Podcast | Technology & Business Solutions (TABS)
Ep. 5: “Automation in Action: The World at our Fingertips: How Automation Fuels E-Commerce”
In an increasingly online post-COVID world, e-commerce accounts for a growing amount of all commerce, and customer expectations are higher than ever. For shoppers, the e-commerce world means any item is available at their fingertips, ready to ship in the blink of an eye. But the myriad steps involved along the journey are tremendously complicated.
In this episode, we hear from Ron Kyslinger, Advisor to THL, and Marc Austin, Chief Solutions Officer at Fortna. Together, we’ll follow the journey of automation from the moment a customer searches out an item, examines ratings and reviews and adds it to their cart, all the way to its arrival at their door.
Key takeaways from this episode:
• How product information is integrated into the seamless shopping experience customers see
• How automation streamlines the shipping process to optimize efficiency
• Why automation is transforming – not eliminating – jobs for factory workers
• What challenges are facing retailers today and how automation technology can help
Ron Kyslinger [00:00:03] We try to carry every SKU known to man and you try and sit it inside of one warehouse. That warehouse ends up being so big or would be so big, you’d be able to see it from the International Space Station.
Jim Carlisle [00:00:16] That’s Ron Kyslinger, president and chief executive officer of Kyslinger Consulting International and an executive partner for us at THL. And I’m Jim Carlisle, and this is Automation in Action, where we pull back the curtain on automation technology and lead you on a journey inside. I’m here with my colleague Syndie Kim, who spoke to Ron about the role in automation in the journey of our purchase from a click on a laptop or a smartphone all the way to when the package arrives on our doorstep. Before we hear from Ron, you’ve had the opportunity to speak with someone who specializes in the front end of things. Is that right, Syndie?
Syndie Kim [00:00:49] That’s right, Jim. I had a fascinating conversation with Marc Austin, chief solutions officer at Fortna and a logistics and supply chain expert. I asked Marc and Ron to walk us through the process of purchasing something online from start to finish and the automation technologies our purchases interact with along the way. As a hypothetical, I chose the barbecue grill that I bought earlier this summer online.
Jim Carlisle [00:01:12] Awesome. You know, I’m from Kansas City. I love barbecue. Syndie, I hope you use that grill a ton over the summer.
Syndie Kim [00:01:18] I did, Jim.
Jim Carlisle [00:01:20] Well, I find it fascinating to think not only about barbecue, but also about the way so much data from product specifications to stock levels to customer reviews all come together to give us a seamless experience when we shop online. We’ve come a long way from the late nineties. THL is an investor in a full range of companies that power the e-commerce lifecycle, including software companies that specialize in product information management, logistics, asset management and user generated content curation, but also robotics companies that power everything from warehouse automation to fulfillment. Fortna is one of those companies, and no one is a better resource than Marc Austin. Let’s hear from Marc.
Syndie Kim [00:02:08] Great to have you today, Marc. Maybe to kick off, what are some of the various software systems that power an e-commerce website? Say Target.com or Walmart.com?
Marc Austin [00:02:20] Thanks, Syndie. They’ll start with the web commerce platforms that are really the front end that you would link to. And that’s the very pleasant user interface that we’re used to seeing with the behind the scenes. They have to connect a lot of different functions and platforms that really grabbed the information and provide it back so that to provide that customer experience. So an order management system, we call it OMS, that checks inventory. It’s going to maybe look into an ERP system that’s going to provide the system with the specifications so you can do a comparison. And then there’s the other of the other types of things that they’ll typically hook in to several outside review sites so that you can go and see information that other customers have fed back and provided their opinion. So it’s really complex as you think about how the interface works together to provide us all that information does a seamless but behind the curve, really there’s a lot of things at play in all of those systems accessing information to give us one upfront interface that that lets us make a decision.
Syndie Kim [00:03:17] And if I imagine a large retailer like Walmart.com or a large brand with a lot of SKUs, I’m imagining that the information of the SKUs that are available as well as the specs of the different products are fluid over time. Could you give us a sense of how those product information gets brought in to the Web Commerce Platform?
Marc Austin [00:03:41] Anyone that’s on an ECOM platform has a really wide range of SKUs available to your point, where a retail retail store is going to have a very focused limited number of SKUs will typically keep a much wider platform. So getting all of that information entered in, getting photography done, getting the specifications that are that all ends up being across several different platforms and the order management systems and the different ERPs and then that feeds back up to that Web Commerce Platform so that you’re able to see all of it in one place.
