raftery: 47:26

Super fantastic gents. That's been really interesting. Thanks a million for coming on the podcast today.

Scott DeGroot: 47:32

Thank you. very much.

Keith La Londe: 47:33

It's been a pleasure. Thank you.