AI-Powered Delivery Logistics: Zero100 x Deliveroo Case Study

Mike Silverman: The Zero100 Community is hungry for a case study on AI and supply chain, and I’m always hungry for a burger, so I met the team at Deliveroo for a big serving of both.

Deliveroo is a leading global food delivery platform based outta London, and I got to speak with operational leaders and delve into logistics technology at their global headquarters, as well as go behind the scenes at a Deliveroo HOP site where, among other things, I was introduced to Aunt Bessie’s Roly Polys.

But beyond the best of British snacks, the team at Deliveroo shared three key insights that we believe will resonate with our operations and supply chain audience. First, how do you utilize machine learning to optimize logistics? Second, how can you experiment with new business ventures to increase data visibility? And third, how do you build a culture of experimentation and data-driven problem-solving? While Deliveroo’s edge is in logistics, there are lessons here for supply chain leaders and operations leaders across functions and industries.

Zero100: This is Radical Reinvention, a show by Zero100 about reimagining the world’s supply chains. If you’ve been listening to our show, you know why we started Zero100: to push the world towards Zero Percent Carbon, 100% Digitized supply chains. This means a world where supply chains power business efficiency, growth, and resilience, and where people’s needs are met, but the planet is still preserved. To do that, we are working alongside the world’s most innovative supply chain leaders, technologists, scientists, and academics to modernize supply chains through digitization. And we are inviting all of you along for the ride. Join us as we work to create more responsive, resilient, and responsible supply chains, one radical reinvention at a time.

Mike Silverman: Let’s hear more about how Deliveroo operates from Devesh, who leads product and technology for the company and also sits on the Zero100 board.

Devesh Mishra: My name is Devesh Mishra, and I’m the Chief Product and Technology Officer at Deliveroo. I have the privilege of running our global technology teams, which comprises of engineers; scientists, right, from data to machine learning scientists; designers; business development folks; and product managers.

I joined Deliveroo in September of 2021, and in my role, I’m responsible for driving innovation and delivering exceptional customer experience through cutting-edge technology solutions. I report to the CEO of Deliveroo, and I sit in their executive team.

Mike Silverman: Devesh’s boss, CEO Will Shu, actually did stop by to visit us while we were filming on the main floor at Deliveroo headquarters, which made me a little nervous.But let’s have Devesh explain what Deliveroo is for those of our listeners who are outside their delivery footprint.

Devesh Mishra: Deliveroo is an online food delivery company which started in 2013. Ten years on, we are in 10 different markets, and we have more than 179,000 best-loved restaurants and grocery partners, and we have a network of more than 150,000 Riders globally.

Our mission is to build the best, definitive online food delivery company. We aim to be the platform where people turn on to when they think about food. Every day, we complete hundreds of thousands of orders globally to bring people the food they love on demand.

Mike Silverman: Devesh explained how delivery operates a three-sided marketplace that brings together partners (the restaurants and groceries who prepare the food), Riders (who deliver the food), and customers (who order and eat the food).

The size and complexity of the food delivery marketplace mean technology is critical when it comes to operating efficiently. The vast amounts of data collected generate a prime application for machine learning to optimize elements across the value chain.

Devesh Mishra: So let’s start with our consumers first. Our aim is to make sure our consumers are able to find and discover the food that appeals to them. And they’re able to place and receive orders seamlessly. So that’s one. The second thing is we work with our restaurant and grocery partners to make sure that we are able to provide fantastic customer experience for our consumers. And also in addition to that, we work with our merchants to help them get their businesses online and help them to grow their business, and drive more profitability for their businesses. And lastly, we have the Rider Network. For the Riders, we develop tools and systems so that they can deliver the order faster, with ease, and with safety.

Mike Silverman: Let’s dive into estimates and economics – it’s not boring, I swear – which really helps explain what goes on behind the app for every Deliveroo order. Deliveroo’s biggest challenge and opportunity is how they manage the three-sided marketplace. Especially interesting are the components that apply to their self-employed Riders. The Riders have the freedom to work when they want and the freedom to accept or reject any order. Based on insights provided by Deliveroo models and algorithms, Riders get a bunch of information that helps them choose. To explain this further, let’s meet Charlie.

