Deliveroo plc (ROO) Earnings Call Transcript & Summary

Okay. Good afternoon, everybody. We will get started. My name is David Hancock. I lead Investor Relations here at Deliveroo, and it's my great pleasure to welcome you to our first Capital Markets event here in our London headquarters. Thanks for joining us here in person, for those of you who are with us. I'm not sure what I should interpret about the expectations that no one dares to be on the front row, but let's see how things evolve. Thanks also for those joining us on the webcast. And in particular, thank you to those joining us live on the West Coast in the very early hours of the morning. I know there are a few of you. I'll start with a few quick housekeeping points. So firstly, on mobile phones. I'm not going to ask you to turn them off because you might want to use them shortly, but I would ask you to turn them on to silent. We also ask you please not to record or to take pictures during the presentations and during the breakout sessions afterwards. And then secondly, a final bit of admin for now. There are no scheduled fire alarm tests today. So if there is an alarm, then please do treat it as genuine. Moving quickly on to the agenda for the day. So we've chunked it into 3 sections. So we'll start with a welcome from Will, then we will cover restaurant, grocery, retail and the delivery experience, and we'll follow those presentations with the first 30-minute Q&A session on those topics. We'll then have a 10-minute comfort break, so water and bathrooms before being back in here for the second set of presentations, where we'll cover Plus, ads, the financial framework, and then we'll wrap that all up in the customer experience. And finally, a wrap-up, then we'll have the second 30-minute Q&A session at that point. And then finally, we will have, for those of you who are here in person, some breakout sessions where we'll go into a bit more detail on the product and bring to life what you have seen on the presentations during the afternoon. We'll also be serving some food and drinks during that session from some of our favorite local partners on the platform. So I hope that whets your appetite slightly for what's coming up, hopefully, not only for the best of Popeyes and Shake Shack and the like to come afterwards. So without any further ado, I'm going to hand over to Deliveroo Founder and CEO, Will Shu. Thank you.

Can you all hear me okay? Great. So we have Shake Shack coming? Okay. Thank you. All right. So good afternoon, everyone. It's a pleasure to welcome you all to our home. This is our HQ for Deliveroo in London. And I'm thankful you've taken the time to come in person. We haven't done this before, so really excited about today. And for those of you who are joining via video link, thank you very much as well. I'm Will Shu. I'm the Founder and CEO of Deliveroo. Three years ago, I stood in front of you guys virtually to share the story of what Deliveroo would be like over the next few years. And those years, frankly, have been crazy. We had COVID, we had the unwind of COVID. We had to adapt to a completely different macro, and we had to do this all in real time. It's been a challenging period, I think, for all businesses, but challenging markets are when winners emerge, and I'm extremely proud of the results that the team has delivered. We'll talk in more detail about what we accomplished in the last few years in silo section, but mostly, we're here to talk about the future and why we're confident in what we're presenting to you today. But first, let me get to the punchline. We are confirming 2 numbers: mid-teens growth in GTV in the medium term and 4%-plus adjusted EBITDA margin by '26. Now I know some of you might be tempted to leave now that we've shared the numbers, but please do stay because we're going to tell you the important part today how we accomplish this. And we're going to get into a bit of detail about our plan, but first, I want to set out why I'm excited about the opportunity ahead and why I deeply believe Deliveroo are uniquely placed to capitalize. So I've been at this 11.5 years, and I still come in to work excited every single day. Why is that? The first one is the size of the prize in front of us. The second is the fact that on-demand deliveries are notoriously difficult to execute well. And third, it's the amazing team I work with. So first, the TAM. You all know we operate in a giant total addressable market in all of our verticals, and it's still underpenetrated. I'm sure it's a bit obvious. I'm also self-aware enough that talk of giant TAMs in this interest rate environment isn't always appreciated, but it is true. The second reason, on-demand deliveries are notoriously difficult to execute, yet we are excited by that. Is that weird? Well, I think getting this right is the key to unlocking the TAM I just talked about. And that unlock comes through building customer trust, which is through price integrity and building a flawless delivery experience. What's on the screen there? That is a pizza that is delivered upside down. And the reason this picture has special resonance for me is I did the first-ever delivery and I delivered a pizza to my friend upside down. And so for me, it's a symbol of how things can go wrong and a reminder that we at Deliveroo and the industry are far from perfect. Carlo is going to talk about price integrity. Camilla is going to talk about the delivery experience later. At the end of the day, this is a complex business with a lot of moving parts with some parts seemingly out of our control or influence, but that's what we challenge each other on every single day. Is it truly out of our control? Can we influence a better outcome? That's our #1 question. And then the more parts of the process we can influence, the better the consumer outcome. What's the third reason? It's the people on the screen here. I'm excited because I have an incredible team who feel the same way I do. They care about every single customer and every single order, and they are motivated by getting the details right. I've worked with many talented people over the years, people that helped create this business with me. But I can tell you this, this team that I work with today, they're the best team I've ever worked with in terms of talent, experience and the way they work together. All right. So I've talked about midterm guidance. I talked about why I continue to be excited. But now I'm going to talk about something that is emotional and perhaps unexpected and certainly not a topic I spend any time talking to investors about. But I do want to go into it. And that topic is: What is Deliveroo all about? Who we are? Because who we are is why I'm confident we can deliver on what the team is going to lay out for you today. I'm going to take you back in time further than 3 years. In some cases, I'm going to take you back to the beginning. Because at Deliveroo, who we are is the result of our cumulative experiences. And over the years, we've developed a strong essence, a common purpose, a set of principles in a way of doing things that's really unique. So let me get into it. Who are we? We are consumer-obsessed. Above all, Deliveroo is a consumer-first business. When we started for more than a year, for more than a year, I was the only person on the ground. It was very lonely. I just signed the first 20 restaurants on Deliveroo walking up and down the Fulham Road. That's Rossopomodoro. That's the first restaurant we signed up. I also had to wear a Kangaroo costume handing out flyers. I had to make product decisions. I was also the sole customer service rep. And this was at the same time as doing deliveries every single day. So that's the Yamaha scooter I used to ride doing deliveries. I used to wear my headphones under my scooter helmet, answering customer calls, doing deliveries at the same time, and I definitely don't recommend that. But what I learned from that was I saw all the pain points of the consumer, and I wanted urgently to fix them. And I emotionally felt the consumer's issues that I still feel them today. And I still do these things, except wear the Kangaroo costume, which I absolutely hated. Now the team, we all do this. The team places a huge premium on primary research and figuring things out from first principles. We focus on edge cases. We do deliveries, we pick groceries, we talk to consumers. We do shifts working in restaurants because all of this keeps us focused on what matters, which is every single customer and every single order. They all matter. And that obsession that was hammered into me by necessity in the early years is part of the DNA of everyone who works here. Now being obsessed about the consumer also means caring deeply for riders and merchants because we can't deliver for our consumers without them. So we're just as obsessed about helping them succeed on our platform. That's why we own a bricks-and-mortar restaurant right here. I talked about us doing restaurant shifts. We own a restaurant on the Finchley Road called Pizza Paradiso, so we can directly understand the pain points in delivery and take action if needed. That's why we're the first platform to roll out free insurance and other benefits for riders and why we're the only platform with a union agreement in the U.K., driving up rider retention and satisfaction. And in our grocery business, we built a proprietary picking app to help grocers increase picking speed, decrease errors and lower costs. We are innovators. We've been innovators since day 1. When we started this business, we were creating a whole new industry, and we had to invent every single part of it. And we haven't stopped since. We invented a rider model that delivers flexibility and security. We were the first food delivery player to launch a subscription program, and plus we're the first scale delivery-only kitchens in a major way with additions, and we pioneered on-demand grocery in our markets. And when we took this business public, I talked to you about our ambition to be the definitive online food company. And food remains at the heart of what we do, but we've been listening to consumers as well. We've been looking at what are they buying from our grocery category. We've been looking at their search queries, and we've been listening to them directly in feedback sessions. The feedback is overwhelmingly clear. People want more than just food. They want to apply convenience to other verticals as well. They want us to service more shopping occasions. And so today, we announced our expansion into nonfood retail. So what does this mean? Well, it effectively means we're bringing your entire neighborhood to your door, not just restaurants, not just grocers, but your local florist, your local pharmacy or your local DIY shop. And that's going live today. So I want you to take your phone out if you have the Deliveroo app downloaded, which I'm sure you do. Go into the app. Scroll all the way over to the right to the shopping category. And here, you're going to see that we've got different categories available now: electronics, beauty, flowers and so on. And for the first time, we have a dedicated nonfood retail section of the app with specific navigation functionality for these categories. What else have we done? We've also added occasions. So for example, if you go to gifts here, you can see that this takes you to the gifting tile. And here, what we do is we surface products where we think we'll make excellent gifts. And the definition of a gift is highly subjective. But using AI and ML models, we're quickly able to scan through millions of items on our platform to curate a list of suitable products. And you can see what's on there in the app. The reason we're really excited about this is this gifting is just one example. But you can imagine a wide range of other occasions where we can quickly create an automated list of different occasions, right? So example being rainy days kids activities or quiet night in, Thanksgiving, that type of thing, and we can do that immediately. So I think what this makes Deliveroo is it's become a much more powerful platform, which is relevant to people for a whole host of new occasions. And Eric and the team are going to show you more about this later on how it works. I'm incredibly proud of what we've built for nonfood retail in a short amount of time. So a big kudos to Eric and the team for doing that. But innovation is nothing without the ability of getting things done. So we are operators. And we're really, really good at this. This is a highly complex operational business driven by technology but executed in the real and physical world where things go wrong. So many different things have to come together for each order to land perfectly and repeating these steps millions and millions of times a day across our markets. And we're able to do this for 2 reasons: one, attention to detail; and second, our technology platform. We have exceptional attention to detail. We are big believers in the collective power of marginal gains. I'll give you an example of this. I used to host a weekly meeting where we went through every single poor order experience in this neighborhood where the office is. We would have data scientists, product people, ops people, a bunch of people around the table. And this meeting was useful because we would uncover edge cases that have global relevance. But I think just as importantly, it's a cultural reminder to the team that every order matters, and I loved those meetings. They were so fun to figure out, let's get to the bottom of actually what happened on the streets. And then we took those street learnings, and we apply them to more data-driven scalable processes as well as our ML models, but this is the physical world, right? And understanding what actually happened in the physical world is critical for executing this business. Secondly, our technology platform. We've built a technology platform that is highly automated, highly scalable and powered by the latest technology. We use automation across our business to drive efficiency, to get the most out of people, but ultimately to deliver step changes to our CVP. So things such as customer care chats, our consumer and rider pricing algos, our rider fleet planning, to developing our hyperlocal CVPs, and you can hear in the breakout session some of the details around this, but also knowing what a consumer wants when he or she logs in, knowing how long a specific dish is going to take to cook, knowing when it should be collected and when it will be delivered. And a lot of this uses AI, and we've been using large language models for years before ChatGPT made the mainstream. Now let's go on to the final trait. We are relentless. One thing I've learned along the way, no matter how well you plan, no matter how much time you spend looking around corners, a lot of unexpected stuff is going to happen. A lot of things you wouldn't imagine. And it doesn't matter how smart you are or how good your ideas are or even how well you execute. You just have to be ready to adapt to push on and to thrive because there's no map when you're creating a business in an industry. And we are relentless and time a time, again, we found solutions to really hard problems. And this is the thing that makes the other 3 characteristics work. I think about this as an umbrella tree. Our relentlessness underpins all of them. It means we never give up. We never accept that things can't be done or that they can't get better and helps us see opportunities where others don't. Let me give you a few examples. So I talked about walking up and down the Fulham Road for hours every single day until we had 20 restaurants in the beginning. Well, for 2 of them, I actually physically installed a WiFi router for them. I drilled holes through the restaurants of these 2. I drilled holes in the walls of these 2 restaurants installing a WiFi router. And I had no idea how to do it. I literally watched a video on YouTube before I went into the restaurant, but I had to do. In 2015, we launched 11 markets with 3 engineers. I was told it couldn't be done, but we had to move quickly, and we did it and it worked and still one of the most incredible feats to me. In 2019, the CMA blocked our funding round that was led by Amazon, $550 million denied for 18 months. It was an absolute gut punch. We had daily Board meetings, we cut costs, we rewrote our plans, and we got through it. Then of course, in 2020, COVID shut down the business. It wasn't clear if restaurants would be able to be open at all even for delivery. Yet we worked with governments and partners. We changed our processes to become COVID-secure. We reopened and grew rapidly. So we have a drive to get better every single day. And our relentlessness means whatever adversity comes our way, we adapt, we continue to improve and we see it as an opportunity. And that's the drive that underpins everything we do. It's the foundation of everything you're going to hear from the team today. So this is my favorite example. Now very near and dear to my heart, it's an example that my team is super sick of listening to me about. And I don't know if they actually have seen this video. So this is our first office at Deliveroo. We launched this business in a room on Cleveland Street in London without heating or windows, and we found furniture on the street. That's how we started this business. Why do I tell you this? Not a lot of people would have wanted to work in that environment. But we wore it as a badge of pride that a few people in this terrible room were energized enough to want to change everything. That's very motivating to me. So who are we? We are consumer-obsessed. We're innovators, we're operators and we're relentless. These characteristics are very deep in our DNA. So what does this mean for the future of Deliveroo? We'll we transformed the way people work, we've transformed the way merchants do business and the way people work. And now we're going to transform the way people shop. Our platform is getting much more powerful. We're reinventing convenience for our consumers. But as my team is up here on the stage today talking about the future plans, I want you to remember this, this is who we are at Deliveroo. And this is why we're confident that we can deliver everything we're going to put in front of you today. Okay. So let me give you a quick summary of what we're going to talk about today. Here are the presenters. Carlo is going to talk to you about the significant potential in our restaurant marketplace and how we deploy technology and a hyperlocal lens to help unlock that. Eric's going to do the same for grocery, but you just saw the retail launch. He's going to give you much more details about that, which is super exciting. Camilla's going to cover my favorite topic, which is our obsession with delivering perfect orders. Emma is going to talk about Plus and our ambitious plans to expand it. Adam is going to talk to you about the progress we made on our ads platform what's coming next. Scilla is going to take you through what does this all mean in terms of financials. And then I'm going to come back to share our vision for how does this come together for the consumer and to summarize everything you've heard. And finally, Dan, for those of you here in person, this is a real special treat. He's going to introduce our breakout sessions, and we're going to give you the chance to meet with members of the team to go under the hood about our technology. I highly encourage you to attend if you want to dig a little bit deeper. So with that, I'm going to hand over to Carlo.

