In this Digital CxO Leadership Insights video, Mike Vizard interviews HyperTrack CEO Kashyap Deorah about what makes managing a modern supply chain so challenging.
Mike Vizard: Hello, and welcome to the latest edition of the Digital CxO Leadership Insights series. We’re here with Kashyap Deorah’s CEO for HyperTrack, and we’re talking about supply chains and that last pesky mile. Hey, welcome to the show.
Kashyap Deorah: Thanks for having me, Mike.
Mike Vizard: We’ve been managing supply chains forever and a day. And there’s always been a problem in that last delivery point, or the challenge of figuring out how to get something to exactly where it is. And there’s usually a lot more people involved than anybody realizes. But we’ve also seen a lot more pressure on the supply chain these days. So is there a lot more focus on trying to optimize that last mile? And what are the challenges we need to deal with?
Kashyap Deorah: Sure, you know, the world’s changed just very rapidly in accelerated fashion. So you know, the good old days where you have a whole supply chain, bringing stuff on to the store shelves, and then you have customers who nicely go into the store, park their vehicles, go pick the stuff themselves, checkout, you know, at these aisles and take stuff back home, right? And now, the way the world has shifted, is we press a button in the phone, the universe conspires, moves around, brings stuff to us; products and services come to us. And we want it now, you know? Not just like next day, same day – we want it – even on-demand is not enough. We want it instantly; there’s instant commerce. And the other big change that’s happened throughout this is the workforce has become more flexible, and gig work. So just like we in tech, or, you know, in any space, workers are choosing more flexibility and pay for what you do. And the field workforce is no exception to that. So what that does do, you know, what that basically means is you, the workers, would wake up in the morning, press a button in an app and say, “I’m available to work.” And then once they log out, and go for their beers, after that, they don’t know if they’re gonna work tomorrow or this week at all. And that’s the flexibility they seek. So what that does to the last mile, Mike, is it just makes things a lot more complicated. You can’t just plan something for the next day sort of FedEx style, you know, create your batches and sequences and routes and rosters and assignment and go; boom, the next day, you know, a bunch of trucks would leave and things would happen, I would know at the end of the day, what did not get delivered. Now, everything’s real time and needs to be orchestrated. And the technology for that just has to be very different than what it was before. And a lot of people in the world are building out that technology for the last mile is our estimate is that there’s about a million developers in the world who are building last mile logistics tech, who are trying to solve this problem within their industry within their region, you know, within their context.
Mike Vizard: Am I extending my existing supply chains or am I replacing things with something that operates more in a real time fashion?
Kashyap Deorah: It’s a combination of both. Listen, I mean, when ecommerce started it, you know, sort of Amazon commerce that was a separate silo where there’s a different way that customers would order; there’s a different way that inventory would get into a warehouse, there’s a different way that it would get from warehouse to your door. And, you know, the store supply chain was just completely different from e-commerce. And then we saw over the last 10 years, the world’s converging where, you know, a store also becomes a warehouse, a warehouse also becomes a store and, you know, there’s micro fulfillment centers to enable instant commerce. So that’s where your warehouse is kind of becoming a store. So the short answer is it’s a combination of both.
Mike Vizard: Are the challenges different for a B2B company than for a B2C company? Are they both having the same kind of issues in terms of that last mile?
Kashyap Deorah: Oh, very interesting question. So, you know, let’s take consumable goods for example, the Seven Eleven’s style General Store could be a mom and pop shop, could be a chain. So, the way the world is shifting, there is the store that would need to stock a certain inventory. And it would count the number of inventory days they have, the inventory turns that they have, and obviously they don’t want to sit on inventory; but usually they will order once a week on some some items or once a month. Now the stores expectation has also become, “I’m just going to order for the day. And I’ll order the previous day, and stuff comes to me the next morning. I put it on the shelf. So I only have to forecast for a day. And I’m not worried about having that inventory on my shelves.” What that does to the supply chain for manufacturers and wholesalers is, think about it, the stores, just like consumers, are becoming more on-demand. Only order the little things that you want right now; don’t think ahead, low attention span, you know – just right here right now and, you know, make it happen for me. So what I’m saying is just as the consumer sitting at home is expecting things here and now, if it comes to my head that this is what I want, I just want it fulfilled right here right now. That psyche is passing along throughout the supply chain in the back as well. Where the store says I’ll just order right here right now, the wholesaler says I’ll just order right here right now. And you know, as you keep going back, that’s sort of perpetuating in the supply chain, if that makes sense.