Charlie Wren: Hi, I’m Charlie Wren. I’m the VP of Delivery Product and Operations here at Deliveroo.

Mike Silverman: What are some of the main operational challenges to delivery logistics?

Charlie Wren: So the main operational challenge, the thing that’s kind of kept me interested over all these years at the company, is the big one, which is that we have a massive network of Riders, partners (that’s restaurants and grocers and customers), and we are trying to get produce, hot food, or grocery from the partner to the customer. The customer’s impatient. We wanna do it as quickly as possible, and all of the actors in the network are independent. We are not directly involved in the transaction at any step of the way, we are just creating the mechanisms and the systems for that to happen. So we have to set things up so that the Riders want to accept to complete the order so that the restaurant gets things ready on time for when the Rider arrives, and so that it all synchronizes so that the customer gets their food quickly and in good condition. Customers are hungry, normally. They want that food quickly, and the hotter we can get it to the customer, the better. So that’s the challenge. I mean, it’s a fascinating one and challenging one.

Mike Silverman: Charlie went on to explain the complex data analytics and behavioral economics at play.

Charlie Wren: As I say, I’ve been here a number of years working on the, kind of, the logistics or the delivery operations challenge at Deliveroo, and on the face of it, it’s this really simple thing. How do you get someone their dinner? But behind that is this massive kind of web of models, of systems. The whole ecosystem is driven by the kind of different behaviors and behavioral economics for all parties involved. When the whole thing works: the restaurants preparing or the grocers preparing that order in time for the Rider to arrive, the Rider loves that. The restaurant sees their food leaving in good condition, it gets the customer, they love it. But to make that thing happen, we need to make sure that all the right incentives are in place and that everyone wants to do those things at the right times.

Mike Silverman: There are numerous calculations, estimates, and models being run in the background.

First, Deliveroo needs to ensure that there are Riders on the road to accept orders, meaning that they have to estimate the fleet needed at a certain time. If there aren’t enough Riders, the system automatically adds extra fees as an incentive. Second is about how long the order’s gonna take to be prepared by the partner restaurant. The time estimate should be as accurate as possible to avoid the Rider arriving too early (and having to wait there) or arriving too late, which causes the food to get cold and can slow down delivery to the customer. Lastly, the system has a travel time model to estimate how long the Rider is going to take to get from point A to point B. The model considers things like the time of day and traffic conditions, and instantly presents the travel time to the customer via the app. The whole system is a number of rapid transactions based on tech models and behavioral economics. All of it keeps the engine running.

I’d like to ask you, what’s a key takeaway you wanna share with our audience? With supply chain leaders who are learning from Deliveroo?

Charlie Wren: Key takeaway? Well, I’ve been to a few different events where I mingle with supply chain leaders, and I always think it’s sort of fascinating, the similarities that we have. Ultimately, we’re trying to get something from A to B, but also the differences that are there.

And so, I guess there are two main things that I always think are really interesting about our model. So the first thing is the point that I’ve already made. We’re talking about creating a system within which completely independent actors are able to synchronize to achieve what we want to achieve, get the thing from A to B in good condition, and that timeframe is really, really short.

But one of the really interesting things I hear is when you talk about sort of traditional supply chain or delivery supply chains, this problem with the empty box or the empty van. And one of the really great things about our model is that we kind of have to create the incentives and have the structures in place for that problem to really not to be a very big problem. Empty box in our model means that the person with the empty box i.e., the Rider isn’t earning. You know, apart from the point at which they’re going on their way to the restaurant, after they’ve delivered at a customer, any time with empty box is time that they’re not earning.

And so what we see is that as our network becomes denser, as we see more restaurants, more Riders, more customers in a given location, the middle of London where we are today is a great example of this, you see that the Rider will go to restaurant A, they’ll deliver to customer A. And then the restaurant that they’re now going to, restaurant B, is very close to customer A. So that pickup or that leg to their next pickup is really, really short. And so, you see this, I mean the smattering or the distribution of Riders, restaurants, and customers across somewhere like London, but it’s also the same in Hong Kong, in Paris, in Singapore, some of these really densely populated places. They’re kind of everywhere. And so, any given journey from A to B for the Rider, when we offer them that order, is very, very short. And that’s one of the big advantages of this system, but also one of the big things that, you know, possibly does have read across into our other examples of supply chain.