Thank you, Will. My name is Carlo Mocci, I'm Deliveroo's Chief Business Officer for Europe and Global Restaurants. All of my career has essentially been in food. I started at McKinsey, where I became a partner of the consumer practice. And at McKinsey, I served grocers and food service operators across Europe and the Middle East. I then moved on to Amazon. Initially, I joined to lead a product team that was setting up Europe as a region, or Pan-EU, as we called it. And then I moved into a GM track, ultimately running grocery, Amazon Fresh for Europe and Japan. I joined Deliveroo in 2020 to lead the U.K. and Irish business. And since earlier this year, I'm also responsible for Mainland Europe and our global restaurant vertical. Today, I'm going to cover how we drive growth in the restaurant business through hyperlocal improvement to our consumer value proposition, supported by technology and innovative scalable programs. You heard from Will in 11 years, Deliveroo has grown into a large global business with presence in 10 countries. However, there's still substantial growth opportunity for the restaurant business across our entire country portfolio. Starting with early-stage markets like France or Italy. In these markets, food delivery penetration is 2 or 3x lower than the further developed markets like the U.K., UAE or Hong Kong. There are no significant barriers to growth. And the growth in penetration is essentially driven by new customers adopting -- entering into the category, and this is pushed by favorable demographics. So what we see in these markets is essentially the same patterns we saw in our further developed markets with a 5- to 10-year lag. And another way of looking at this, on the right-hand side of the chart, is to look at the consumer penetration. Here, you can clearly see how many adults in these markets are yet to try our service. But the growth opportunity is also huge in our further developed markets. Taking the U.K. as an example, in the U.K., our penetration is below 50%. So there's an opportunity to develop the customer base by adding new customers and reactivated lapsed customers. But even for many adults who are Deliveroo customers today, ordering a takeaway is not yet a habit, a weekly habit or a monthly habit. So we can do a lot here, improving our consumer value proposition to drive frequency and in turn, activate substantial business growth. So I'm going to spend the next 10 minutes explaining how we drive this consumer value proposition improvements. Our business, you've heard from Will is a presided marketplace connecting consumers, merchants and riders. Because of the hyperlocal nature of this marketplace, it is required to optimize investments and prioritize actions at a hyperlocal neighborhood level and not at the national level. So later on, one of the breakout sessions will actually be on what we call the hyperlocal engine, which is the engine that powers all of this. And these breakouts will be led by our product and data science teams so you will have ability to go into more details. But let me explain how it works at a high level. So for each neighborhood, on a weekly basis, we reassess continuously our performance in terms of absolute and relative performance on the vectors of the consumer value proposition. We obviously look at our market share and we look at our profitability. How are we doing? What's our dashboard for each neighborhood? And then periodically, we also reassess the intrinsic potential value of each neighborhood. This is done essentially on 3 dimensions. The first one is demographics. So we look at what customer segments, what is the level of affluence, what is the predictive frequency they can develop. We look at the merchant types available in a neighborhood as that informs our margin mix. And then we look at the topography because that is a good predictor of when the neighborhood develops the maturity, what are the types of network efficiency we can extract from it. So on the basis of performance and intrinsic potential value, we then determine where to focus our resources and what levers to pull to drive the improvements to the consumer value proposition. So today, I'm going to cover 3 specific levers within the CVP, selection, price and consumer experience. Let's start with selection. This is probably my favorite slide. Customers can find an incredible selection of national and local high-quality restaurants across various cuisine types, price points and occasions. And we have so many different places here in the team. Many of these restaurants are actually exclusive to Deliveroo. And I'm sure you will recognize many of these brands for the markets and cities you are familiar with. And I'm pretty sure you can also confirm these are very high-quality brands. The real question is how do you actually identify the best restaurants at scale across 10 markets and hundreds of cities? How do you do it without having an army of people? Let's see how it is done. This slide illustrates how we use data science to inform our selection and account management process and decisions. So we started using multiple data sources to identify the new restaurants that we don't work with yet. Then we use a machine learning model that takes -- ingests a number of data points internal and external, and assigns a predictive score that essentially predicts how a restaurant is going to perform on the platform. The beauty of this is that we use the actual performance of the restaurants to back test and improve these machine learning models continuously. We then assign the highest ranked leads to our best sales team. And this team is also incentivized in terms of commissions on signing the best restaurants that we predict will do better on the platform. And so we really drive their actions. And finally, once restaurants are signed and they are on the platform, we use another machine learning model to prioritize commercial and operational investments on the best performance. And again, this supports -- some of this is automated and a lot of this supports our account managers on the next best action to pull. This approach is important not just because it informs how we grow the restaurant portfolio overall, the selection count, but most importantly, on how to improve the intrinsic quality of the portfolio. And so if you look at the results over the U.K., over the past 2 years in the U.K., we've grown our number of high-quality restaurants but double the pace than the growth of the total selection. And so we're really improving the quality of our selection. And in fact, if you look at the performance of the restaurants, these restaurants drive 2.5x the GMV on a per site basis versus the lowest growing restaurants. So these are the U.K. results, but these approaches already rolled out to all of the markets internationally as well, which allows us to gain leverage on the technology investment that underpins this approach. Now let me move to the other areas of the CVP. Consumer trust. The best e-commerce marketplaces deliver great selection at great prices. They offer flawless and consistent service. And if any issues arise, they resolve them gracefully. By doing this, they establish consumer trust. And in turn, that trust leads to higher retention and higher frequency. So building consumer trust is really the key to unlocking long-term growth potential. As a food delivery industry, you saw Will's pizza. We haven't done a great job at this as an industry. There are a lot of things that certainly can go wrong in the industry. Delivering hot food fast is hard. However, the defect rates are simply too high. And then when it comes to pricing, my favorite topic, 1/3 of customers tell us that markups versus the dine-in prices are a barrier to ordering more frequently or not ordering at all or barrier to ordering at all. They're not convinced that the price they're paying is fair, yet many restaurants on the platforms are still deciding to charge significant markups. If you take those 2 things combined, those 2 things erode consumer trust and they limit growth. So we think that we can change this. We think that we can change it, and our teams are working very hard to do that. I'm going to explain how we're going to do this on the 2 sides of the consumer trust equation. Let's start with delivery experience. Improving the delivery experience has always been a focus for the company. It's Will's favorite topic, and we're proud of the continuous progress we made over the years. My colleague Camilla will talk more about this later, but I'm going to give you a few tidbits. First of all, we constantly aim for better speed. Over the past year, we have managed to reduce average order duration globally by 1 minute, which is a really meaningful reduction if you think that this is an industry where duration is measured in seconds, not in days. This has been done despite 2 things. We've expanded the catchment areas, and we've grown the overall selection that we carry. And we've done this by reducing inefficiencies like, for example, the time that riders wait at merchants to collect orders. Secondly, and more importantly, we've actually become more reliable. We've taken a big stride here, and we've reduced the defect rate by 10%. And what is most important is that 10% reduction comes in the most trust-busting consumer outcomes, such as orders never delivered or missing items. So these improvements have driven an improvement of global NPS by 11 points, which clearly demonstrates how we're building consumer trust over and over on this side of the equation. Now let's look at the price integrity side. Think what customers really want. Customers want savings. They want promotions, they want low prices. We directly deliver this through Deliveroo Plus and targeted promotions. Deliveroo Plus is our subscription program, offers free delivery and other benefits. We've been the first to introduce a loyalty program in this industry. My colleague Emma will talk later after the break about our plans for Deliveroo Plus. And regarding our promotions, we are on a constant journey to innovate on how we target them, how we improve the efficiency. Essentially, what we want to do is to give more to the customers whilst at the same time, maximizing the efficiency of return on investment. These direct levers, key focus for us, they continue to be -- they are critical to deliver value to the consumers. But today, we're actually going to talk about something which is really innovative and unique. We have developed tools to incentivize our merchants to provide fair prices and price integrity in combination with create service. We do this through 2 programs. The first one is the commercial architecture. This is a framework that we use to work with our largest-scale partners. The value program takes the same principles and then helps scaling those principles from the large partners all the way down to the entire selection. And so think of tens of thousands of partners being touched by this. We really believe that this is very unique in the industry. We think that it's very hard to replicate, and we think that this gives us an edge. Let me dig into those 2. Firstly, commercial architecture. The commercial architecture has the objective of improving consumer outcomes. It is a commercial framework that aligns our incentives with those of our largest partners. And this is cascaded from the HQ down to a franchisee level, if the company is a franchise model, and all the way down into individual store levels. So there's a full alignment of incentives. We provide marketing and commercial investment, in return for better service and value. This is done in a standardized, transparent way across every restaurant segment. This approach builds consumer loyalty, builds consumer trust, accelerates the business growth for us and for the partner. And this is a great example of the type of innovation we are driving in our commercial approach. When I speak to partners, many of them are really, really excited by this new approach. They are embracing it, and they see this vision fostering future growth. These work well with the largest partners, but the problem is it doesn't scale to the smaller partners because it's intrinsically complex. And so how do we actually scale the same principles without the complexity over tens of thousands of merchants? We do this through our newest innovation, the value program. This is really something I'm highly excited about, and the work our teams have done here is incredible. The value program incentivize restaurants to provide better service and great prices to our customers. This program went live in the U.K. at the beginning of November. We have plans to roll out this program to other countries in the coming year, and the first results are really encouraging. Let me walk you through how this really works. So the program rates the partners on a value-for-money scale along 3 vectors. Price integrity. This is essentially the price markup you pay buying on Deliveroo versus a dining price you would pay. This is something we hear loud and clear from customers as one of the key barriers to ordering. Customers don't understand why they would be charged the markup when they're already paying delivery fees and service fees for a delivery. And so this is really, really key. The second element, consumer ratings. This is essentially the quality merchant provides and is directly rated by consumers. So we have plenty of data here. And the third, the service. Here again, we got a lot of data. We are improving our data signals. You will see it in the breakouts. And in this specific case, we really focus on availability, we focus on reliability, speed, and in general, low defect rates. So the restaurants we are providing fair prices or great prices, high quality and great service, see their traffic increase. We direct more traffic to them with dedicated merchandising. We have dedicated offers, and so they enjoy the benefits of this program, and they see growth into this. But the opposite is also true. Now if you look at the app and you browse on the rest of the section, you will actually see a new tab called Deliveroo's Choice Restaurants. These are the restaurants that provide the highest scores across 3 dimensions. They have no markups, they have great quality. They have great service. And I'm someone who's really into this, this has essentially become the way I shop the app. How have we done this program? I hope by now, you understood this is actually very complex in terms of amount of work that goes behind it. So we've developed multiple tools and processes to be able to acquire dine-in prices at a huge scale, from national chains all the way down to the local single-site independents. We think this is really unique. It is pretty hard to replicate, complex to build. And this program, together with the commercial architecture, provides a comprehensive way for us of delivering great value for money in partnering with our partners and, ultimately, enhancing consumer trust. Let me wrap it up. Four takeaways. We operate in big addressable markets with significant growth potential across the entire country portfolio. We use data science to improve consumer value proposition at scale on a hyperlocal level. Winning consumer trust across delivery experience and price integrity is the key to unlock and accelerate category growth. And finally, our innovative commercial programs, the commercial architecture and the value program through those will align commercial incentives across restaurants of all scale to improve consumer outcomes. With this, I'm going to hand it out over to Eric, who's going to talk about retail and grocery.