Syndie Kim [00:04:11] Marc, where does the order management and logistics software come into the picture?
Marc Austin [00:04:17] So it’s writing between the web services side and the ERP. Looking at, hey, we’ve we’ve received these orders, we’ve been paid for these orders, now we need to take a look at where we have the items. The OMS is really looking at the what and the where along with the ERP. So, hey, this is what the clients purchased or what our customers purchased, where do we have it and then how are we going to go and ship that to meet a customer, to meet the client’s desires, to meet the customer service profile that they’ve chosen? It’ll take a look at if we have that inventory in multiple places where’s the best place for us to ship it from. It’ll link up to what we call post-purchase customer experience platforms, which is really dives into what, where and how to ship it, because that’s also a key. May be able to ship from the East Coast. It may need to ship it from the Midwest. It’s always going to look at what’s the best place to ship it from, but it also has to look at what your service level commitment, their service level commitment to you. It’ll work back and forth between the web service, web commerce platform and the and the ERP to try to make those decisions.
Syndie Kim [00:05:24] And then what happens inside the order management system to be able to direct that order to the right place?
Marc Austin [00:05:31] It takes a look at levels of inventory, service level requirement, operating hours, all the different business roles that it would take to get a product out based on when you need it. So if you’ve ordered it for I need it tomorrow and you’ve ordered that 1 p.m., it may only have certain choices if you say you need it three days later, it has a lot of options of where to ship it from and is probably going to look at the the least amount of touches and the most efficient process.
Syndie Kim [00:05:56] That’s fascinating, Marc. And similarly, how do inventory levels get exposed to a web commerce platform? I’m a user. Maybe I don’t see exactly how much is left in stock, but some of that information obviously needs to feed into what gets shown to the end customer. So I’m curious what the integrations might be there to enable that on an e-commerce website.
Marc Austin [00:06:21] They would be constantly linking back and forth with the OMS or ERP to understand what we have in stock, what’s already been allocated to previous orders or other customers and what do we have available and where. When you’re getting ready to place that order in the background, all of those things are being considered when you actually say I commit, when I click to buy and it will go ahead and say, “What’s the best place for us to take this from? Where do we have available?” Typically, they’ll know obviously upfront is that we do not have it available and it won’t be an option for you to buy it, but they may have it in multiple places and there’s no reason for you to know all of that. They’re just going to balance the business they to get you what you’re asking.
Syndie Kim [00:07:01] You’ve worked with a lot of customers that’s focus on e-commerce Marc, over the course of your career. What are some of the major complications or bottlenecks that these customers face around building an e-commerce infrastructure that works seamlessly from the front end to the back end?
Marc Austin [00:07:19] You want to give as much info to a client as possible and on the spot, right, to help influence a buyer. But that’s a large stake. Pulling together shipping info cost service stock information. Do we have it? They even ship it like we talked about earlier with inventory levels, product characteristics, all those details come together. What we want to buy is complex. It’s tough enough to launch these systems, but then there’s a constant work to maintain and keep it to date. So new products are coming in daily. Old ones go away on a regular basis. This has to be constantly managed, which is a really complex effort. When you look at it from a big picture, it’s not just an update on a website. You realize that to make an e-commerce system work really well, it isn’t just product experts deciding what will sell this week. It takes buyers planning, distribution and many others all to be in sync, to pull it off efficiently and to meet customer promises.
Syndie Kim [00:08:07] How do companies automate product information, and why is that essential on the back end?
Marc Austin [00:08:12] A lot of different things they need to do, but a good example of using automation update or maintain product information is you need to understand queue or volume metrics for an order. So traditionally the accuracy of the information has been troublesome with things come in. They’re not in the same box that we thought that you may get the same SKU from two different locations or or two different ports of origin, and it may just be different. So we have a large SKU base that problem gets magnified. So many companies utilize technology like a Cubiscan to quickly check and verify dimensions and the weight on on the fly. So we ensure accuracy not only in the distribution center, but so they can accurately apply the information when they combine that item that you’ve ordered with the other things you’ve ordered to get it to you efficiently.