Mike Silverman: We talked about the delivery business and Deliveroo’s three-sided marketplace, as well as some of the optimizations that happen, especially for their Riders.

Now let’s jump into their dispatch system. It’s powered by machine learning, and it’s called Frank. I spoke to Mahana, who told me all about Frank and the science behind decision-making at Deliveroo.

Mahana Mansfield: I’m Mahana Mansfield. I’m VP of Science at Deliveroo. So I look after our data and science organization, and the mission of that organization is to enable the highest quality human and machine decision-making.

Within the data and science organization, there are five disciplines, or five skill sets that we have. The first one is machine learning engineering. The responsibility of our machine learning engineers is to build our production decision-making software in our consumer app and our Rider app, for example. We have data scientists who are responsible for doing the, I guess, the complex statistical analysis, which might include experimentation, econometric analysis, causal inference, to help us make the really, really hard decisions. We have data analysts who are responsible for building the tools and helping humans make decisions day-to-day all throughout the organization. We have analytics engineers who are responsible for making sure that we have the right, robust, reliable data sets to use. And then, we have our data management team, who are responsible for making sure that we have the right kind of data integrity and the right governance around our data sets.

Mike Silverman: Can you talk a little bit about how Deliveroo uses data to operate and what data is really critical for the business?

Mahana Mansfield: Yeah. Great question. So how we use data, I guess, is to advance on the spectrum from decisions made via intuition and guesses through to data-informed decision-making to scaled decision-making where we can allow our humans to say, this is the goal that we wanna achieve. And then, you know, these are the kind of the data points that feed into it, and this is how we can scale these decisions. A great example of that is our dispatcher, which we call Frank. So initially, it was people and spreadsheets, and then it was informed by data, and then it’s automated via these algorithms and these machine learning models.

And the next thing that we’re looking to do is make that decision-making faster. So if you can imagine, our dispatcher has to consider so many combinations of Riders and audits, and if we can find some maths tricks to speed up that consideration of the components and potentially eliminate some potential combinations early because we know that they’re just not feasible, then we can make these decisions even faster and get even better outcomes overall.

Mike Silverman: Can you talk about how Deliveroo uses artificial intelligence and machine learning today?

Mahana Mansfield: So obviously, artificial intelligence, machine learning, lots of terminology that’s kind of used all the time at the moment. I have a bit of a thing about artificial intelligence in that I think a lot of what is talked about artificial intelligence is probably not intelligence.

Artificial intelligence, to me, it means trying to replicate human intelligence in a really kind of smart way, and a lot of what needs to be done, in fact, what is done is simpler than that. It’s you have a goal and you wanna try and achieve that goal. So, you know, for example, for us and dispatcher Frank, our goal is to match the right Rider, the right orders, and then we have the kind of the data that feeds into that. So yeah, we use more of the machine learning side of things.

Mike Silverman: Mahana shared how Frank is enabled by machine learning and deliberately optimized to empower better decision-making. What’s really cool is that at Deliveroo, this is applied cross-functionally. So while everyone reports up to Mahana, there are a number of her team that support folks outside of her organization.

Mahana Mansfield: We work in a fully embedded way, so we work in cross-functional teams. That might be with engineers and product managers, but it might be with marketing specialists or operations people, or finance people. And we are really embedded across the whole company in the day-to-day. The people on my team are more so working with their cross-functional partners than they are necessarily working with other people from my organization.

Mike Silverman: So that enables kind of everyone to get the benefits of machine learning and of better decision-making.

Mahana Mansfield: Exactly.

Mike Silverman: Our audience is a lot of supply chain leaders, supply chain practitioners. A lot of ’em are especially interested in AI machine learning capabilities. Do you have like a takeaway message for them or any recommendations?