Great. Thanks, Carlo. Hi, everyone. I'm Eric French, COO of Deliveroo. Very excited to be with you here today talking about our grocery business and also talking about our new retail business. So as you heard Will mention earlier, retail is our new vertical, where we're going to be offering customers a range of new nonfood products and categories, including things like electronics, gifts, flowers and a bunch of other categories I'll go into here in a few minutes. Before I get into grocery, let me maybe just start with my quick background. So I've been at Deliveroo for 3 years now. Before this, I spent 15 years at Amazon, spent most of my time at Amazon in and around the grocery and consumables business. I started off working on the original launch of Amazon Grocery back in 2006, spent many years working on building and scaling and growing that business. And then my last role at Amazon was leading the U.S. consumables business, which is about a $60 billion business for Amazon. So as you can probably tell, online grocery is a space I have a ton of passion for, very excited to hopefully share some of that with you all here today. Okay. So before I get into where we're headed with grocery at Deliveroo, and maybe just start with a bit of history and a bit of context, which I think is important to kind of start with here. So if you rewind about 6, 7 years ago, really the only way for customers to buy groceries online was through the kind of legacy big basket model. So this was next day scheduled delivery from players like Tesco, Ocado, Amazon, et cetera. In 2018, we had the idea to run a small test of listing a handful of grocery items on Deliveroo through a partnership that we had with Co-op. Now if I'm honest at the time, most people internally did not think this was going to amount to a whole lot more than maybe selling some late-night beers and kind of salty snacks. What emerged really pleasantly surprised us. Customers were very excited about buying normal grocery items on Deliveroo. Some of our top sellers that quickly emerged were things like milk, eggs, spread, chicken, all the kind of normal things you would buy in a grocery store. And the fundamental insight that we took out of this was that customers love the fact that they could use Deliveroo to basically avoid planning in advance. They loved that it can be 4:00 on a Wednesday afternoon and they can open up Deliveroo and get the ingredients they needed for dinner on Wednesday night delivered in this incredibly convenient manner. And in exchange for that incredible convenience, customers were willing to pay a bit of a price premium. And I'll share more here in a few minutes about how we think about the price premium, what I mean by that and some of the things that we're doing in that space. Okay. So that was how we got started. Fast forward 5 years, we've built a huge business. So grocery is 11% of Deliveroo's GTV in the first half of this year. There's still lots to be done, lots of work, lots of runway here in front of us, but we're really proud of what we've built so far. We've built a big business in a short period of time here. Okay. So here, I want to talk a bit about how we built the grocery business, which is particularly important because you're going to see some applicability in our approach to grocery for how we're going to approach retail here. So really 3 main points to make here. So first, in building grocery, we had to build out the supply side of the marketplace. So we had to bring on new selection. And so this was large grocers, people like Morrisons, Waitrose, Carrefour, it was also your local independents. So with the supply side built, we then needed to drive demand to meet that supply. And in building grocery, we spent a bit of money on paid marketing. But the huge advantage we had was that we were able to take the daily organic traffic we have coming to Deliveroo day in and day out and simply point those customers towards this new offering that we had. So supply side, demand side. And then the third bit is on the technology side. One of the other big advantages we had in building our grocery business was that the grocery business was basically able to ride for free on top of all the technology we have built for restaurants. So things like payments, our rider app, et cetera, the grocery business basically plugged in seamlessly to the technology we had already built for the restaurant side. But there were 2 places where we had to invest time and effort to build out the grocery -- the tech side of grocery. So first, on the customer-facing side, we had to create basically a brand-new customer experience that enabled customers to shop through the thousands of grocery SKUs that we had, which is a fundamentally different experience than shopping or than using the service for restaurant delivery where you're scrolling through a much more limited kind of menu assortment of items. And then the second thing we had to do, and Will mentioned this a bit earlier, is build out this dedicated picking app. So we had to create technology that enabled the in-store worker at the grocery store to walk around the store and do all the pick-pack of the item and deal with use cases like out-of-stocks and substitutions. So those are 2 big things we had to do. But again, in general, the grocery business was able to be built on top of our existing restaurant technology, which was a big advantage we had. Okay. So I want to talk here a bit about the industry, the TAM and kind of where we see things headed. So on the left side of the slide here, a couple of data points, probably won't be hugely surprising to anyone here. So grocery is a very large TAM. So $600 billion in the markets we operate, and online is still relatively underpenetrated within grocery, certainly relative to other kind of verticals of e-commerce. On the right side here, have some data from a recent McKinsey report that suggests that on-demand will increasingly take share from online grocery. And we obviously tend to agree with this point of view. But I think there are 2 really important points to make here and 2 really important things that we need to do in order for on-demand to take share within the overall online. So the first is we need to continue to enable customers to fulfill more and more of their shopping missions through an on-demand model. And I'll get to this in a minute by what I mean by this. And the second thing that we need to do to unlock more of the TAM and push into more of the opportunity here, and this kind of relates back to this price premium in exchange for convenience, is the extent to which we can reduce that price premium that a customer has to pay. We will unlock more and more of the TAM over time. So more shopping missions and reducing the price premium are 2 of the huge unlocks to growth, and I'll go into more on both of these here in a minute. Okay. So here, I want to talk a bit about the unit economics and some of the benefits that the grocery business drives back on to kind of the main restaurant business. So on the left side, you can see the unit economics of grocery next to restaurant, 2 points to make here. So first, the unit economics of grocery are a bit under our restaurant unit economics, but not by much. And second, you can see we've driven improvement not just in the restaurant unit economics year-over-year, but also in the grocery unit economics year-on-year. On the right side here, sharing some data that points to the benefits that the grocery business drives back onto the rest of the platform. So we like the stand-alone unit economics of grocery. But in addition to that, it has these kind of larger flywheel benefits. And those are the -- 3 things. So first, the grocery business is basically entirely incremental. Second, our grocery customers are our most engaged customers. And third is when a customer starts using grocery, we see an increase in the retention rate on the platform overall. Okay. So now I want to come back to where we're headed and how we're driving growth. So again, one of the most important things that we can do to unlock more growth here is enable more shopping missions over time. And so on the left side of the slide here is basically what I would characterize as our focus to date. So we have created a customer experience where customers can come in to Deliveroo, shop from about 3,000 SKUs at any given grocer. We have more SKUs than that overall, obviously, but 3,000 SKUs at any given grocer, build baskets of up to about 30 pounds and get that all delivered within the convenience of on-demand delivery. The opportunity in the future, as we see it, is to push more into this middle bar -- this middle part, in addition to continuing to offer the left side here. So we see a customer experience where customers are able to shop from around 10,000 SKUs at any given grocer, build baskets in the GBP 30 to GBP 60 range and importantly, continue to get it all delivered with the delivery speed and the convenience of the on-demand model. Now there's a really important point here. And one of the things that gives me just a lot of confidence and optimism is we have high conviction that we can build the business here in the middle without any fundamental reinvention of the on-demand model we've built already. The things that we have done to get us to where we are now in grocery, we have confidence of the same things that will enable us to go after the larger basket size here. And so this is things like continuing to expand the assortment in the range with grocers. It is continuing to optimize the customer-facing technology, continuing to optimize the partner-facing technology and then doing all that with a high degree of operational excellence, which, as you heard Will mention earlier, something that's just in our DNA. Okay. Then the other growth driver we have here is to reduce the price premium that a customer has to pay in exchange for the convenience of on-demand delivery. And so here's how we think about it. So we have very clear opportunities to expand the unit economics of our grocery business, pushing into larger baskets and higher AOV will improve our per order unit economics. We have all sorts of opportunity to reduce our delivery cost per order through network efficiency initiatives, and Camilla will go into some examples of that later. And then as you're going to hear Adam talk about later, we have a very big opportunity in front of us with the FMCG advertising opportunity. So clear opportunity to expand the unit economics. We are then going to take that unit economic improvement and do 2 things with. First, we will drive improved unit economic, will drive increased profitability over time. And you saw a couple of slides ago how we've already been doing that. So we will continue to drive improvement in our grocery unit economics. At the same time, we will also take some of that improvement and choose to reinvest it into growth. And the way we kind of think about this is we spend a lot of time thinking about the interplay between the commission rate that a grocer pays, the price premium that a customer pays and the unit economic improvement that we're able to drive. And there's no magic kind of perfect answer as to how we balance those variables exactly. But what I will say is we will drive both margin expansion and choose to reinvest in growth. And where we do choose to reinvest in growth, we view it as a very explicit investment decision. It's not something that we take lightly. We spend a lot of time putting rigorous analysis behind our decisions to reinvest some of the savings on the left into driving growth. Okay. So to wrap up on grocery before I move to retail here. So 4 key takeaways. First, we built a very large business; second, there is a very large opportunity with attractive unit economics; third, we will continue to drive growth here by unlocking more and more customer missions and customer occasions over time by moving into medium-sized basket; and fourth is we will also drive growth by reinvesting some of the profitability improvement into less of a price premium for customers and better value. Okay. So that's online grocery. Now really excited to spend some time talking about our new retail vertical. So as you heard Will mentioned earlier, we have just this morning launched this new service in the U.K. and also the UAE. And so very excited to walk you through some of how we think about it and some of the progress we've made thus far. Okay. So maybe just starting at a high level, the way we think about retail is it is simply just a very natural extension of the marketplace we have already built. So similar to grocery, it's an opportunity for us to build a big business by leveraging two sides of our marketplace, our big customer base and our big network of riders, while also adding in new partners and new selection to the third side of the marketplace. Okay. So one of the things that gives me the most conviction in our ability to go after the retail opportunity is the fact that the playbook in the set of frameworks that we've developed in building our grocery business are very clearly applicable to how we're going to go after retail. So you obviously recognize this slide from the one I shared kind of a few minutes ago. And the point to make here is the approach to retail will be very similar to our approach to grocery. We need to add selection and new partners. So this is both big retailers as well as local independents. We need to drive demand to this new selection. We'll spend some money on paid marketing, but we will largely leverage the existing organic traffic we have coming to the platform day-in, day-out. We need to build some new technology. Although there's an important point to make here, which is the retail business will be able to leverage the technology we've built for the grocery business just about entirely. There are some places that we need to do things a bit different for retail. I'll share some examples of that later. But I would say, in general, the incremental build of retail on top of grocery is smaller than the incremental build required to build grocery on top of restaurants, which is a very exciting thing to us. And maybe the last thing I'll mention that gives me lots of confidence and lots of conviction is it is the same team building retail as the team that built our grocery business. I'm very excited. Hopefully, you'll all have a chance to meet some of them here later today. They'll be at the breakout session later. Okay. So one of the other things that gives me a lot of conviction in the opportunity here is what we're hearing from customers. And so as Will mentioned earlier, we already have all sorts of signals and data points we have from customers indicating that on-demand retail and on-demand retail from Deliveroo is something they want. And so in our decision to go after retail, we started with the customer and really kind of did the work to understand what they want. And sharing here just some of the data points that we've been looking at. So we have search data with customers already searching for nonfood products and categories. We have done qualitative insights work actually getting out there and meeting with customers and sharing some of the examples here of what customers are telling us in terms of their desire to see an on-demand retail opportunity from Deliveroo. And then on the right side here, we are simply already selling many nonfood items in relatively meaningful quantities. Now in fairness, the kind of cynic could probably look at these numbers and say, okay, relative to fill in the blank, e-commerce player. These are relatively small numbers, which is, I think, fair, but two important things to keep in mind here. So first, we have done basically nothing to make any of this assortment discoverable. So up until this morning, anything that was outside of food was highly buried within our experience and yet customers are still finding it. And second is, it is very, very early days in our building out the assortment and the selection. And so some of these search queries you see here are oftentimes ending up in a not terribly robust set of selection that customers are looking for yet customers continue to come back and search for these items. So early sparks here of a big opportunity in front of us. Okay. So here, I want to share just a bit of data on the TAM, how we think about the TAM and then just how we think about the industry trend overall. So clearly, a big opportunity from a TAM perspective. So GBP 700 billion in the markets we operate. This excludes some things that we think are not addressable with an on-demand model, so large appliances and other things that are not terribly accessible through our model. We think it's a GBP 700 billion opportunity. The maybe kind of bigger point to make here is the way we think about this is on-demand retail is really just the next chapter, the next evolution of a trend that has been going on for basically 30 years at this point. Customers simply want to buy more and more online. And the things they buy online, they just want to come faster and faster. And so 30 years ago, we were in a world of Amazon selling books and it would get delivered within 3 weeks. We now live in a world of getting your groceries delivered within 30 minutes. And we have conviction that in the very near term, we're going to be in a world of customers wanting and expecting to get a wide range of items delivered with the incredible convenience and speed of on-demand. And I guess, personally, someone who spent the last 20 years working in e-commerce, the fact that -- or the kind of concept that this is where things are going, it's just crystal clear to me. It could not be more clear the direction that all of this is headed. So it gives me a lot of conviction and the opportunity. Okay. So here, I want to share just a bit more specifics on how we're thinking about going after the opportunity and kind of how we're thinking about pursuing things. So if you go back to how we built grocery, it really started with what we kind of think of as the emergency need use case. So this was the -- it's Friday night, I'm throwing a party, I've run out of beer and ice and Deliveroo kind of saves the day basically. And as we added assortment and customers build the habit over time, we, on top of this emergency need use case, have built out what we think of as everyday convenience. So again, this is -- it's 4:00 p.m. Wednesday, and I'm able to open up Deliveroo and do my grocery shopping for Wednesday night dinner basically an hour or 2 in advance. And so that was how we built grocery. We see the same opportunity existing within retail. So the kind of most obvious opportunity here is to go after the emergency need use cases. And there are dozens, if not hundreds, if not thousands of missions and customer journeys to go after here. So there is a huge opportunity solely in the lower left here. But we are not constraining the kind of ambition here to only go after the emergency need use case. We also see an opportunity to build everyday convenience within on-demand retail as well. And conceptually, the way we think about this is as we build the assortment and again, as customers kind of build the habit of turning to Deliveroo for on-demand retail, they will increasingly look to us to help them just simply avoid the hassle of planning in advance and having to go to a busy store when maybe they have other better things to do with their time. So here, I want to share a bit about just the specific categories we're going after and also how we're thinking about prioritization. So the categories I've listed here are the ones that we intend to go after in the kind of very immediate term. This is not the totality of the opportunity. But these are the ones we're intending to go after most immediately. The -- maybe two points I want to make here are just how we think about prioritization because, obviously, there's a lot to go after and doing this right is going to require focus and discipline. And so there are really 2 big signals we'll be looking at to kind of prioritize us and guide us. So first is, again, just going back to these customer signals. So it's all the things I shared a couple of slides ago. Search data, insights, existing product sales and doubling down on categories or product types that seem to be working well. And then the second thing we're going to do is just simply going after the most low-hanging fruit. I think there is so much opportunity by just going after this emergency need piece in the near term that there is just all sorts of low-hanging fruit here in front of us. And so this will be things like last-minute gifts, things like my kid is asleep and I need nappies. It's things like it's a rainy Saturday afternoon, I need activities and games to kind of entertain my kids. It's a mid-DIY project, and I've run out of nails or whatever -- I'm not a DIY guy, but run out of DIY supplies. I will still go with that part. Okay. So -- all right. So I've shared kind of a bunch about kind of the opportunity and kind of conceptually how we think about it. Now excited just to get into just a bit more of the kind of tangible progress we've made. So on this slide, I want to share the progress we've made on the selection side. So if you go back to kind of the 6 building blocks slide, the upper left part there, this is the progress we've made thus far. So kind of 3 examples I want to share. So on the left, our existing grocery partners already have a large range of nonfood items. And so sharing an example here of this really cool LEGO selection that we have from Morrisons available like, I think, across 100, 150 of their sites already, great selection to have on the site here, right, as we head into Christmas. The middle bit here, share an example of a really cool local independent with some kind of fun giftable kind of products here. And one of the partners you'll see in that gifts tile that is in the new experience that Will was walking through earlier. And on the right side here, I want to share an example of the sort of kind of large partner that we are going to be working with soon. Now Screwfix is actually not live on the site at the moment. They will be live soon in the coming weeks. But they are one of the big partners that we're very excited to have. And for those of you not from the U.K., Screwfix is a big U.K.-based DIY retailer that covers about 95% of the U.K. population within a 20-minute drive. So very excited to have them on the platform here soon. Okay. So this might be my favorite part of the presentation. So the next few slides I'm going to share some of the new things we've done from a customer experience perspective. I'm going to whiz through this relatively quickly. But again, we have the breakout session later. The team can go into any of this in more detail, if you want, I'll be available for questions as well. And then again, as Will mentioned, all of this is now live in the app. So after the presentation, feel free to go and use the app and play around with it and so you can kind of see it for yourself real time. But just to kind of walk you through maybe a few of the examples and kind of how we think about this. So on the left side here is our new shopping home page, this one Will mentioned earlier. You can see kind of the breadth of categories that we now have, a really key part of building awareness and kind of driving discoverability for customers. On the right side here, you see an example of the sort of merchandising that we now have live within the app, again, going back to the kind of tapping into organic traffic bit. This is an example of the sort of thing where we will be highlighting the new range of products we have to customers that are live in the app. Here's an example of search. So we've been spending a bunch of time on search, and this is actually a good example of the sort of thing that will be a bit different for retail than restaurants and groceries. So retail will be a more search-driven customer experience. You'll see more customers doing kind of spear phishing-type customer missions. And so we've had to build a whole new search experience here, probably looks quite different for those of you who are long time Deliveroo users, much more kind of item-led experience to allow customers to find the right items. Sharing here our new browse experience. So yes, retail will be more search-driven than grocery and restaurants, but there are absolutely going to be categories where customers need a bit more inspiration. They're maybe not sure exactly what the specific item is. So things like toys, flowers are examples of categories that will be more kind of browser-oriented than search-oriented potentially. There's also actually really important nuance here, maybe just to spend kind of 30 seconds talking about. So behind this kind of seemingly simple and clean experience, there is an area that we spend a lot of time investing in over the last year, which is our product catalog. So there's technology that basically determines what category to surface for these various items. And there is a lot of complexity that we have helped solve for behind the scenes by creating this catalog, which then when coupled with some of the machine learning models we've built determines whether or not a given item and partner shows up if I'm searching within toys versus baby and child or electronics, for example. And by the way, we can go deeper on all of that later. Okay. The last bit here, I want to share this new give-a-gift functionality that we've built. A couple of fun things we've done here. So there's this new integrated message flow in the experience, and then there is this kind of fun digital unwrapping experience on the right. Okay. So I'll wrap up here with just a bit on how we think about the unit economics and size of the prize for retail. So we have confidence that we can build a business with unit economics that are on par or better than our grocery business. So we think the AOV will likely be higher just given the underlying kind of product type of the retail items versus grocery. We think the commission rates will likely be better just given the underlying margin characteristics of retail versus grocery. We think there's an opportunity for the delivery cost to be lower just given that these items are not perishable, certainly relative to restaurant food and relative to much of grocery as well. And then we also see a large ads business. We see a big ads business in grocery as well. But we see a large ads business for retail. And then lastly, just bringing it all together, we see an opportunity to create a business in the region of GBP 700 million of GTV by 2028. Okay. So to wrap up on retail, we will go after the opportunity using the same approach as we use for grocery. There is clear appetite for on-demand retail. We're really excited to share with all of you today the new in-app experience, which launched this morning is now live, and we will build a large business with attractive unit economics. So thanks very much. I'll now hand it over to Camilla.