Syndie Kim [00:08:55] Marc, you also help onboard a lot of customers who are starting up their e-commerce infrastructure in an automated way, maybe for the first time or the second time. What are some of the upfronts advice that you have for these customers, either around the physical infrastructure and the automation systems that go in or the software stacks that get integrated into the backend?
Marc Austin [00:09:21] You know, it really depends on the journey they’re taking. It’s such a big change for a retailer who has been used to shipping to stores larger orders. There’s a lot of items going multiple times a week to a store. There’s a limited number of SKUs. I’m going to pick 100 items or 200 items or 1000 items per store. That’s and I’m not really worried about a customer experience at the end. Your life as a distribution head or someone in supply chain changed dramatically when they say, okay, I’m going to now ship a final product to the customer, and instead of it being I’m averaging a thousand items per store, I’m averaging 3 to 3.5 units per person. And I’m going to now have to deal with all the overhead that that means like every order needs paperwork processing, maybe special handling.
Syndie Kim [00:10:10] Marc, has there been any examples of customers that have had challenges around the integration of the front end and the back end that has led to some some challenges of physical delivery of products?
Marc Austin [00:10:24] There certainly has. And without saying names, there’s there’s a lot of clients in the industry that have really struggled to make sure that they’re linking up proper. You know the order management systems and the front end web services are being linked up to proper order levels and inventory levels, but also being able to make sure that the information is provided to an automation system or distribution center in time to meet service levels. Quite honestly, I think one of the big changes in the industry of the past few years is many of these projects, those things, all of those pieces, the front end of the web order management, all of the things that manage transportation. And then you jump down into the automation side. A lot of them are being very managed in silos and you realize that all of them have to work together. So if you had one piece break in and you had a project or a big effort where someone was all not really focused on the overall effort, they just wanted to work on their thing. And that’s where you see a lot of fail because they hadn’t taken each other’s needs into account and how they’re going to interface. I think that’s been one of the big focuses is how do you create behind the scenes a system that when you and I go deal with it, we want to purchase a seamless to us.
Syndie Kim [00:11:34] That’s awesome, Marc. I learned a ton from our conversation today.
Marc Austin [00:11:37] Thank you very much. Enjoyed it.
Jim Carlisle [00:11:41] Marc, thank you so much. Syndie, that was a great conversation.
Syndie Kim [00:11:45] It really was, Jim. I had always thought about the front end and the back end in a separate way and never had connected that link of, hey, this broadening of demand and SKUs on the front end really impacts the way that goods need to flow through the entire system and ultimately get delivered to the end consumer.
Jim Carlisle [00:12:02] All right, let’s get that barbecue grill straight to your door. Next up, Ron Kyslinger. As I mentioned, Ron is an executive partner to the Technology and Business Solutions Vertical at THL. And he has spent his entire career in robotics and automation, starting in the automotive industry and in moving to third party logistics. He was recently interviewed in the New York Times about this whole space, and he’s held multiple positions supporting and assisting startup companies on back end automation and logistics.
Syndie Kim [00:12:31] Ron was the perfect person to help me understand what happened after I ordered my grill. The logistics and technology involved in getting it to my doorstep.
Jim Carlisle [00:12:40] Walmart, where Ron was the global officer and SVP of Fulfillment Engineering, is a great case in point. They have over 350 million SKUs in their online marketplace, but it’s impossible for a warehouse to hold even close to that much. So there are clearly a lot of logistics to consider for such a massive retailer. And to manage those logistics, it takes sophisticated automation technology.
Syndie Kim [00:13:06] Yes. And as you hear from my conversation with Ron, it’s quite a complex process for one product, like my grill, to make its way from a warehouse to your door. And when you add multiple products, the process expands.
Jim Carlisle [00:13:20] Let’s hear what he has to say.
Syndie Kim [00:13:27] Ron, thanks a lot for joining us today. We’d love to learn from you the physical steps and technologies that enable the fulfillment process that a common kind of e-commerce retailer site. I’ll give you an example. A couple of months ago, we were personally gearing up for the summer and we wanted to get a new barbecue grill. And I went to Walmart.com. I read a dozen reviews to make sure the grill was compact and portable enough for us to go camping with it while still enabling proper cooking. I finally ordered a grill after a few hours of research and it arrived at her home in about a week. Could you maybe help us understand how the item would get to a warehouse, get stored their pick packed and shipped and get delivered to our house ultimately?