Mike Vizard: Is it also impacting the customer experience? It’s not uncommon, for example, to go to a store that belongs to a chain, and they’ll tell you that they don’t have that particular item in stock, but they’ll tell you that the one across town does. But then they’ll call over there and they’ll say it’s there. But by the time you drive over there, it’s not there; it was never there in the first place. So how do we synchronize all this effort when we have so many tears in the system? And it’s not clear data can synchronize that anyway?
Kashyap Deorah: Oh, yes, absolutely. So you know, if you remember about 10 years ago, if you go to a store, and they don’t have it, let’s say 20 years ago, if you go to a store, they don’t have it; like well, we don’t have it right? But that other store might, and they can actually look up inventory. And then you go back to your car and you drive over there five years ago. They would say, well, the other store has it, do you want me to get it for you? And they would sort of source it and get it to their store. So somehow the supply chain that’s delivering to all the stores, was able to tag that inventory and say, “Okay, now it belongs to this store and needs to be moved from there to here.” And then you’d sort of go back to the store, you know, and pick it up. Now what started happening is, if they don’t have it, they’ll say, you just want me to send it home? It’s in that other store. And if you give me the address, I’ll just get it shipped to you. So what’s happening in this evolution more and more is systems, which were earlier siloed, are starting to integrate with each other and, like you said, synchronize with each other. And if you think about it, inventory needs to be synchronized, the customer data needs to be synchronized, the transportation mechanisms need to be synchronized. And the more of these systems start talking to each other, the more it’s you know, you’re going from a distribution centric parking supply chain to a fulfillment centric supply chain where the customer is at the center. And, you know, all the systems are conspiring to make it work for the customer.
Mike Vizard: And not everybody works for the companies involved, right? I mean, I see a lot of times there’s an Amazon truck, for example. And if you look closely, it’s operated by some other company, and they have their own systems, and then everything’s gonna get integrated back and forth accordingly. And we see every other type of store now trying to do something similar, where it’s same day delivery service, but the people doing the delivery are not usually in the same systems as the folks who sell the thing. So how do I create this kind of federated supply chain?
Kashyap Deorah: You look exactly in the supply chain, you know, as a thought experiment, if we have something in our hands at home that we ordered for delivery, how many different companies did it touch? How many different people did it touch before it got to you? And it’s a fairly large number. You know, depending on the item, it’s a whole bunch of companies and people that that have to hand it off before it gets to you. And, you know, it’s the fragmentation of logistics and supply chain that is just fascinating, you know; just blows my mind how we think that it’s a win or take all kinds of thing. Seeing a few companies on X percent of the market share, etc. But, you know, on the back end of things on the supply chain of things, it’s it’s super fragmented. Everyone uses different technologies. And you know, every possible combination of who’s using what tech and who’s using which logistics service, you know; you’ll see that in action, whether it’s food or groceries or any other supply chain. And I think the way the systems are talking to each other is API’s and sort of that’s the world now, you know – that Hypertech is in what I’ve realized is, you know, each one of these in a software architecture way think of them as sort of microservices doing what they do best, and all talking to each other in context of the transactions that they fulfill the best. So it’s all web services API’s, where each of these services just looks at themselves as a micro service. So from your Instacart, to the world, to the the local logistics service provider to a small store, to a transportation company; that’s that’s enabling all this, all using similar API’s for their systems to talk to each other.
Mike Vizard: So that makes it possible for somebody like you to then pull that data and aggregate it in a way that will allow you, I assume, to apply some analytics – is that right?
Kashyap Deorah: That’s right. So the role of HyperTrack in this whole thing is actually location and mapping Infrastructure as a Service. So as things get more on demand and gig fulfillment requires a lot more live location, and real time orchestration, using location and mapping. And we bought a micro service for that, we essentially enable that, and that ends up being used by whoever is building the logistics stack. Whoever’s building the tech for planning and assigning and tracking orders in the last mile, you know, once the order is accepted, to the point, you have it in your hand, that entire lifecycle; whoever’s building that technology, uses HyperTrack, and all of these systems are, you know, sitting on data, which then has a natural need for an analytics layer. And now a learning layer; how do you not only get the goldmine out of that data, and figure out meaning from that information, but also use that in a way to make your next order better, or your overall operations better? Right? How you look back at that has become a very important part of where the future is.
Mike Vizard: Is there an opportunity to throw AI and machine learning deep learning algorithms and all this to optimize this stuff? Or is it just too dynamic right now? And it’d be too hard to kind of say for sure that this is what’s going to happen without some human intervention?