Mahana Mansfield: Sure. So I’d say don’t be scared of it and don’t feel like it’s intimidating and it’s hard to understand. So if someone’s explaining something to you in the AI machine learning space and you don’t understand it, then ask them to explain it again, and ask them to explain it again. And, you know, for me, thinking about my team, one of the key skill sets is to be able to explain something in a way anyone can understand. And it’s not actually that complicated. It can be explained really well in plain English. So don’t be intimidated. It’s so useful. But make sure you understand what’s happening.

Mike Silverman: After visiting headquarters, we got to stop by a Deliveroo HOP location in central London. Deliveroo HOP is a dark store that fulfills customer grocery orders received either via the app or an in-store kiosk. The orders are delivered by Deliveroo Riders or can be picked up by customers directly from the counter. What’s really cool is that Deliveroo HOP is increasing the number of insights the team receives from the physical location. That allows them to increase the speed at which orders can be fulfilled.

I talked to Suzy and got to go behind the scenes to more clearly understand the technology underpinning store operations.

Suzy McClintock: My name’s Suzy McClintock, and I’m VP for Grocery, HOP, and Additions at Deliveroo. So that means that I run our global grocery business, which is third-party store pick marketplace. I’m also responsible for our HOP locations, which is where we are today, and our additions cloud kitchen business.

Mike Silverman: Tell us about the space we’re in today. What is a HOP location?

Suzy McClintock: Sure. HOP is Deliveroo’s own dark store, which means that we service groceries out of this site. It’s our first consumer-facing location. It’s based in the city center, so on a high street, very close to Oxford Street, one of the major shopping locations in the UK. All of our other sites are located in sort of industrial locations, which are close to the city center but not close to where consumers actually shop on a day-to-day basis.

Mike Silverman: Can you talk to us a little bit about the life of an order? What does the customer see, and what happens back here behind the scenes?

Suzy McClintock: So the customer lands on the Deliveroo app, you know, thinking, I want groceries, or I need a cold drink, or whatever it is. They find the HOP site. So this site will have a hero image on the app, click on the hero image, and they come into the store. And from that point, it’s like browsing any grocery store. So we have a shop-by-aisle experience. The customer can think, okay, I want to order fruits and vegetables, cold drinks, et cetera. Add them to their basket, hit order, and then the order comes through to the site, at which point it’ll end up on one of our picking devices.

So then one of our Pickers will get that through on their device, and they’ll go around the store and pick the order, typically in less than two minutes, which I think for some of your listeners, viewers, et cetera, will seem incredibly fast compared to your normal grocery operation in a much bigger warehouse. But the fact is that this site is relatively small, and we carry a pretty limited range of fast-moving items, which means that we can pick the orders really quickly. From that point, once it’s picked, it’ll be taken to pack at dispatch, put it in the bags, Rider turns up, hand it over to the Rider, and off it goes to the customer.

Mike Silverman: What are the benefits to customers for using Deliveroo HOP?

Suzy McClintock: Absolutely. So I think it’s really interesting in that we operate within the Deliveroo grocery ecosystem, but there are three specific benefits for HOP.
So number one is speed. We deliver orders typically in under 20 minutes and that’s because we have a small range, we’re located close to customers, and we also watch our pick speeds so we can deliver really quickly.

The second thing is inventory accuracy. So it means that there are far fewer bad outcomes because we don’t know what items are on the shelves than there would be in a typical store pick grocery store. So in a store pick grocery store, it’s almost impossible to have 100% inventory accuracy. Whereas here, what we have is 100% availability and inventory accuracy at all times. So we know what’s on the shelves, and therefore consumers can’t buy something that we don’t have available.

And then I think the third thing that’s important to remember is value. So unlike some other dark stores, we partner with a brand. So here we partner with Morrisons, and it means that we offer the full range of Morrisons own-label to consumers, which are great value products. So it means that consumers can get the products that they would typically find in a normal supermarket, you know, crumpets, pitta breads, some surrounded by them. And that’s something pretty unique in the dark store ecosystem today.

Mike Silverman: Let’s talk about the Riders. What are the benefits to them? Is picking up here easier or different than their other Deliveroo orders?