Hi, everyone. I'm Camilla and I joined Deliveroo in 2016. So Will's intro brought back a lot of memories, mostly fond. And actually, over the years, I've led a number of teams here, tech and nontech and across operations, marketing and commercial. And today, I'm on our exec team and cover a number of areas, but I'm here to talk specifically about the Deliveroo experience. As you heard, it's Will's favorite topic, and it's actually also mine. So what do we mean by the delivery experience? This is everything that happens from the moment a customer places an order right through to them receiving and consuming that order. And so you can see from that, that this is obviously at the heart of everything that we do and really core to the consumer proposition. And as Carlo said, delivering flawless orders, effectively eliminating defects is key to building consumer trust and therefore, represents a growth opportunity. We're actually already ahead of the pack in terms of service, but eliminating defects, creating more flawless orders still represents a big opportunity, an opportunity to drive retention, to drive frequency, to drive spend and actually also to drive new customer acquisition, if you think about the propensity to recommend us to friends. What's more, eliminating defects actually saves us money too, because it reduces the amount that we have to spend on compensation costs. So the cost of putting things right with customers. There's actually an even bigger margin opportunity in this space, though, and that is from driving efficiencies, which allow us to reduce delivery costs, which are our biggest cost line. More on that later, but first, let me talk to you about defect reduction. I'll be honest, I did consider changing the title of this section because defect reduction obviously doesn't sound very positive. But I think it's important to be transparent about what's going on here. The overwhelming majority of orders go perfectly, but there is a small percentage that do go wrong. And whilst we're ahead of the competition, as I mentioned, it's still the case that a single defect can be a real trustbuster for customers. And so defects effectively can have an outsized impact. Eliminating defects is hard. Delivering food from A to B sounds simple, but it actually takes dozens of things going right in the background for an order to be flawless. The fact that it's hard just makes it an even bigger opportunity for us though. And that's because we're uniquely well placed to go after it. There are a few reasons I say that. The biggest factor might sound soft, but it is a real differentiator. And it's that obsession with the customer experience and every operational detail that underpins it. As Will mentioned earlier, it's not a fad. It's deeply, deeply ingrained and is part of our DNA. And it actually manifests in lots of practical ways. It manifests in how we think and behave. So to give you an example, every delivery employee will regularly do deliveries as a rider or do a shift in a restaurant or a grocery store or take customer chance, which means that we all live and breathe the marketplace. It's also reflected in how we engage with merchants and riders. Carlo talked about the value program that we have. And that's an example of that on the merchant side. And on the rider side, we have features like post-order tipping, which incentivize riders to go the extra mile. And in addition, we bring together merchants and riders to jointly solve operational challenges. Most importantly, perhaps, it's reflected in our processes and in our tech. Processes and tech that we have built and iterated on and improved over years. And I'll touch on a few examples in this presentation, but you'll also get to experience it firsthand in the demos if you're here. Of course, our logistics model also helps because it means that we have real-time data on every order at every stage, and that means that we can detect if things are going wrong and intervene. To give you a taste of how we're eliminating defects, I want to talk through a recent case study, which is around eliminating OMDNR. You're probably wondering what on earth an OMDNR is? Well, it stands for Order Marked Delivered Not Received. And there's 5 little letters, which are not so little on the screen, represents the worst customer experience there can be. A customer just doesn't receive their order. OMDNRs used to happen for a few reasons. For example, a customer might have inputed in an accurate address or maybe the customer's address was really tricky to find and the rider ended up giving up, or there might even have been foul play on the customer side where the customer would say they hadn't received the order when actually they had just to get the compensation. So it was super hard to tell apart those scenarios and to understand the root cause, to know whether it was the rider or the customer, legitimate or dirty behavior. And that made this a really tricky defect to unpick. Whilst OMDNR doesn't occur often, they were super expensive for us. And that's because we had to pay the merchant for the order, the rider, the delivery fee and then compensate the customer the full value of the order as well. By far, the biggest cost, though, was in the form of broken trust with genuine customers. And in a cost of living crisis, that trust is more important than ever. So at the beginning of this year, we decided we had to really go after these edge cases. There wasn't one silver bullet. It took a concerted push on all front and a multitude of fixes. So to give a few examples, we changed the way that customers input their addresses to increase precision and accuracy. And we also nudge customers who identified as having problematic addresses to update those to avoid issues on future orders. We also rolled out a feature that we call positive handover, which you might have seen in the app. It's when the rider asks you for a code to confirm that you've received the order. I'm really proud of what the team did and the impact that they've driven, which you can see on this chart here. So you can see that the OMDNR incidence rate dropped steadily and dramatically over the period. And alongside that, compensation costs also fell. That represented an annualized saving of GBP 20 million. And alongside that, you can see the increase in customer satisfaction, which goes to show that whilst OMDNRs were rare, they were having a real impact on customer experience. OMDNR is just one case study, but it reflects the approach that we take to all defects. We relentlessly push to improve and to reinvent and to raise the bar, not just for ourselves, but for the industry as a whole. Now on to how we're driving delivery efficiencies to reduce cost and drive margin. Let me first talk a bit about what we've achieved in the past few years in this space. I have to admit, it's a track record I'm pretty proud of. So let me talk you through each of these charts. On the left here, you can see merchants available. That's the number of merchants that a customer sees when they open the app. And you can see that, that's increased significantly over the period. That was partly achieved by enabling customers to order from merchants that are further away. And so other things equal, you would have expected this delivery cost per order, the purple line here to be increasing, but it hasn't increased in spite of inflation. And why is that? It's because of the delivery efficiencies that we've realized. So effectively, the cost per kilometer has dropped, but that's hidden by the mix shift towards longer distance orders. Crucially, alongside all of this, we've increased earnings for riders. So you might be thinking great, but how much juice is there left here. And the short answer is that there are still lots of opportunities for us to drive savings. And let me be absolutely clear, those savings are not going to be realized by jeopardizing the customer proposition or the rider proposition. Rather, we're going after smarter savings, savings that either benefit all sides of the marketplace or at a minimum do no harm. So let's start with the win-wins. Reducing the amount of time that rider spend waiting at merchants is the perfect example of this. It drives efficiencies for us. It means merchants have fewer riders waiting in their stores. It means riders can increase their earnings potential, and it means customers get their orders faster. So it's actually a win, win, win, win. So how are we actually reducing these wait times? There are two things here. First, by collecting new and better signals, and second, by incentivizing merchants. Carlo already talked about the commercial architecture and value program earlier. So I won't dwell on the incentives. But let me talk a bit about the signals. On the rider side, we're introducing a new physical check in to confirm the rider is definitely in exactly the right place, and that's an industry-leading innovation. And on the merchant side or the food readiness side, we're increasing -- we're introducing more automated signals using things like kitchen display systems that allow us to know the exact stage that an order is at in real time. And so using those new signals, we can improve our machine learning models to smooth the handover, but also crucially, they underpin the bolstering of those incentives that I just mentioned. Unfortunately, not everything is a win-win, which means that sometimes there are [ trade-offs ]. But the real opportunity here is in creating new capabilities that allow us to drive savings without jeopardizing the consumer or rider experience. And there are two major areas of innovation that I want to talk about here. The first is smarter stacking and the second is smarter merchandising. So starting with smarter stacking. Stacking, which is sometimes called batching, is when a rider collects and delivers multiple orders at once. And it drives delivery efficiencies because it reduces travel time. We've been stacking for years, and we're pretty sophisticated about how we make stacking decisions today, but there's still so much more that we can do here. So let me give you an example of a change that we recently rolled out, which unlocked lots of efficiencies. That's multi-pick up stacking. Previously stacked orders had to be picked up from the same merchant whereas they can now be picked up from different merchants. In practice, that means that if there are 2 orders from neighboring restaurants going to customers on the same street, subject to ensuring good customer experience, of course, those can now be stacked whereas they previously couldn't. So they could be delivered by the same rider. The first variant of multi-pickup stacking is live or ready, but there are others to follow. I won't tell you more about that because there is a demo on it later, so I don't want to give too much away, but I do encourage you to go and see that. Now let me talk about another area of opportunity. It's around how we're using generative AI to enrich the information we give Frank, our dispatch algorithm, to make decisions. And we're using that generative AI to better understand exactly what's in order, how well it's going to travel, it's size, it's weight. And we're combining that with other data sources like on outdoor temperature, for example. So to bring that to life a bit, previously, Frank might only have known that your order was placed from a sushi restaurant and would have assumed that the food was therefore cold and could safely be stacked without jeopardizing its quality on arrival. Now if Frank knows that actually your order contains miso soup and the outdoor temperature is freezing that might fundamentally change his decision of whether to stack the order or not. In addition to generative AI, grocery and retail present opportunities to increase stacking without jeopardizing the customer experience. And that's because maintaining temperature is much less of an issue for those orders. And it's also likely that there are more that will be slightly less time sensitive. So that's a bit on stacking, which honestly, I could talk about for hours, but I'll move on to merchandising. Merchandising is all about how we present and promote merchants and content to our customers in the app. The most obvious example is the choice around the order in which we present merchants on what we call the home feed, what you see when you open the app. And that ranking is already cost aware, but only based on a customer's location, not based on the customer's preferences. So to bring to life the opportunity here, let me talk about 2 different customers. Imagine the first customer is coming to the app specifically to order a Shake Shack Burger. And we know that because that's exactly what they've typed in the search bar. In that case, even if the restaurant is far away, we should just get out of the way and make it as easy as possible for them to place that order. Now imagine a different customer comes to the app, and they're just looking for a burger. Maybe in that case, we do include Shake Shack in the list of choices that we return, but below another burger joint, that's closer and is going to deliver faster and hotter and also to be cheaper for us to deliver from. Effectively, we can nudge the customer. There's room for ever more personalization here. So this is a really exciting area for us. Combining smarter stacking and smarter merchandising with the win-win that I talked about on the rider merchant handover, you can see that there really is a lot of juice left here in terms of driving savings. That's a lot of information I've just given you, but I wanted to give you a peek under the hood as it were. But to summarize here are the key points. Most orders are perfect, but defects are real trustbusters. So we are methodically and doggedly going after them to drive growth and margin. To further drive margin, we're going after win-wins and creating new capabilities that allow us to drive savings without jeopardizing the consumer or rider experience. And we're uniquely well placed to go after these opportunities because of the obsessiveness and the relentlessness, which are so core to who we are and what we do. Perfecting the delivery experience often goes overlooked. But in my opinion, it's where the biggest opportunities lie and we're going after them. And with that, I'll hand back over to David, who's going to chair our Q&A.

Great. Thank you, Camilla. So we're just going to bring some chairs on, and I'm going to ask the speakers from the first session to come on stage and also to ask Scilla to join us as well. The finance section follows after this, so we ask you to hold your finance specific questions until after that section. And also please hold your DIY questions for Eric because he will take them offline. And because of the number of people we have in the room and on the webcast, I'm going to ask you to limit yourself to one question each to give as many people a chance to ask questions and a prior warning, I'm going to be strict on that. And please just wait for the mic and announce your name and your institution. So do you want to come to the front, please, and we'll take first question from Andrew.

Andrew Gwynn from BNP Paribas. Just coming back on to the basket size and maybe also thinking about profitability within the retail business, if you look on the app, obviously, you think about Boots, for instance, you could pick up some pain medicine. It's obviously a very, very low basket. So how would you balance the two to make sure that the expansion into retail, you are going into decent gross profit opportunities versus actually the convenience for the consumer?

Eric, do you want to take it?

Yes. So there will certainly be some retail categories or partners that will have products with lower average selling prices. And I think the way we will generally think about that is quite similar to how we run the grocery business. I mean the grocery business obviously has a wide range of low ASP products and kind of subcategories. And the -- we have MOVs, first of all, is one of the kind of important things that we have to kind of put some guardrails on the unit economics. But maybe from a bit more of a kind of customer-centric perspective, the bits that I mentioned earlier around creating this new customer experience for grocery was largely oriented around just helping customers browse through all the items and kind of build the basket that they want to build. And so that same experience will apply to pharmacy or will apply to other categories where there is an opportunity for customers to build baskets of kind of lower average selling price products.

Do you want to pass just behind to Monique?

It's Monique from Citi here. It was a question for Camilla on the stacking and the big opportunity that you see from stacking on the retail and the grocery side. I guess I'm just interested because Eric was also talking about increasing the average basket sizes for grocery and talking about the fact that the retail basket sizes are hopefully going to be at least as big, if not bigger, than the grocery one. So I'm just trying to think about what the opportunity is from stacking? Are you just still going to be using riders? Are we going to be moving more to drivers to larger sort of drops?