Ron Kyslinger [00:14:15] Let me start by bifurcating the types of orders. The fact that you mentioned that it’s going to arrive in about a week, that tells me that it’s probably not a first party SKU. So what it means is that it’s coming probably direct ship from a vendor or even maybe a 3PL to your location. So Wal-Mart, one of the strategies that we set and we weren’t trying to be the everything store. We try to carry every SKU known to man and you try and set it inside of one warehouse. That warehouse would be so big that you’d be able to see it from the International Space Station. And so at Walmart, we want to carry the top one million fastest moving SKUs and other SKUs be on that will be direct should vendor or filled through 3PLs and other assistance. So let’s start with your grill specifically is probably coming from one of those 3PLs. So they’ve ordered products. It comes into one of their sites. They will pick it overbox it, probably to protect it. Sometimes they’re capable of shipping in their own packaging. But I can tell you, grills are one of those ones that are a little tricky. It’s like TVs. They tend to need a little bit of additional packaging to make the voyage. Middle mile and last mile delivery processes can be a little brutal on some of these larger, heavier products. The SKU would be ordered. And let’s assume that one of your SKU is in the top one million fastest moving, and that’s what they had inside their building. That means they would order from a vendor and they would be trying to order probably somewhere between 12, 14 as much as maybe 16 or 18 days of supply to their building. It would arrive if your barbecue’s portable, it can be auto recognized on the way in, meaning that we know what was on the manifest of the truck. We actually have three dimensional scanners and weight machines, so it’s coming in and it might say, “Hey, it’s a 20 inch by 24 inch by 18 inch box.” And so once that is actually all entered into the computer, we know what it is. It’s associated to that SKU. The SKU now can be sorted for inbound. What I say you get sorted, it can go in multiple different directions. If it’s needed immediately to the fill in order, it can go directly to a pick module to be put away. It normally will be put away in a random position, meaning that the computer system, the WMS, will say, okay, what spots are available? Any open spots available? A human can actually walk up with that item, scan an opening of the spot and say, okay, it’s on the second floor of the pick module. It’s getting stands scanned into this location. It’s in aisle four, section seven in spot B, and they put it in there and now the WMS knows that’s where it is. So whenever it needs to be retrieved, it can now say, take that exact same information, everything that was just stated. And it will now ask that person to make pick. Now, a lot of those things today are automated. So what I just explained is what’s happening in a manual pick module, somebody physically going, putting it on the shelf, scanning it to a shelf. An automatic storage and retrieval system will actually bring that in, store it away in a location, will automatically bring it out goods to person or in many cases, if the grill’s large enough and can ship in its own container, it will actually come down to the conveyance, get labeled, and no human hands have to touch it. That conveyor will take it for outbound, depending on what facilities you’re at. Sometimes that outbound is shopped. So what that means is that it will actually go on the outbound conveyor to where it is actually, again, three dimensional size and weight to confirm that everything that’s in the box is what’s supposed to be in the box. Let’s say your grill was supposed to weigh 10 pounds and your barbecue tongs weigh another 2 pounds. All of that together now should have been 12 pounds. Coming across the scale, it should weigh this amount. If it doesn’t, it will get kicked out to a human to check. And so they’ll make sure that they and accidentally put an extra item in or forget to put your barbecue tongs in. And then if it’s all packed correctly, it will determine what is the most economical way to transport it to you. And it will shop those rates and then determine which lane down the lane to send it to. So basically it’s now on a large sorter, a shoe sorter. The shoes are these little packs that sit on one side of the box and as the box is going along, those packs will automatically push the box down the lane that is associated with the truck and the company that you were supposed to have your box shipped through.
Syndie Kim [00:19:24] That’s great, Ron. And it sounds like a pretty complicated process for my grill to get to where it needs to be. What do you see as the main challenges retailers are facing today around the fulfillment process? Could you hit on some of the automation technologies that are up and coming to be able to address those challenges?