Suzy McClintock: Yeah, so we think about the Rider experience quite a lot, and we’ve actually designed the sites with the Riders in mind. So it’s not a great example at this site because it’s obviously a consumer-facing proposition as well, but what you’ll see at the other sites is that, actually, the Rider pick-up dispatch area has been thought about. It’s nice, it’s warm, there’s water available, and what Riders really want is work, right? So they also know that when they come here, the order will be handed over to them really quickly when they turn up, and that for them is very attractive because there’s no waiting around.

Mike Silverman: Amazing. I’d love to move on to talk a little bit about the technology here. So can you tell us about what technology is being used here at this HOP location?

Suzy McClintock: What technology isn’t being used? So I think we’re really fortunate in that the business is relatively new. So we could kind of build it from the ground up with the current technology that’s available in mind. So everything from the warehouse management system that we use, the picking technology that we use, I mean, we use a forecasting model that forecasts at a SKU level, which some of, I think your people listening will understand, that’s actually relatively difficult, particularly when you have a new business. We obviously have the Deliveroo platform technology that we plug into, and we do that in a really specific way. Because we know how long we take to pick orders in this site, we actually plug in a different point to your typical restaurant so that we call the Rider before the order’s being picked because we know that it’s going to be picked in time.

Mike Silverman: Suzy, can you tell us about the data that you collect here at the HOP location?

Suzy McClintock: Sure, absolutely. So we collect a lot of data. All of your kind of typical demand metrics, obviously order volume, gross transaction value, et cetera. Then you move into the onsite metrics. So things like pick speed, inventory turns, forecasting accuracy, fill rates on RPOs, et cetera. And then, we also look at consumer demographics, so the types of consumers who are ordering, the time of day. And we use that consumer data to constantly iterate on our proposition. So really understanding what the local consumer wants and then looking to fulfill that through our range.

Mike Silverman: Awesome. Can you tell us a little more about what you’ve done with some of that consumer data here in this location?

Suzy McClintock: Yeah, so this location is really interesting actually because it’s a city center location. So it means that the demographic here is relatively different to the demographic in our other sites, which will typically deliver to people’s homes. And we had a look at where people’s primary addresses and secondary addresses were and found that typically people are ordering here when they’re in the office, even if they’re ordering in the evening, so maybe they’re ordering food for dinner, food for lunch, et cetera. And typically in our other sites, we didn’t have a big range of sandwiches, salads, et cetera. And now what we are looking at in this site is how do we get a big range of sandwiches? How do we get more food to go, more chilled drinks, et cetera, that will really speak to the consumer demand in this location.

Mike Silverman: Can you talk about an exciting project that you’re working on now?

Suzy McClintock: Yeah, so an exciting project that we’re working on at the moment is called Code Name Project Joey. So like the kangaroo, but it’s our top-up orders experience. So for people in America, they would understand that as DoubleDash. So it would be a UK first, and it means that you can top up your restaurant order with groceries from a HOP site or from one of our grocery partners. So, you know, imagine you order your pizzas in the evening, you’re like, oh, I really want some cold beer. You can add them to your order, there’s no additional delivery charge, and eventually, the same Rider will deliver both of them for you. Or you are getting lunch at the office, and you really want some fruit, and you know, some food for later on in the evening. So we’re really looking forward to it being a great experience for consumers.

Mike Silverman: At this point, a Deliveroo Picker interrupted our interview so he could reach some of the items that were next to Suzy. Suzy took the moment to talk about what was in the customer’s basket.

Suzy McClintock: What you can see here is a typical grocery basket. I think we’ve got some halloumi, some green tea, some grapes, some water. And I think what everyone assumes about dark stores is that it’s gonna be impulse purchases. So like cigarettes, alcohol, sweets. But what we see is it’s not. I mean, this person’s getting facial tissues, water, washing liquid, coffee, you know, that’s a grocery shop, that’s someone topping up their grocery shop.
These are really good brioche buns, by the way, just saying they are.

Mike Silverman: I mean, who doesn’t love a good brioche bun? Sadly, it’s not allowed on my summer diet.

Anyway, I wanna tell our listeners what is all the noise and what’s happening around us. What’s the activity in this HOP location?