Yes, so I think the first thing, which is maybe not obvious is that there's a very -- well, an imperfect correlation between value and size. And so we can have larger order value baskets that aren't necessarily larger volumetrically. And this is why I do think that the generative AI that I mentioned and our cataloging inventory is really key because it means that we have great, basically, insights on exactly the -- well, the size and the weight of items and can, therefore, make stacking decisions based on that. And so that's how I think we can have great confidence that we can effectively navigate that and make optimal stacking decisions.

And so you still will just be using riders [indiscernible]

So we actually already have a mix of vehicle types today. So we have everything from walkers and bicycles and scooters who tend to have a sort of backpack right through to cars. And our dispatch algorithm, Frank, that I mentioned, we call it, is vehicle aware. So we'll consider that. So if you pair great signals on exactly what's in order and the size and weight of it with information about the vehicle, Frank can make the decision of when to assign a car versus a bicycle, for example.

[indiscernible] from Panmure Gordon. One for Eric, if I can. I think the major category of retail that stands out to me is not in your current plans is apparel. Any word on how you view that in the future?

Yes. So you are right. I mean apparel, clothing, footwear is a large category within the kind of larger retail opportunity. We, I think, do see an opportunity to get into that over time. It's just probably not going to be where we start. It's not -- it's just simply kind of not one of the easiest categories to go after. I think you can kind of debate maybe the kind of near-term desire for kind of on-demand kind of immediacy there. But it is a challenging category given the kind of managing the different sizes and colors, like the kind of SKU proliferation that kind of emerges within clothing is actually a pretty tricky one. So it's something that we intend to go after over time. Certainly, something like accessories is probably an area we'd kind of get into sooner. But I guess -- maybe the point is there is so much low-hanging fruit in my mind that it's -- there's a lot for us to go after before we go after some of the stuff that's a bit more difficult.

If I can ask a follow-up, if that's all right, David?

Okay.

Would it be something that you could white label into an existing reseller website [indiscernible]?

Do you want to take that?

Yes. So it's a good question. It is certainly something that we, I think, will explore over time. I mean it's pretty day 1 in our kind of push into retail. And so don't know that, that is kind of the most immediate thing here, but certainly, it's an opportunity.

Behind Mr. Joe?

Joe Barnet-Lamb from UBS. Just one from me. When you were talking about the economics, I was sort of slightly surprised with some of the commentary around the commission rates for retail. So expect it to be broadly the same, I think you said. Could you give a little bit more color on your expectations for commission rate in retail and the degree to which margins in those industries perhaps act as impediments?

Sure. So Yes. I think what I said earlier was that the we envision the commission rates in retail being a bit favorable to grocery. And the reason I state it is that the kind of underlying margin characteristics of nonfood tend to be higher than food. I mean, food is a bit of a notoriously kind of challenging margin category. The other thing I had mentioned that I didn't mention earlier is part of the commission rate, part of the kind of logic behind the commission rate we charge is the partner's cost to serve and their kind of fulfillment economics. And the extent to which retail is oriented towards kind of lower unit per order baskets than we have on the grocery side. There's actually a bit less cost incurred from them having to have someone go and kind of walk around the store and fulfill the order. And so I think that is also something that we see as kind of a tailwind to the commission rate for retail versus grocery.

But just a clarification, which I imagine is okay. The commission rate relative to food rather than relative to grocery would be a little bit lower still?

I'm not sure that we're going to speculate on that kind of...

Maybe let me handle that.

Yes, go ahead.

I don't think we're going to give that exact breakdown. I think, Eric, really helpful to describe the different things you should consider. The one other build I would have just in terms of the mix versus grocery is I would anticipate that there will be more of a mix towards what I'm going to call independents rather than national chains, and you know and understand the kind of commission and dynamics between the two of those.

Will?

William Woods from Bernstein. So Amazon failed to build any kind of rapid delivery business with Prime Now and kind of pulled out of it a couple of years ago. Why do you think you're different to that offer?

I think I can [indiscernible] that one. Although the former Amazon guys, feel free to chip in. I think fundamentally, we just operate a very different model than Amazon. Amazon was always trying to solve the traveling salesman logistics problem, right? So how do I minimize my fuel cost for 15 drops, right? And what we do is how do we get something to you in 20 to 30 minutes? And that math problem fundamentally is just very different. And I think because of that, it's not that I don't think Amazon could have figured it out over time. But just given where they were coming from, I think it was more of a lift and this is ultimately our native experience. I don't know if you guys want to add anything?

The one thing I'd add is the same things that were a challenge to Amazon in building out the Prime Now business are some of the same things that are going to lead us to be quite cautious and disciplined about the extent to which we lean into our Hop model, our dark store grocery model for retail. We think there's an opportunity to potentially have some of the absolute kind of best sellers within that model. But fundamentally, the dark store model, having these hyper local kind of fulfillment hubs, only works when the velocity and the inventory turns is of a high enough rate. And if you look across the kind of totality of the landscape, the fastest selling items are always grocery items. And so pushing kind of into retail items and getting the kind of required inventory turns is challenging.

But I suppose -- sorry, the question was just more on the TAM, and it felt like there just wasn't the demand from customers even throughout COVID for that kind of service. You're kind of talking about demand creation in this rapid delivery space. What makes you...

I would say the Amazon proposition was quite different, right? I don't think Amazon ever launched an on-demand type of business. And I don't know enough about the details, but from -- as a user of Prime Now, the earliest you could really get was 4 hours out, you picked an hour slot. I think they were solving a different mission than maybe what we're trying to solve. So I don't view them as maybe the same type of consumer propositions. I think they're both very valuable, but they're just different.

Do you want to pass in front to Luke?

It's Luke Holbrook from Morgan Stanley. Just a couple of questions. One mainly on the U.K. and the UAE has just been launched with your retail proposition. Just be interesting to hear how you're going to scale that across the different markets towards the GBP 700 million opportunity that you see by 2028?

Sure. Yes. So we are live as of this morning in the U.K. and UAE. We absolutely see this being something that we will scale across certainly our biggest markets and ideally across all of our markets. We are starting in those two because, first, obviously, U.K. is kind of our biggest and most -- our biggest and kind of strongest market, so it makes sense to go there. And the UAE is a market where on-demand retail is actually a bit further developed. And it's something where we have confidence the consumer demand is going to be there from kind of day 1.

Okay. Understood. Just quickly on the dark store front. How does your dark store proposition in terms of your selection on offer play a role in that expansion? Are we to expect more category add-ons in your existing dark store network and the selection to increase?

Yes. As I mentioned earlier, we see Hop playing a role in retail, but probably not an outsized role. And maybe just to kind of share a bit more about Hop kind of generic because I didn't really have time to get into it in a lot of detail earlier. But we love the customer value prop that Hop delivers, and it is largely the fact that it solves the in-stock problem more than this kind of 10-, 15-minute delivery marketing message that was kind of a bit more out there a couple of years ago. The -- we took a very kind of disciplined and methodical approach to rolling out Hop, which we are very happy that we did because it allowed us to focus on the unit economics of that business, which are challenging and require you really kind of rolling up the sleeves, and we are pleased with the fact that our most mature sites are now profitable for Hop. And we think the role it's going to play for Deliveroo overall is it will be the kind of pillar of our grocery strategy for our most kind of dense urban areas, but that the store pick model is the majority of the business now and will remain the majority of the business over time.

Andrew?

It's Andrew from Barclays. Mine is just a follow-up on the chart you put up showing the kind of long-term gross profit per order in retail would be slightly higher than grocery, a bit lower than restaurants. And if you kind of think through the mix you're saying retailers are higher basket in grocery, higher commission, lower delivery fee and only slightly lower advertising. I guess if you have a bit enough bridges consumer fees. So is the intention that there aren't going to be kind of many consumer fees in retail? What am I missing in terms of getting that all to add up?

Scilla, do you want to take it?

Are you doing the contrast with grocery or are you doing the contrast with restaurants online? Both?

I would love both, to be honest with you.

So again, you -- I think you walked through the kind of building blocks sort of in the right way. So look, it's day 1. So let's face it, it's really hard to be really precise on how we expect this to evolve, but our expectation is that the AOV that's attaching to these things will be pretty strong. We've gone through some of the impacts in relation to commission rate. So if I looked at it versus grocery, probably a bit more of a skew to independents, maybe bit of put and take versus grocery in that regard. And then I think the other key thing to sort of highlight is where Camilla started. So we do think that there's an opportunity here in relation to stacking. So typically, I mean, clearly, it's nonfood it's largely not perishable. I spent quite a lot of time dealing with flower delivery in my life, but it's largely not perishable. And therefore, you get some benefits through that piece in the economics. But in terms of the delivery construct in terms of the way that we would approach MOVs, fees, all those things, no difference.

How come it's only slightly better than grocery is something I don't get, if the fees are the same and everything else is better?

Slightly better, slightly better.

Giles?

It's Giles Thorne from Jefferies. It was a question for Carlo actually to come back to your slide on the framework you have for ensuring quality restaurants. It'd be interesting to hear basically what broke down as KFC passed through that framework earlier this year and called a few headlines by coming off the platform?

Sure. So going back to that framework, essentially, the way we really approach the business is about building consumer trust for the long term. And so it comes down to, of course, offering the larger selection, but also being quite obsessed about delivering service and price integrity. So that is really what is driving essentially our commercial approach and all of the indirect levers that I presented today, whether it's the commercial architecture, which is the approach for the largest partners, or the value program for all the partners on the platform. You mentioned KFC. KFC is not on the platform in the U.K. We actually work with KFC in plenty of other countries. We went on record about a month ago on essentially the fact that we were very happy to offer lower commissions to KFC in exchange for performance along the drivers I spoke about. We -- they weren't ready to take our offer at that stage. But we think that for the long-term health of the business, doing what we are doing in this space is the right approach. And as I mentioned, the -- what I'm really optimistic is there are so many large partners that have been completely embracing this approach, they see the upside for them. Frankly, they understand fully why the category is not growing, what are the bad years, and they see this as the way to grow our joint business for the future.

And the other thing I'd just say on that, Giles, is we obviously have a program that is transparent. It's very clear to everyone what we're trying to do. And I think we just have a really long-term view because we think this program marries the interest of merchants and consumers in Deliveroo. And I think, like as Carlos said, a lot of people are really enthusiastic about it, right? But of course, there's going to be a few people that disagree and it's our job to obviously convince them over time, but it's also a long-term initiative, and it's really early.

Should we go to Marcus?

It's Marcus Diebel from JPMorgan. Maybe another question for Eric on grocery. Obviously, it's a very difficult business. You need to be successful. To have a good customer experience, you need to be integrated in supermarket's inventory management to make sure everything that the customer wants is actually physically there. You've built this sort of from scratch. You integrated it. How much of an entry to barrier is that? I mean shall we assume that the supermarkets that you have under contract today are pretty happy not to go with anyone else on the back of this? And then the second point related to this is pricing. I mean, clearly, the Coke that the customer orders is much more expensive than in the store, which is fine. There's a premium for the delivery. But you mentioned to really drive this business, you need to do something on price. And what is actually your toolkit to convince that the growth is to at least cross-promote this a little bit as well? Would they be accepting to have a loss-making order? Probably not. But really, your tools to bring the price down in this regard.

Marcus, I'll just take these two questions. And Eric, please jump in. I think your question around -- I think you're asking kind of a few different questions on the grocery side: one is how valuable are the integrations, right; and then secondly, does that prevent a grocer from being nonexclusive. I think about grocers and QSR restaurants as quite similar in the sense that their throughput is very high. So if you're a grocer, I think -- I don't see grocers in the long term working with necessarily one player. I think they will do in the short term as they sort of understand an industry and they get up to scale. But I think long term, it sort of doesn't make sense for a grocer because ultimately, they can always just pop more items on. If you're talking about a casual dining restaurant, if you're talking about an independent restaurant, I think the dynamics of the kitchen is quite different. And then just on pricing, and Eric, feel free to jump in, I think really, what we're talking about is we have a situation where we have immense demand for our product, which is delivered in 20 to 25 minutes, yet there is that markup of, call it, 10% to 12%. And when we think about the interplay between commission rates, delivery fees, product-level margins and advertising revenue and what commission the grocer can pay, ultimately, in order to drive down that markup, something has to give, right? And we think one of those interesting dynamics can be the FMCG advertising revenue. So what that looks like over time, it's hard to say. But I think our view is, if we can increase the TAM of the industry by driving down some of those markups, we think that's a really good thing, and we think that creates a much more resilient consumer base.

We'll take one more question before we break. Georgios, do you want to go?

It's Georgios, Numis. I just wanted to understand what the key bottlenecks are with bringing nonfood supply onto the platform and why this hasn't been an initiative that's been a greater focus for the business previously.

Do you want to take that one?

Sorry, the first part of your question, I'm sorry, say it again.

Sorry, what are the bottlenecks for nonfood retailers coming onto the platform. So why isn't everyone listing today? Today, you've announced Screwfix. But there's leading electrical retailers, pet food retailers that we've seen go on in the U.S. Why aren't they already on today?

Yes. That is a good question. So I think in a lot of ways, we needed to get as much kind of progress and traction with our grocery business. The success we've had with our grocery business has given us the conviction to go after retailers. I think that is a big part of the kind of why now versus some time kind of previously. And then the signals that we've been getting, some of the most important signals we've gotten have been the nonfood assortment from our existing grocers. So again, not just the kind of the framework and the approach but the actual underlying product demand in people's kind of willingness to kind of think of Deliveroo as this convenience platform as much as a food platform is just something that has strengthened over time. In terms of some of the kind of bottlenecks to growth, honestly, many of the conversations we've had with partners have been really positive. I mean, certainly, with some of the kind of independents and the local independents, there's been a bit of, "This sounds fantastic. We're ready to move quite quickly." With the larger partners, there's a bit of kind of a discussion with them and helping them. Certainly, some of the people who thought of Deliveroo at this kind of foodservice for a long time, kind of getting their head around this kind of pivot in some of the approach takes a bit of a kind of learning curve. But in general, I don't see any kind of huge structural kind of barrier.

I think Eric said it though, right? We've been at it 11.5 years. And for a really long time, the opportunity in front of us was restaurant food than casual dining, QSRs, grocery. And I think it took us a bit of -- it took a lot -- it took a bit of us by surprise when we looked at the data and saw the search terms and seeing what people wanted. And so it's very clear that consumers want it, but we were going after a lot of different things, right? And so now we've invested and we have the same team as grocery building this out. But I'd say it's a question of culture and a question of focus probably. But also, I think it took people a while, consumers a while, to just understand on demand from a food perspective, and the next logical category is some of this stuff.