Kashyap Deorah: Absolutely, actually because it’s dynamic. And because it’s complex, the role of AI actually becomes somewhat more significant, because, you know, pretty much a large proportion of the problems you’re solving are what I would call stochastic in nature, they’re probabilistic. They’re not deterministic, right? You’re, you’re sort of going by thresholds and probabilities and going, okay, is this possible that it’s this way, let me go do that, right. So when an Uber is finding a driver, who’s the right fit for you, there’s a lot of that happening, once the driver is dispatched. From the time the driver gets from there to you, there’s a lot of assumption around what routes the person will take and how long it will be before they get there and so on, right? And then when you get from point A to point B, how much time will it take, and that matters too, of course, that transaction as well as a lot of things beyond the transaction. So, you know, let me give you an example in the world of HyperTrack. So, for every fulfillment, the better you know, the precise location of fulfillment, where is the entry point where is the parking lot, where is the door at which the pickup will happen or the drop off will happen? To knowing the precise location helps you out better – helps you navigate the driver better, and then knowing the service time for that location, a building takes much longer to check into and service than a home location. A certain store on High Street will take longer than a certain store in the suburb, right? So how much service time will it be once you show up? And finally, the routes that you take to get there. So if you break every transaction into the precise location of where it’s happening, the service time, once you get to the location, and the route you will take to get to that location. If you model these three core sort of items and keep learning and getting better for every address and every customer, you know what those things are, and keep improving on that, that can be reinforced back to better route planning better. It’s just better overall efficiency. Did that did that sort of answer your question?
Mike Vizard: Yeah. How hard is it to set this all up? I mean, do I need somebody with an MBA? Or can mere mortals attempt this?
Kashyap Deorah: It’s, look, I mean, like anything in the knowledge economy, it goes with clarity of thought. Who, especially who, solved a similar class of problem before? They say, “Oh, this is easy stuff.” And at the same time, those who might be new to all this, and still understanding the context say, “Wait, why are we doing this?” It can, it can take a really long time. But you know, it’s the tools, and the skills out there have made it quite real and quite possible to do this very quickly.
Mike Vizard: So what’s your best advice to organizations that are kind of struggling with their supply chains? How do they get from where they are to up-to-date in this brave new world? And, yeah, how long might that take?
Kashyap Deorah: Sure. So look, I would say, the first step is to draw out a blueprint of your logistics stack. What you would find in many of the enterprises, many of the mature businesses right now is, you know, logistics is, is one of the oldest industries. And it’s one of the first things you have to put in place to fulfill orders to customers. And so there’s a lot of legacy built into this technology. And with every new sort of frontier that comes in for fulfillment, you know, people keep sort of applying an incremental version of that legacy technology. And then also, you know, traditionally, if you think about it, the front-end of how I’m gonna get more customers to market to more customers; sell more to them, is where technology investment has been much higher than how I will fulfill to those customers, especially the last mile. So you know, if you take marketing tech, as a proportion of all of e-commerce, it’s a much higher proportion than if you take logistics tech as a proportion of the entire logistics market. So an 11 trillion logistics market is serviced by $25 billion of logistics tech, which is like a very small fraction, like point one 2.2%, of all of logistics is the logistics tech. And that is changing very rapidly. It’s going from 25 billion to 200 billion in within this decade. Right? That’s, that’s the forecast. And what that means is, you know, a lot of people are building that right now, in the next few years; there’ll be more logistics tech that will get built that has been built in the last few decades. So one of the first things that I would say, leaders need to do whether it’s the CTO, or product executive, or its supply chain, or operations is the need to draw out the blueprint of, you know, for the fulfillment I’m doing right now. What does that technology stack look like? Who does what part of it? What part of this am I going to build? What part of this am I going to buy? For the things that I will build, be very clear about. What’s the IP; that’s core? For me, that’s truly the differentiator for me. And what is stuff that I can just buy through API’s, right? So even if I’m building stuff, and building with something, you’re not really building software from scratch these days; you’re building using API’s, cloud technologies, open source software, etc. So the better you know, what part of the build is core differentiator to you, the better you will also know what parts are the ones you need to get through API’s and cloud technology. So drawing out that blueprint, I think, is the first order of business and then you know, what that also helps do is give a lot of time to make this happen. You need many different stakeholders in the organization to come together; different kinds of engineering skills set different, you know – product design, engineering, leadership, and, of course, the operations people, and that involves the drivers, the ops managers, the customer support staff and so on. And finance; people are obviously keeping an eye on how much all this is costing, how much improvement and ROI this can make. So having that blueprint also helps sort of brings the team along to say, you know, “Hey, this is the vision and collaborate on it.” And then, of course, you know, going in and building out that blueprint.
Mike Vizard: Alright folks, you heard it here. First, the most important thing is to have a blueprint to prevent yourself from getting lost in the first place. Hey, Kashyap, thanks for being on the show.
Kashyap Deorah: Thank you so much. All right, you can find this video podcast on Digital CxO. We invite you to check out all of them. And once again, thank you all for spending some time with us.