Suzy McClintock: I guess you can hear the beeping in the background and the rustling. That’s someone picking, I can’t actually see what he’s picking off the shelves, but he’s picking at great speed. You’ll probably hear some rumbling of the trolleys. So we use, we tried actually a lot of different carts, and we do use different carts in different sites to fit the locations. So we tried like a double-stack cart, but we discovered that these little mini trolleys actually seem to be the most efficient. I guess that’s the noise of the fridges in the background. So this is actually one of the only sites where we have fridges. Now, most of the other sites we have walk-in chillers, and some of the sites we have walk-in chillers that you can range into fridges. So they get replenned from the back, which is pretty cool. But then the Picker can open a fridge at the front to take the product off the shelf, which is really good for FIFO. What else is there to talk about? These eggs, Burford Browns very, very popular. I had a real fight to put them on the menu because they’re my favorite and they are quite expensive, but they fly off the shelves, so that was a good decision.

So what you’ll see in the dispatch area is we’ve got big screens up on the wall and that means that managers and also Riders can see exactly what’s going on in the building, like what status an order is in. Is it being picked? Is it ready? Has it gone out the building? It also means that we in head office can see where every order is, and we have the same display on our laptops, et cetera.

Mike Silverman: It was amazing to see everything that was happening at the Deliveroo HOP store. We got to see the speed and efficiency of how things were laid out, how data enables everything to move so quickly, both within the store and for Deliveroo Riders who are picking up orders. With that, I said goodbye to Suzy… after I picked up a few snacks for my friends back at the office, of course, hoping that they would eat those brioche buns before I do.

Visiting the team at Deliveroo was amazing. Not only because we got to understand what they’re doing, but we really got to delve into how they’re doing it. How they do the work is really important. And three things stood out to me.

At the heart of Deliveroo’s DNA lies: one, a heavy focus on customers and customer behavior; two, a data-driven approach to decision-making; and three, a culture of experimentation.

Firstly, a large helping of customer obsession. Decisions are constantly being evaluated on their ability to improve the experience of customers across the three-sided marketplace. Taking a customer-centric approach myself, I asked Suzy about how Deliveroo is evaluating what customers are doing at this specific central London HOP location. Seeing that more customers are ordering from their offices, Deliveroo knows that they should be stocking sandwiches and ready-to-go items in addition to grocery store items.

For the Pickers, the store is laid out in a way similar to a grocery store that customers would visit because it makes more logical sense. It allows people to find the items faster and in a more intuitive way, as opposed to if the layout were optimized by a computer.

Deliveroo Riders are able to pick up faster from Deliveroo HOP locations as they get a really accurate estimate of when orders will be ready. And this applies down to the entire Deliveroo team. Mahana spent time in the We Are Deliveroo program working a customer care shift, and she noticed some critical information was being shared but on different screens, which made it harder to toggle between different things and to answer a customer’s question.

The Deliveroo team is always looking for elements of friction or delays that can be eliminated with questions as well as answers arising from the data they collect. Beyond standard data analysis, machine learning has the capability to improve and hasten decisions across the organization. Mahana told us to lean into machine learning and to find ways where we can make or allow us to make better decisions more quickly.

What I really loved is that Mahana’s machine learning teams are not siloed within data science, but they support functions like marketing and finance as well. They bring machine learning and data-driven decision-making power, at scale, across teams. This has massive implications for digital talent, especially for supply chain leaders, as Devesh explained.

Devesh Mishra: So the world is becoming more and more interconnected, more technical, and more automated. And the skills that worked in the past is not gonna work in the future, especially when lots of decisions are made through computers. And that’s where we need new skill sets for our employees on how the humans and machines can work together to drive the flywheel faster. How you can audit and inspect machines at scale.

Mike Silverman: Zero100’s forthcoming Rewired 2030 report showed that 48% of Chief Supply Chain Officers discussed top talent on a weekly basis. Yet, only 17% believe they’re ahead of their competition when it comes to recruiting for tomorrow. Further analysis has revealed a strong correlation between supply chain, digital maturity, and its leading position in talent. Many supply chain leaders are hungry, pun intended, to bring digital capabilities and IT within their own organizations. The shared service model at Deliveroo is a great example of bringing machine learning firepower to various teams, including supply chain and operations.