Okay. Thank you. So thanks for the questions. We are going to break now for around 10 minutes. So we will come back in the room at just after 5 past 3. There is water just behind you on the left side as you walk out here, bathrooms to the left and to the right, and we will help direct you to that, but we will see you also back on the webcast at just after 5 past 3 U.K. Time. Thank you. [Break]

Okay. So welcome back for the second session in the room and welcome back on the webcast. I'm now going to hand straight over to Emma Simmonds who's going to talk to us about Plus.

Hello. Good afternoon, everyone. I'm Emma Simmonds, VP of Product, with responsibility for loyalty. I've been at Deliveroo for about 5.5 years now, and I'm delighted to get a chance to talk to you today about our subscription program, Deliveroo Plus. So first, a bit of Deliveroo history. In 2017, we were the first food delivery player globally to launch a subscription program. And 6 years on, Plus is now live in 8 of our markets. We've got 2 program tiers, global partnerships and multiple benefits. Our Plus customers spend 3x more with us than non-Plus customers, their retention is stronger, and Plus now accounts for 40% of our global order volume. Plus is at the very heart of our business and is one of our most strategic and valuable assets. And we still see plenty of opportunity for further growth. So today, I'm going to announce our important ambitions to become a Plus-first business, with the majority of our orders coming from Plus customers by 2026. But before I go into how we intend to grow the Plus program, I'd like to spend a few minutes outlining what the program is today and exactly how it works. So today, we have 2 tiers designed with different target customers in mind. Plus Gold is our core tier. That's the one we built first. It offers customers free delivery at a relatively low order threshold. So for orders over GBP 10, they receive free delivery. And that's great for all sorts of ordering opportunities, whether that be lunch or breakfast or snacks. And in the U.K., it's currently priced at GBP 7.99 a month. Plus Silver is our lower-cost tier. It has a higher minimum order threshold of GBP 25, which makes it ideal for families or for groups or couples. And Silver is the tier that we currently distribute through key partners, such as Amazon, who offer it as a free benefit to their Prime customers in the U.K., Ireland, France, Italy and the U.A.E. But Plus isn't just a free delivery program. There are a whole host of other benefits as well. And the advantage with having a tiered subscription program, as we do, is that it gives us more flexibility to build out the right proposition for the customer needs. Let's take Hong Kong as an example. So Hong Kong is one of our most popular Plus markets. Now I won't go through every single row of this table, but you can see that Gold customers get certain benefits beyond just a lower free delivery threshold that the Silver customers do not. So here in Hong Kong, we have a 5% off grocery discount because we're really focused on driving grocery adoption there. We also have a 5% off for pickup orders for Gold customers. That's actually a unique benefit to Hong Kong. And it's because in Hong Kong, the extreme density of the city means that customers walk past often hundreds of restaurants on their commute. And so pickup has proved to be really popular there. And then they also get an on-time promise where we offer them proactive compensation if their order is late. So as you can see, even though Plus is our global program, we tailor it and make it hyper locally relevant for the customer needs in each market. So now that you understand what Plus is, let's look at the economics of the program. And this is the Gold program specifically and see how it works. So I should say I'm going to walk you through the Gold program now. But actually, all tiers of our program are 12-month profit-accretive. So this chart is based on global Gold data and shows you the 12-month impact from the cohort of customers who joined Plus. As you can see, the profitability increases by approximately 1.3x. I'll go quickly through this, but the left bar is the profit before they join Plus when they're on pay as you go. We then have a cost when we offer a free trial period for some of the customers. We then experience a dilution to our unit economics, predominantly through a reduction in delivery fees and service fees and any of the other benefits that we might offer. We then gain a new revenue stream from the subscription fee that the customer starts to pay. And then the kind of most interesting part of the program is how it changes customer behavior. So here, we generate incremental orders. Customers immediately increase their ordering frequency once they sign up to the program and their retention also improves. And so as you can see, the profit after joining Plus is 1.3x larger over a 12-month period. So this chart is over 12 months. But as I mentioned, we've been running our Plus program for a really long time now, and we know that our Plus customer retention is consistently stronger than non-Plus customers. So the delta between Plus Gold and pay-as-you-go has been consistently around 20 percentage points. So speaking of other benefits, there are also wider business benefits generated from our Gold customers because they are our early adopters, if you will. So here, we're looking at a 4-week window and a look-alike group of non-Plus pay-as-you-go customers who are active in the period. The Plus Gold customers are 117% more likely to try a new-to-platform restaurant. And that's great for getting the flywheel spinning for those kind of great new restaurants that Carlo talked about us signing earlier. They're 46% more likely to order grocery, 124% more likely to order pharmaceutical products, and we make 54% more monthly ad profit. So to bring this to life a bit more, we believe that the behaviors exhibited by Plus customers make them exactly the type of consumers that advertisers will want to target, and we think they will appeal to advertisers across all of our verticals. So while we are very pleased with the impact that the Plus program has already driven for our business, we intend to grow the program much more, and we'll do it through the following routes. First, we want to enhance the savings Plus customers get. We're already finding new and innovative ways to offer value to our Plus customers. One example is our new students program in the U.K. where students are currently benefiting from free Plus. We chose to invest in students because we know they are the affluent customers of the future. The majority of the students we have on our books move on to a more profitable demographic segment for us once they graduate. And we want to entice them to our ecosystem early on through our Plus program. And it's not just us who feel this way, our partners do, too. And so we've got a great program of perks lined up with partners like Nando's and Wagamama. And we definitely see lots of potential for both us and our partners to offer more value to our Plus customers. The second route we intend to use to grow the Plus program is segmentation. We know so much about our Plus customers, and we intend to keep innovating to deliver targeted propositions that deliver on their needs. So take the U.K. as an example where here, we are the market leader with more premium customers. We know that our top customers spend around 25% more each month on Deliveroo than the top spenders on competitor platforms. These customers are overrepresented in our Plus base, and we're currently exploring how we can understand them better and make sure we build features to meet their needs, just actually like we did with our on-time promise feature which we did based on extensive research with Gold customers that found that they were looking for reassurance, that they would get great service outcomes from us. Plus customers are also much more engaged with our marketing, so we also intend to leverage some of the insights that I talked about earlier to make sure we can kind of leverage the fact that they are early adopters to help design relevant targeted campaigns to help support our new verticals. And lastly, the third route to grow the program will be through partnerships. I talked about Amazon earlier. They were our first major partner just over 2 years ago. But in the past 9 months, we've launched more partnerships, this is just a few of them: with Revolut in most of our European markets; with Gojek in Singapore; and with HKT in Hong Kong. And we intend to continue this strategy of distributing through partnerships over the coming years. So now bringing it all together, we've built a really attractive program with large customer base that results in the compelling financial performance that I talked you through a minute ago. And we intend to continue building a program that will resonate with even more customers in the years start ahead. And I think we are uniquely well placed to do this. So we have over 6 years' experience now in the subscription program. We've got strong underlying user demographics. And we have sophisticated tiering, which offers us plenty of opportunities to both customize the proposition, so what the customer gets, but also customize the commercials of the program. So to conclude, our plan is to further invest and to grow our Plus program and for us to be a Plus-first business by 2026, with the majority of our orders from Plus customers. I'll now hand over to Adam.

Thank you, Emma. Good afternoon, everybody. Nice to see you. My name is Adam. I'm our Chief Product Officer. I've been with Deliveroo almost 2.5 years. And prior to my time at Deliveroo, I've predominantly worked in branded FMCG and media sectors. I spent 13 years working in FMCG for Procter & Gamble leading some of their strategic accounts and also their U.K. and Ireland beauty care portfolio. And then after that, I spent 5 years working in the media for BBC Studios as the SVP, leading their channels and on-demand services through EMEA. And it's with great pleasure that I'm going to spend some time with you now getting under the hood everything that you see behind me here, which is our rapidly growing advertising platform for brands, something which I know we all agree has a high TAM and a huge revenue potential for Deliveroo. I wanted to start by answering the question why do we believe we're an exciting platform to advertisers. And really, it's characterized on these 3 stats. We've got 7 million monthly active users; almost 200,000 restaurant and grocery partners on the platform; and we run about 300 million orders per year. There are two good facts about those 300 million orders. The first being, look, all these orders, they have the potential to have an advert tied to them as well as the fact that we generate a commission revenue from them. And then also all these orders are for high-intent consumers who skew towards being premium. Now most importantly, we enable brands that advertise with us to talk to consumers and guide them towards a transaction on our platform all in one place. Here, you can see who our consumers are, and we over-index with these 3 groups: so number one, affluent professionals with higher disposable incomes, you can imagine a very interesting cohort that an advertiser would want to engage with; young families, larger households, typically larger baskets, again, very interesting to an advertiser; and then high potential, so I would think of these as students or graduates, their spending power today might be a little bit more in line with the average but in terms of their future spending power, again, very interesting to an advertiser at the point that they're making a lot of choices about what brands they want to set themselves up with. Our advertising journey started 2 years ago. We started enabling restaurants to advertise on the platform in late 2021, quickly followed by grocers as we've built our grocery platform. And then right now, we're in the middle of accelerating our FMCG business and our entertainment business, too. And I'll show you some examples of some of these later on. The entertainment thing is quite interesting in the sense that we're looking at adjacent consumer brands that work very well with our overall proposition. And as you heard Eric talk about earlier, we're very excited about the prospect of what retail brings to the advertising platform. And just to zone in on that, when we talk about that, we, of course, mean both retail stores who want to advertise on the platform but also retail brand manufacturers as well. I'm going to talk you through what an advertiser wants and wants from us and how we believe we can meet their needs. Now you can imagine I talk directly to a lot of advertisers and advertising agencies. But more importantly, the sheer scale of our advertiser partner base gives us minute-by-minute data and insight here, and we can use that to really optimize the product, taking both those kind of qualitative insights and the quantitative insights we get. I'd characterize our advertisers' needs in these 4 asks. So number one, they want a rich product suite that enables them to talk to consumers at a brand level with very brand-driven messaging, all the way through to how they help guide the consumer to transact and convert. They want reach, but they want quality of reach. They want relevance. So for example, if I'm a sushi restaurant, I want to be able to talk to consumers who might have a propensity to eat sushi. They want performance transparency, so I spend GBP 1, what do I get back, and they want a seamless experience that's simple for them in an ever fragmented and more complex media buying world. And then they want brand safety. So by that, they want to know that their brands are being treated well on the platform and that they're adjacent to content that they're comfortable with. And as you can see on the right of the chart, that's how we design against, that's how we meet their needs. And due to our ability to understand and segment our consumers, we can deliver a more tailored experience to both our consumers but for the benefit of our advertisers. And you heard Emma just now talk about Deliveroo Plus. And I think that's a really great example of a program that we run that we can use to deliver a more tailored, more premium experience that connects our advertisers to our most loyal consumers on the platform. Now we believe that in building an advertising platform, you do have to drive a very healthy tension between consumer experience and advertiser load, specifically how many ads you show and in what format that you show them. Now that's the right thing to do for the long term for consumers and to drive shareholder return. This will be an interesting experiment, when you're on your way home this evening and you get your phone out and you're on the train, you probably want to catch up on the news from today or catch up on your hobbies and interests. You'll probably do that on a publisher side or a social media platform. It will be interesting for you, I think, to take a look at, of those platforms, which ones are really thinking about the balance between long-term consumer experience and advertiser load correctly and which ones aren't. If you're anything like me, what I find is when that balance is wrong and someone's over-anchoring on advertiser load versus consumer experience, those platforms start to become unusable and you rarely return back to them. This is where it's hard and this is where we have to be science- and data-led. And there's an inherent tension here. But as you've heard throughout the day today from Carlo, from Eric and Camilla before me, whenever we need to make these kind of trade-offs, we always side with the consumer. So said differently, what we're not doing is chasing easy money at the expense of long-term consumer trust. And we believe that when we do this right, advertising can actually enhance the consumer experience and not get in the way of it. And that's definitely the standard that we're holding ourselves to. And as a result of that standard, we invest in machine learning to guide what we expose our consumers to and how we make it a relevant experience for them. To tie that all together, here's a look at the suite of products that we offer our advertisers with some real-life examples from across the U.K., from Hong Kong, from France. I'm not going to talk through every single one of them. But I'll give you a little guide of what we've got. So if you look at the top of the product set here, you can see examples of how advertisers can use our platform to engage with consumers in a very brand-led way. Now that might be a very physical experience, so for example, putting a physical sample of a product inside one of our bags; or it might be something that's more digital and display driven, so for example, using our social media channels or maybe advertising on our order trackers or the screen where you can track the riders' progress as they make their way to your home. And then you can see all the way down this funnel to how we help brands drive sponsored positioning on our platform, as in higher-ranking positions on the app, to drive conversion for that brand. And remember here, to make the point again, that we take an advertising revenue here, and we continue to take commission for any transaction as well. Today, we're announcing that in 2026, our advertising revenues will go above 2% of platform GTV. And we intend to deliver this by five mechanisms: number one, new advertising formats, both on our app and off our platforms, I'll give you some examples of what that looks like later; we're going to safely increase advertising load, as I've just talked about, just making sure that we grow that with consumer experience in mind; we're going to continue to increase advertiser return on spend, which we do that for two reasons, one, we know it increases loyalty of advertisers to our platform stickiness but for another thing, it increases our pricing potential as well; and we're going to increase advertiser penetration, both monthly spend that we already get from existing advertisers but also bringing new advertisers onto the platform, and we're very happy with the weekly and monthly return rates that we see from the advertisers that use us today. As you heard earlier, we're very excited about the plans that Eric walked you through earlier on retail and what they mean for advertising with us. And again, with retail, that means both retail stores advertising but also retail products as well. Right. I wanted to walk you through some advertising case studies, and I thought I'd start here. So let's walk through an example of how a restaurant or a grocer sets up an advertising campaign. So you can see the kind of platform facing for them. And you can see here, this is actually a really simple and an intuitive experience for an advertiser. With just a few clicks, they can choose a duration of an advertising campaign. They can set a budget. They can set how they want to bid in our advertising bidding auctions. We can actually do that for them as well if they want, and we can automatically bid. And in not a lot of time, advertising campaigns can then go live on the platform. This is, therefore, what a consumer sees. And you can see here some examples of sponsored positions, so as in increased prominence on the app, both on our home screen and our search returns and in our cuisine listings as well. Let me walk you through some case studies with some FMCG advertisers that we've worked with as well. Now we often talk about when using our advertising platform, advertisers can really seize a unique moment in time. And by that, what I mean is we can help an advertiser engage with the consumer about their brand in a specific moment in time and help the brand convert them there and then in one place or on our platform. So here's an example from Coca-Cola with Coke Zero. And let's just talk through the kind of moment in time that we're talking about. It's probably one that we've all done together, and I'm sure we'll do this week, which is kind of evening time, browsing through the app for dinner choices and thinking about what soft drinks might pair with the food that we pick. And you can see here that Coke have invested in a banner, so to be able to talk about their brand in a very brand-led way, but also invested in sponsored positions through our grocery platform at the same time. And this kind of contextual advertising has led to a 6x return on ad spend for Coke. And you can see a testimonial here from them about their partnership with us and working with us. This is another example of a very creative one, I think, of seizing those unique moments with consumers as well. So in this example, we've placed an advert for ITV in our order tracker. So our order tracker again, that's where you can track for the progress of the rider to your door. For people watching on the stream, ITV is the major commercial broadcaster in the U.K. And it's also a non-endemic brand, so that means it's a brand that we don't obviously sell on the platform, but it's very, very much food adjacent. And again, let's unpack that unique moment we're playing with here. You've ordered your food for the evening. You're excited about it arriving. You're tracking the progress of the rider as they make their way to you quickly. and you're deciding what TV show you're going to watch when the food arrives. And you get on the sofa and you turn the TV on. And in this example, ITV are talking to our consumers right there in that moment when they're making those choices. And again, you can see that, that kind of contextual benefit for ITV has led to that advert across the campaign being viewed by across 9 million orders. Now that's a really interesting statistic. I get a bit excited about this when I talk about it because I'd like to kind of dimensionalize that and say, okay, if I thought about traditional kind of legacy media, what would I have to invest in to get 9 million people to see an advert. And the answer is you'd have to buy an ad in the middle of the Strictly Come Dancing Live! final to be seen by 9 million people; the largest, highest viewed show in the U.K. But actually, that comparison is ever so slightly unfair on us, too, because given how high intent these users are in the sense that they're in the moment, thinking about what show they want to watch, that's a very high quality of reach there. So to actually get that quality of reach, you probably have to own legacy media by 2x, 3x that number. But we can do that all in one place on our platform, which again makes us very attractive to our advertising partners. And here's how we're doing so far. We've got 65,000 partner sites running campaigns with us. And as you can see, almost 300 global FMCGs working with us as well. We're very happy with the weekly and monthly return rates that we see from our partners, and you can see some examples below of the type of brands that we're working with as well. So I thought I'd end with giving you a sneak peek of what's to come from advertising. We're building some new advertising formats, as we talked about. Here's an example for the restaurants, which I'm really excited about, actually. This is going to enable our restaurants to have a more brand-driven message when they talk to consumers and our app with far more engaging and emotional advertising. Again, this is a good input of where we'll take both an advertising fee on top of commission revenue that we generate from sales. We're also building new ways for brands to run storefronts on the app as well. So here, this is a good example from retail where a retail brand can create a shopping experience that's tailored to them and in their own look and feel. Again, we'll be able to generate both app revenue and commission revenue here. And we're going to continue to invest in our audience segmentation and targeting capabilities as well. So I will leave you with our kind of 4 takeaways from this session. Number one, in 2026, advertising revenue will go above 2% of platform GTV. We're going to continue to leverage our technology to deliver our advertiser value proposition. At the same time, we're going to leverage our technology to make sure we protect and enhance consumer experience. And we're going to continue to invest in exciting new ad formats as well. Thank you very much for your time. I'm now going to hand over to Scilla to walk through our financial framework.