Ron Kyslinger [00:19:46] One of the biggest challenges right now is labor. So if I were to design a fulfillment center that’s capable of doing a million units a day, I might be able to get close to 500,000 units out of it on day shift. But trying to staff it for nights and weekends would be extremely difficult. I would probably be lucky if I got even a half of that or a third of that output on that shift. And the other problem is that I don’t care what software you use trying to determine where to put your next fulfillment center. Chances are it’s going to be close to a highway. So you have good access, which tends to put you right on top of everybody else. So if Amazon’s not already there, they will be and I still have a site there or Grainger has a site there. There’s a whole bunch of people fighting over the same demographic of person to work on that same type of process. They start getting into basically wage wars, trying to maintain the people if you’re in constant turnover mode, constant training mode. So automation tends to help solve those problems. When I speak publicly about this and people say, “Oh, you’re the person that eliminates everybody’s job because you’re bringing in all these robots and automation.” What I do is I make sustainable jobs. I don’t care how many humans you’re capable of finding if they have to run around picking up 40 pound bags of dog feed. The human body is not made to do that repetitively over and over. She’ll see the people’s neck start to hurt, their backs start to hurt their legs, their shoulders start to hurt. And eventually they’ll say, okay, I’ve had enough of this. And we create automation that does the dirty jobs, if you will, the things that are very repetitive in nature. We created, for instance, an auto packaging machine. So I talked about packaging your own barbecue grill and tongs together, and somebody’s having to build a box. If I’m doing that manually, some people do that between 32 to say 35 parcels an hour that they can build the boxes and putting the items in. And that’s on an average of 3.7 to 4 items per parcel. We co-developed the machine while I was there that can do a thousand parcels an hour without human hands having to touch it. Now I can actually afford to pay them more because I upskill them to run a machine. And now really, they’re a technician. They’re not really doing the human labor part. They’re just keeping the machine running. And they can probably keep multiple machines running. So now their productivity is much further increased, which means my variable costs actually come down so I can charge less to the end customer for that product. I can facilitate faster so I can be more efficient inside the building. It’s a much more enjoyable job, you know, working with really cool robots or working with really cool automation as opposed to lifting 40 pound bags of dog food all day and manually building pallets. And there is some really cool solutions out there now. A human can actually observe what is happening with the robot or get called in by the robot when the robot needs assistance. For instance, the robot might be capable of doing everything that I’ve programed it or other people have programed it to do. But all of a sudden it comes into a situation it can’t solve. And every robot, when it hits that situation, acts the same way that it stops in a neutral position and waits for help. And that help can come either through an HMI, which is a human machine interface. So now you can have a person with an Oculus headset with virtual reality. You can have a little Vive game controller and you can now manipulate the robot remotely.
Syndie Kim [00:23:52] Ron, robotics capabilities seem to be transforming e-commerce in ways that we haven’t imagined before as we walk through a couple of examples. Do you think this is mostly a hardware story? Is it also a software story?
Ron Kyslinger [00:24:07] Great question. It definitely is both. It is a hardware and software play. You can’t really do one without the other. You can, but you’re not going to be various effective. But where the developments really come in, I think for the AI and leveraging AI is in the intelligence, meaning we have systems that collect all kinds of data. And I joke about it, but we drown in an ocean of data where we have limited real time information for someone to make a decision. So what you’ll see is the AI is starting to learn from patterns and it’s starting to get smarter through execution systems.
Syndie Kim [00:24:55] Ron, one last question on on my end. You talked about WMS systems, WES systems and WCS systems which all work together inside a warehouse. What are the types of data or information that gets collected by each? And do you see a particular piece of software or data that you think is valuable for the AI to leverage in the next five to ten years?
Ron Kyslinger [00:25:25] Sure. Good question. So a lot of people are concerned that it’s collecting personal information. So it’s like, hey, I know what Syndie ordered and that can start getting tweeted across the Internet or whatever. Right? But that’s not how those systems work. And we make sure they don’t work that way. The WMS, as I mentioned, is going to be looking at what orders it needs to process through the building. Now, keep in mind that it’s through that building, right? So the WMS may not be aware of what’s going through another building next to it and another building. Right? So those systems need to be coordinated. And there’s a company by the name of FourKites, and they can identify that if you ordered three things that let’s say they were the barbecue grill, that you got patio furniture and it came with a, you know, a table and four chairs. If the chairs were in separate boxes, the table is in a separate box, it could actually track that order and see that one of the chairs got left behind. And you can actually flag the shipping company to say, hey, oh, by the way, you’re you’re missing one. Get it on the truck or we’re going to end up making two trips to Syndie’s house rather than one. Those systems are getting smarter and smarter, and more and more people are focusing technology in that space. And is that that focus technology starts getting smarter and smarter with AI, then it will become more efficient. So it will save more money because a lot of the things that we do today that are inefficient will get rectified.