My final thought is about Deliveroo’s culture of experimentation. Devesh explained how they utilize technology to keep a pulse on the customer’s real-world experience, but always with the aim of experimenting and being able to fail faster and cheaper.

Devesh Mishra: When you are innovating and defining the future, there’s no playbook for you to apply, and in that case, you have to experiment a lot. You have to try out new things rapidly. And when you are experimenting a lot, by definition, you will have a lot of failures. And the way I think about that is it’s not about the failure itself, it’s about the learnings from those failures, and how can you make sure that you are able to fail faster and cheaper? And the way we do that is a combination of things. First is, we use a lot of our forecasting and machine learning models to try things virtually. That’s one. Then we experiment in real life to try those things faster. And in addition to that, we also do a lot of primary research. By primary research, what I mean is spending time with our end customers to really understand how they are experiencing our products. And by using the combination of these three things, we are able to try out things faster and cheaper.

Mike Silverman: All right. I’m getting hungry, so let’s wrap this up. There are three key takeaways for the Zero100 Community.

First, invest in machine learning to accelerate better decision-making. Mahana encouraged supply chain leaders not to be afraid of AI and machine learning technology but to apply it wherever it can help, especially across different silos. The main thing is to really discern where you want computers, not humans making decisions. And beyond that, where can it enhance or speed up human decision-making?

Second, increased data visibility across your supply chain. The reason Deliveroo HOP works so well and can respond with such speed to their customers is because they have real-time visibility into their inventory at each store. Supply chain leaders are already on the road to increase digital visibility across their own operations, and there’s important lessons here when it comes to partnering with other businesses or expanding your own ventures. Consider these opportunities that allow you to gain increased visibility, increased data, and therefore allow you to improve operational efficiency.

Third and finally, embed data-driven decision-making and experimentation into your team culture. The entire Deliveroo team I interviewed spoke in a common language, even across our individual interviews… a little freaky. The culture of data-driven decisions shapes how leaders explain their work, and a focus on experimentation inspires an appetite in their teams to constantly invent and reinvent.

It’s worth asking yourself, are these values you care about? Are they in place in your own organization or something you need to develop further? Deliveroo’s fast-moving hyper-local delivery business has applications for operations and logistics leaders across various industries. Their focus on advancing digital technology capabilities, coupled with a data-driven and experimental culture, are a powerful combination that can accelerate our community’s path to Zero100 supply chains.

I wanna say thank you to all of our guests today, the team at Deliveroo, and the team at Zero100. Since we’ve got a bit of time, let me share a clip of Suzy and I scanning through the British treats on the Deliveroo HOP kiosk, some of which needed some translation across the pond.

Suzy McClintock: It is really funny when you see stuff on here, and you’re like, I really need to talk to the Vendor Manager and ask why we decided to range Aunt Bessie’s Jam Roly Polys. But cool.

Mike Silverman: What is a Roly Poly?

Suzy McClintock: You don’t know what a Roly Poly is?

Mike Silverman: No.

Suzy McClintock: This is my favorite thing with Americans.

Mike Silverman: You guys do have the best snacks, I think.

Suzy McClintock: Crumpets. Crumpets are a thing that I feel Americans don’t really understand.

Mike Silverman: You know them from a nursery story, that’s like about it. Yeah.

Suzy McClintock: It’s like, ooh, a crumpet.

Mike Silverman: I just had a hot cross bun at our office.

Suzy McClintock: Hots bun to the other one.

Mike Silverman: I didn’t. I thought it was just a piano song, you know.

Suzy McClintock: No, I mean, hot cross buns are delicious, also.

Zero100: This episode of Radical Reinvention was produced by Brian Egan, Catherine Parry, Mike Silverman, Ursalaan Khan, Duda Rodrigues, and Nick Heinemann. Ko Takasugi-Czernowin composed our theme music. To find out more about Zero100 and to check out our content library, go to Zero100.com. If you’re interested in joining our community of contributors, send us a note at hello@zero100.com.