Just before Scilla does the financial section, for those of you on the webcast, if you are having a problem with the slides update, you just need to refresh your browser. I think if you navigated away during the break, then the slides haven't caught up. So if you just refresh your browser, if you're having a problem on the webcast, the slide should catch up. I will now hand over to Scilla.

Thank you. So thanks, Adam. Thanks, David. And good afternoon, everyone. Sorry, I'm just going to put my water down. So you've heard now about the growth opportunities that we have ahead of us, and I'm shortly going to look at what they mean from a financial framework. But I first wanted to take just a minute to look at the performance we have delivered and importantly, how we delivered that. It's been a combination of strong foundations, a consistent strategy and that relentless focus on execution by the team. These factors have allowed us to deliver a step change in financial performance over the last 4 to 5 years in what I've called here a shifting environment, and that's arguably a bit of an understatement. We've more than tripled GTV and we've made significant strides in margin. And as you know, we reached adjusted EBITDA profitability in the second half of last year, which was a year ahead of plan. And we've shown capital discipline, we've made tough calls to exit markets and we returned a total of GBP 300 million of structurally surface capital this year. So a lot of progress in difficult circumstances and a lot for the team to be proud of. But one of my observations since joining the business earlier this year is that we're only really getting started, and there's a level of excitement across the business about what's to come. That's in part because of those consistent strengths that we're building on. We've got strong cohort fundamentals, thanks to a highly engaged consumer base and we're addressing that large market opportunity. We have a disciplined market focus and leading positions in strong profit pools as well as the efficient logistic networks that we've built. But even more than that, the excitement comes from the opportunities ahead that we'll bring to bear over the next few years. And we touched on some of these highlights already. So that was the further step-up in CVP that you've heard about from Carlo, Camilla and Emma. It's expanding into new verticals and use cases like retail and larger basket grocery that Eric discussed. And it's the ads opportunity that Adam literally just highlighted that we're well placed to capture. But there's also things that I'm going to spend some time on today, that's the further margin levers that we have, both gross and net. So what does it all mean? It means we're confident we can accelerate GTV growth to the mid-teens in the medium term. It means we have conviction in our 4%-plus adjusted EBITDA margin target in '26. And we're on the cusp of turning free cash flow positive. And over the next few years, we expect to reach a point where we're operating and generating strong, sustainable cash flow. But let's turn first to the growth opportunity. So as you know, the fundamental attraction of the business historically has been the cohort dynamics. And on the left side here, we showed where we were in total GTV for the UKI up until 2019. And you can see that the GTV of each cohort increased each year. So let's move forward now, roll forward to 2023. And you can see that GTV from those 2019 earlier cohorts has grown about 40%. And on top of what I'm calling that like-for-like growth, we've added a whole load of new cohorts. So that was the shape. But we all know that things change quite a lot in 2020, so that slot in those intervening years. In that chart, you can see that the pandemic did indeed boost the size of new cohorts and it drove a strong increase in the GTV of existing cohorts. Since then, if you've heard us say, we've had a partial unwind of that COVID boost and we've had the cost of living pressures. Those have both been headwinds to cohort growth rather than COVID growth. But standing back since 2019, we've still had that like-for-like growth of about 40%. In the chart on the right, we've tried to simplify the moving parts here a bit and, at the same time, to answer quite a common question that we get asked, how should we be thinking about the building blocks of growth going forward? We think about GTV growth as having 4 building blocks: new customers, lapsing of existing customers from earlier cohorts, frequency and GTV per order. From when we started the business, really right up until '21, GTV grew over 50% a year and sometimes more than that. By far, the biggest driver in those years was new customers. The second driver was increasing frequency, which we saw with consistency across markets. Partially offsetting that was that lapsing of existing customers. And then finally, GTV per order was sometimes a positive, sometimes a negative, depending both on mix and underlying food inflation. Post COVID, notwithstanding the headwinds that I've briefly touched on, we've continued to add large numbers of new customers. And with inflationary pressures, as you know, GTV per order was also up more strongly in '22 and '23 than previously. So the slowdown that we've seen in GTV growth self-evidently was driven by those other two factors: a reduction in frequency and higher lapses. Of course, some of that was COVID and the unwind from COVID, but the bigger headwind in the last year or so has been the impact of those cost of living pressures. So let's briefly cover that. Even though it's clear from what customers are telling us that we have a product people want to buy, there's been an affordability issue, with food price inflation across Europe running in the mid- to high teens while wage inflation has been running in the single digits. So the chart on the left shows the dynamics, those dynamics in the U.K., but it's similar in other markets, especially in Europe. Those dynamics have had a negative impact on frequency and on retention, as I've said. But you can see from the chart that those inflationary pressures are now beginning to stabilize. And on the chart on the right, you can see that come through in our UKI cohorts, in the fact that frequency headwinds are now starting to bottom out. So there are signs that the higher rates of lapses also that we saw in the last couple of years are also starting to ease. So what do those things mean for growth going forward? Well, as we move into '24 and beyond, we do expect GTV growth to accelerate again, reaching a mid-teens growth in the medium term. I talked a few slides ago about thinking about GTV growth in those 4 building blocks: new customers, retention of existing customers, frequency and GTV per order. So our growth recipe for those building blocks has always come from a leading CVP, innovating with an existing market and existing verticals and expanding to unlock new ones. So what's different now versus the past? Well, in terms of verticals, before 2019, clearly, we really only had 1 vertical for growth, restaurants, and now we've got 3. As you've heard from Eric, we see an even further opportunity in grocery through the expansion of larger baskets and the opportunity to move into retail. Within the growth building blocks, that should, in particular, support frequency and GTV per order, but we think it will also help drive new customer additions as different missions attract different types of customers. And when it comes to CVP, historically, our strength was very much focused on leading selection, whereas now we've added much more on the other pillars as you've heard from the team today, and those two will reinforce that growth dynamic, that growth algorithm going forward. Improving value for money or price integrity and strengthening Plus will both drive improved retention and frequency, as you've heard from Emma and Carlo, as well as attract new customers to the platform who may previously have been put off by a certain perception of value. And on delivery experience, as Camilla has talked about, fewer defects, or even more perfect orders, will definitely improve retention, but it's also going to drive increased frequency. So let's take a very quick look at the top down view of growth, and I'm sure you're all familiar with the fact that we operate in large, under-penetrated markets. We started the business by opening up one of those restaurant delivery and now, just over 10 years of operating in that market, we've still delivered a restaurant GTV growth in the last 4.5 years of 40%. After that, we began to unlock on-demand grocery, and we think that can be equally large. We've taken the business from a standing start to over GBP 800 million of GTV in less than 5 years. That means Grocery is now growing to be 11% of our group GTV and has added 4 percentage points of growth to our group GTV CAGR since 2018. We started with a focus on the U.K. and Ireland, but we've recently been increasing the focus on an investment in Grocery in our international markets where penetration is even lower than in the UKI. And now, as you know, we're entering that new market online, on-demand retail, where -- which is nascent, but we think we've got a huge potential here. What's exciting about all of those markets is that online penetration is still low. And in grocery and retail on-demand penetration is even smaller, but with good reason, as Eric said, to believe that it will increase meaningfully from where we are today. So that's a top-down view. But implicitly, that relies on assuming that we can boost frequency. And a lot of what the team has spoken about today is exactly that boosting frequency through better value, through plus, through reducing defects, through expanding use cases. So what gives us the confidence that we can drive frequency in that way, and that's part of our history. The tables here show frequency by cohort for the UKI on the left and the UAE on the right. 2 markets, you remember that we highlighted previously with our '21 prelims. You can see on the left-hand side in the U.K. and Ireland that really up until '22, frequency increased very consistently across each cohort. And in fact, that frequency averaged an increase of over 10% a year across cohorts. That was true up until '22 and '23 when we saw that partial reversal due to the COVID unwinds and the cost of living pressures. And as I've touched on, the effects of those things are beginning to bottom out. The table on the right-hand side shows the same data for the UAE, where you can see a very similar trend even from a higher start point. So a clear pattern of increasing frequency, which was true across all of our markets, not just the UKI and the UAE, that pattern of increasing frequency along with all of the things that we're looking to do to drive frequency going forward is those things that make us confident that that's going to be an important building block of GTV growth going forward. Now one question that we often get asked is that come on, didn't everyone who is going to try delivery, do so join the pandemic. And I think the chart here is a pretty clear answer. It shows the new customer additions we've had historically for both the UKI and for international. In 2020 and '21, you can see that we did indeed see an uptick in new customers, people trying the platform for the first time during the pandemic. But what's really interesting is that in '22 and '23, if you look at '23 on a run rate basis, we actually added as many new customers as we did in 2020, of course, that first year of the pandemic. And we -- and in '22 and '23, we've added more new customers than we did in any year pre-pandemic. So I think it's very clearly the case that we're still able to add new customers to the platform and to do so in large numbers. Another way to look at the opportunity to drive that growth in consumers is to look at our active consumers in any one month compared to those active in the quarter or in the last 12 months, as Carlo mentioned. So on the left-hand side, you can see that we've got the opportunity to drive that frequency and convert our annual or quarterly active customers into monthly active customers. And on the right, you'll recognize the charts from Carlo's section earlier. They look at the existing customer base compared to the adult population in the market. And as we said, there's still a large part of the adult population across key markets that haven't tried delivery. Some want to try the category at all, some just want to try it with us. So a large opportunity to bring brand-new customers to the platform, but there are also those [indiscernible] customers, people who tried us once more than a year ago, and we know from customer insights that many of those people who've lapsed did so because of issues with their delivery or affordability issues. And you've heard from both Carlo and Camilla how the actions that we're taking to address those issues and how we're going to be fixing those going forward, both of which we think means that we're going to be better placed to reengage with those lapsed users. So optimistic about our ability to continue to add brand-new customers to platform and to reactivate lapsed users and drive the frequency of infrequent users. So lots of ways to think about our levers for growth, but trying to tie it all together, this is a business with a strong growth algorithm. Growth has clearly been more difficult in the last couple of years, but those macro headwinds are now starting to stabilize, and we're confident that we can drive that acceleration in GTV growth back to the mid-teens in the medium term. What gives us that confidence is the large number of initiatives that we have that are fully aligned with driving growth across our CVP improvements and our business verticals. They've always been the drivers of growth in our business, but we have more great levers now than we've had before. So if you take each of those levers individually, they will do things to contribute to our growth. But there's also the collective compounding effect from having so many levers now all coming together at once, something we've not had in the past. Switching gears, and I now want to turn to profitability where you know that we've made significant progress. That's partly due to the way that we manage our markets hyperlocally. We've always been of the view that not all GTV is created equally, and we've built strong hyperlocal positions in the largest profit pools. So as you know, those are the ones benefiting from a more affluent skew in the consumer base, a higher population density and a greater proportion of independent restaurants. Some of you may recall that back in our '21 prelims again, we illustrated the hyperlocal approach with this slide. It shows the gross profit margin by towns and cities in the U.K. and Ireland ranked from highest to lowest. And we [indiscernible] a few different towns and cities with different population and demographic characteristics. So here, I've repeated that data but now I'm going to overlay the first half of '23 data to show the progress that we've made over the last 18 months, underpinned by that hyperlocal approach. If you stand back, the gross profit margin in the UKI over that period improved from 9.1% to 11.1%. Loss-making towns and cities in the first half of this year accounted for 0.1% of orders, and that compares to 1.7% of orders in Q4 '21. And we've done all that while continuing to expand coverage, which is why the curve spreads further along the X-Axis. We've also made improvements in strong profit pools, and we've made improvements in strong profit pools, and we've made improvements in areas which have lagged slightly in terms of profit. So to give you an illustration of the towns and cities, we illustrated on that slide 18 months ago, 3 have seen a margin improvement of up to 100 basis points, 2 have increased by 100 to 200, and 3 have seen over 200 basis points of improvement. Now I do want to say that while we're very pleased with the progress that we've made in terms of gross margin, we don't have limitless room to continue to expand it in the U.K. It remains a highly competitive market. But we do believe that the way that we manage the business on a hyperlocal level allows us to more optimally balance profitability with growth. Turning next then to profitability in our International markets. And on the left side, you see the progression of adjusted EBITDA for each market within our International portfolio. The dark bars show the EBITDA by market in 2019, including the markets we've exited since and the [indiscernible] bars show the first half of this year annualized. You can see that back in 2019, all of our international markets were loss-making. If you fast forward to the first half of this year, we've improved profitability in every market, and we have no businesses making substantial losses. If you turn to the chart on the right-hand side, that zooms in on the progress that we made in our top 4 markets over the same period. And you can see clearly that we've grown GTV and improved profitability. Good progress, but more to do. The good news is we still got plenty of opportunities to drive profitability, in particular, with that reinvigorated focus on efficiency. But in order to maximize the impact that we do have on profitability, we need to look at cost clearly in the broadest way. And to us, that means looking at everything that gets deducted from our gross revenue. I'm first going to focus on the things that get deducted to drive gross margin, and then I'm going to move on to talk about marketing and overheads. So within gross margin, we have levers across revenue take rate and cost of sales. On the take rate side, in the last couple of years, we've moved to optimize consumer fees, and we've balanced that with investments in the CVP. We've also seen some adverse mix effects on commission rates, including growth in the QSR segment and in Grocery, both of which you know have lower commission rates. But more than offsetting that, being the progress that we've made in improving our cost of sales line, both as a percentage of GTV and on [indiscernible] basis. And we've made a good start on ad revenue with more to go for, as Adam said. To date, our gross margin improvement has been in large part achieved by driving efficiencies in our rider network, in particular, by reducing [indiscernible] time at restaurants. But what you've heard from Camilla is that there's much more to come on delivery efficiencies. Rider dispatch, as you know, is a complex math problem, but we've got the data and the expertise to help us address that problem algorithmically. And that I encourage you to attend the breakout session shortly with peers where you'll hear more about our data science approach here. But it's more than just the algorithms. For example, as Camilla already described, we're collecting new and automated signals on both the merchant and the rider side. We're also building incentives into our commercial architecture to reward on-time preparation. And we're getting smarter at stacking and at merchandising to reduce the time the rider spends traveling. So lots more to go for, and we're confident that, that will further support our gross margin even while we continue to make investments into our CVP. Then turning to the efficiency opportunity below gross profit. And I thought it would be helpful to share a bit more detail on what makes up the line that we call marketing and overheads. In marketing expense, we have what you'd expect, spend on [indiscernible] digital marketing and some promotional spend. The next big bucket is people expenses. So wages and salaries of staff and other people expenses, which includes things like spend on customer care agents, contractors, travel and recruitment. And the final bucket is non-people expenses, just under half of which is IT spend, so things like business apps, and in areas such as AUS compute costs and the rest is on things like professional fees, insurance and running costs of additions and Hop sites. That gives you a sense of the breadth of spend areas that we have across the business. And as I've said before, we do have room to optimize here. Now it's not to say that we haven't made any progress here recently. On the marketing side, as you know, we've made some good first steps to improve the efficiency of our spend, particularly in digital. For example, we've driven significant savings without impacting the top line through performance marketing optimization by improving or targeting and also introducing optimization signals linked to individual customer value. We've also enhanced our ML models in CRM that better predict how consumers will respond to promotions, which has driven both cost savings and incremental GTV. And on the people cost side, you're all aware of the action that we took earlier this year on removing roles and reducing the use of more expensive contractors. What we've recently kicked off an efficiency program with a company-wide focus targeting our key opportunities, the 2 big buckets of costs are clearly marketing and people-related costs. On the marketing side, I'm excited about the opportunity from things such as user level targeting, where in the future, we'll be able to leverage our data to tailor and deploy different promotional campaigns to individual customers. On other costs, we still have room on our allocation strategy, but we're also looking at upgrading our tooling and increasing automation. Although that will require some modest investments in the short term, it does drive meaningful longer-term benefits and will help ensure that we don't need to add costs as the business grows over the coming years. So bringing it all together, this slide shows our starting point of profitability in '21 when we first guided to that 4% plus adjusted EBITDA margin in '26. In the 18 months since then, we hit that important milestone of reaching adjusted EBITDA breakeven 12 months ahead of guidance, primarily driven by our progress on gross profit margin. We do see scope to see -- to move this forward further, driven in particular by delivery cost efficiencies and growth in the ads business, and that's going to more than offset the investments in the CVP that we'll continue to make. Ultimately, we expect gross margin in '26 to be somewhere in the range of 10% to 12%. As we expected, marketing and overheads is lagging behind gross margin in terms of contributing to our progress to that '26 adjusted EBITDA margin target. And in large part, that's been driven by investments that we've made since '21, in particular, in our tech team. But of course, that investment has helped us to deliver some of that gross margin improvement. But as I've just described, we began to move forward on the efficiency program and that's going to help unlock the opportunities across the business, and I'm confident about what we can deliver here. We've said before, it's not going to be a straight line from the broadly 1% margin today to the 4% plus target in '26. And that's for a few reasons. It's the macro shape, it's the time it takes to see the full effect of the growth and cost initiatives, but it's also because of our retail investment. Whilst we're building our retail business, leveraging our current capabilities and playbooks, as Eric said, I do expect an investment in retail next year in the low teens and millions of pounds. But altogether, we remain confident to deliver that margin target in '26. Turning finally then to cash flow and capital allocation. We've said many times that getting to EBITDA positive is just a milestone and that we're very much focused on getting to sustainable positive free cash flow. You may recall that in the first half, free cash flow was an outflow of GBP 28 million. So we're not there yet, but we've made significant progress and expect to be positive free cash flow next year. The cash flow dynamics of the business, as you know, are attractive. Capital requirements are limited. We typically build one centrally and deploy across countries. But on top of that, we're really disciplined in how we deploy the capital in the business, and you can see that through the measured approach we've taken to the rollout of Hop and what we've done on country exits. And then as we move to true profitability, the other important part of free cash flow is going to be the cash taxes we pay. So just a brief reminder there, because of the historical losses in the business, we had unrecognized tax losses of over GBP 1.6 billion at the half year stage. Those can be used to offset against taxable profits in the future, but do remember that the U.K. rules mean that we can only offset about half of any tax payable in any given year. So a positive outlook for cash flow generation. And I talked about capital allocation in August with the interim. So I'm just going to briefly reiterate here. Investing for future growth remains our priority. We want to preserve the financial flexibility to pursue the growth opportunities ahead, of course, with the right investment disciplines. We'll preserve appropriate headroom as any business would to deal with unforeseen events and any crystallization of liabilities. And I hope you'll agree that the extent of capital returned this year, which is broadly 1/3 of the cash at the start of the year, demonstrates our commitment to capital efficiency. So yes, with our allocation approach, a priority for growth, with appropriate headroom to sustain the business, but with a commitment to delivering this efficiently for shareholders. And as I've said, we'll continue to keep the capital structure under review. So to summarize. We continue to believe there's a large growth opportunity, which aligns perfectly with the initiatives you've heard about today. With that growth opportunity in mind, we've updated our medium-term GTV growth target to the mid-teens. We've got multiple profit levers at our disposal, which gives us confidence in that 4% plus adjusted EBITDA margin target in '26. And as we approach free cash flow breakeven, we'll continue to exhibit the capital discipline that we've shown over the last few years in order to drive sustainable free cash flow generation. And with that, I'll hand you back to Will.