Syndie Kim [00:27:04] That’s awesome. Thanks a lot, Ron. This is amazing.
Ron Kyslinger [00:27:07] You’re more than welcome.
Syndie Kim [00:27:11] Hey, Jim. So I know that THL has been an active investor within the e-commerce ecosystem, and you’ve served on a board of a number of companies across front end and back end, as well as warehouse automation companies. What is your take on the space?
Jim Carlisle [00:27:26] Well, it certainly evolved a lot, Syndie, over the past couple of decades. The amount of information that’s required for companies to be able to manage their product footprint, get good detailed information to consumers to help them in their decision. You know, they know that people like you were out doing a couple of hours of research online before they make the buy decision. And they also know that people like you may want that information in order to get the grill in to their home in time for the camping trip exactly that weekend, not weeks from now. And so there’s a lot of work that goes into that process Marc and Ron have given us great insights into that. From a customer perspective, the companies are trying to figure out how can we be better positioned than our competition to help the consumer in that journey? And there’s a few things that that those customers are thinking about. They’re thinking about how can I provide the best information online? They’re thinking about how can I get the package out the door as fast as I can? How can I improve the throughput of the facility that I’ve invested in? But they’re also thinking about some other things like how can I do this cost efficiently? How can I manage more inventory, more goods to be able to distribute them on a regional basis when labor is so, so tight? And again, that’s where automation comes into play. We’ve seen that the investment in this technology can produce a really strong return on investment because cost can come out of the system and the product itself, the service that they’re trying to provide can do a better job than what existed before that investment in technology. So not surprisingly, when really hard problems have to be solved, technology can be a solution, whether it’s in the digital world, which is the front end or the physical world, which is the back end.
Syndie Kim [00:29:15] Jim, that was very helpful. I think one thing that we all realize is during the pandemic there was this huge uplift of demand in e-commerce, and that has come down to more normalized levels in the past, you know, 12 to 18 months. How has that impacted customer decisions from your standpoint?
Jim Carlisle [00:29:34] I think what’s happened during the pandemic is the percentage of total retail sales that were conducted on an e-commerce platform took a step function up and that resulted in an absolute ton of growth and a lot of investment in technology from front end to back end to be able to fulfill that consumer demand. It’s not that the amount of e-commerce as a percentage of total retail sales has fallen, it’s just that that rate of growth wasn’t sustainable. And as a result of that, companies are trying to figure out, okay, how do I manage my physical footprint and how do I manage my digital footprint in order to accomplish the dual headed objectives, if you will, of driving revenue and saving money? And what we think and as we see in our own portfolio companies, is that the problems that we’ve talked about on a global scale, really, whether it’s labor shortages or rising cost footprints, given inflation, all those problems need a technology solution. And the amount of competition for a consumer eyeball has never been higher. So the investment doesn’t go away. But in the current economy, I would say things have slowed down a bit. Companies are being really deliberate about where to prioritize their investment. And our expectation is, is that for many companies, the types of investments that Marc and Ron described are at the top of that list. They’re absolutely necessary to be able to fulfill their company’s mission. For others, it’s been pushed a little bit to the right. I don’t think that means it goes away. I think it means that it’s delayed. And so, you know, we think that this is a space that was right for continued investment for a very long period of time.
Syndie Kim [00:31:23] Jim, one of the things that Ron mentioned that was really insightful from my end was that the automation in the warehouse really helps provide sustainable jobs and helps to alleviate labor shortages and create value for the end consumer.
Jim Carlisle [00:31:39] Well, with that answer, Syndie. I hope I’ve gotten through a set of questions. It’s been a really good discussion, especially between you and Marc and Ron. So I’d like to thank all of you for a job well done, and I’d like to thank our audience as well for listening to Automation in Action. If you enjoy what you hear, please like, subscribe, do all the things you need to do down at the bottom. Thank you so much. Automation in Action is brought to you by THL. To learn more about THL cross-sector strategy to uncover opportunities in emerging technologies visit THL.com/automation.