Okay. So you've heard from the team about some key themes that we're focused on over the next few years and how we work day in and day out to make continual improvements. But these are the abstractions, right? These are the high-level operational and financial outputs. What does it actually mean for the consumer? How does it come together? So we wanted to share with you what the consumer experience could look like in the future. So our platform is going to become much more powerful, as Eric pointed out. What does this mean? We're going to sell a lot more products. We're going to offer a much broader choice for a wider set of occasions. But how do we make sure that this is useful for people and not overwhelming? A few things. First, we are going to make sure our platform is going to become much more personalized than it is today. And then secondly, we're going to surface that expanded choice I just talked about in new ways. We want to change the way people engage with the platform. For example, connecting with friends to discover new recommendations, hearing directly from chefs or being able to order from options that meet your dietary targets. So I'm going to bring this to life now. I'm going to show you what this could mean through the eyes of one of our consumers, Maria, [indiscernible] user. The reason I keep saying could and could is this is a direction we might take, right? The delivery of the future may or may not look exactly like this. Some of these things you'll see, we're already thinking about when we've already trialed such as in that video content. Other things are longer term. So this whole concept of enabling a consumer to follow their friend or an influencer in their recommendations and app a longer-term thing. Let's turn it over to Maria. She'll show us how she uses the app. [Presentation]

All right. So that was a snapshot of a few days in Maria's life, a hypothetical power user. So enhancing every aspect of the food delivery business and going into new verticals like retail creates a much more powerful platform as we just said, more choice. And how did Maria take advantage of all this expanded choice? Well, she is hyper-focused on her health and her finances. So the ability to integrate with her health app and her banking app takes a lot of stress. She can also explore content much more easily. She can engage with restaurant owners directly like the one from Kricket. She can see what her friends recommended -- because before she used to just WhatsApp her friends in the neighborhood and say "what's good", now she can look on the delivery app and see from other people as well. In addition, we now have a very good sense of what her patterns are. So we can surface what she actually wants and when she wants it. So for example, she's sitting at work. She wanted a quick lunch that fit her fitness goals. We did that. Friday evening, we knew she wasn't in a rush, she had time to browse, we're going to engage her with great social content. And so like I said, these are just some of the things we're either working on or thinking about longer term, which we think will significantly enhance consumer engagement. And again, we're not saying we're going to do all of these things. This is a direction, but I hope what this is, it's a fun way to demonstrate that we're getting -- just getting started in creating the most powerful, personalized and contextually relevant platform. Okay. Now I'm going to wrap up everything you've heard today. So we've got a bunch of takeaways here on the screen. A lot of exciting stuff. I've pulled out some of the key ones. I'm not going to read it. I guess I'm going to leave you with this. Why do we think we can do this? And why are we confident in our ability to deliver? Some of that is the track record. We've executed well. We've continued to grow the business despite some external headwinds. We've over-delivered on our profitability guidance. And we've done so as we constantly have improved our CVP and constantly improved the delivery experience. But there is a more inherent reason. And it goes back to what I talked about to you about 2 hours ago. It's because of who we are. So why are we confident? We are consumer-obsessed. You heard this in Eric's presentation about how we designed a unique occasion-led browsing functionality. You heard Eric talk about how we [indiscernible] through customer queries to understand where the opportunity is in retail. You heard this in Emma's presentation, she talked about how we're segmenting our Plus consumers so we can grow the program in a much more personalized way. And you heard from Camilla about our obsession with perfect orders. We are innovators. You heard about many of the things we've pioneered as well as what we're doing now. Carlo talked to you about our commercial architecture and value program that aligns Deliveroo, partner and consumer interest. And you just saw the video from Maria. So our plans for a more powerful and personalized platform. We are operators. I think this comes in loud and clear from the team. You heard about that we're leveraging this new check-in restaurant technology to drive efficiencies in the rider network. You heard about how we're using this grocery playbook and tech stack to launch retail. And you heard about some of the results. Order durations are down, defective orders are down, rider cost per kilometer are down. And we are relentless. This underpins everything. It sums up everything we do and how we do it. It's our ability to thrive and adapt to unforeseen circumstances to see opportunities where others don't. Now we have the conviction and confidence that we will deliver. I'm excited about where we are today, but I've never been more excited about the future. So thank you very much. Now we're going to do our final Q&A.

Okay. So we're going to get the speakers from the second session back on stage. And same direction as before. [Operator Instructions] Okay. So maybe we start with Lisa.

It's Lisa Yang from Goldman Sachs. The question is for Scilla. I just wanted to hear your thoughts in terms of how you see the progression in GTV growth over the next few years to get to that sort of mid-teens? I understand you're saying COVID unwind is bottoming out and inflation is also easing. So do you think in 2024, we could get to maybe a double -- next to a double-digit growth? Or is it still going to be in the high single-digit level? And also if you can clarify to get to that acceleration to meetings, how much of that is coming from basically Grocery and Retail versus the core restaurant?

That sounds like 4 questions is. So let me try. So just to be really kind of clear in what we're saying in guidance, I'm guiding to mid-teens GTV growth in the medium term. So medium term means in 3 to 5 years. And to say what I'm not guiding to then implicitly is I'm not saying that, that mid-teens is a CAGR between now and then. And nor am I saying that next year, I'm about to print kind of mid-teens. So that's what I'm kind of going to say on that, kind of a bit of shape. But I think implicit in your question is, is what gives us the confidence to move from the 5% that we were in Q3 to that mid-teens in the medium term. And I'm going to try and break that down maybe into 3 chunks. So the first bit is what we talked about in our own CVP. So the things that you heard about from Carlo, from Camilla, from Emma in terms of Plus. And there are a combination of things there. There are some things which we've already started on. We've already made good progress, so when you heard from Camilla about the progress we've already made in reducing defects. There are some things that we've started on, but just this year with Carlo in terms of those value for money pieces. And clearly, with Plus, we've been doing it for a number of years. So confident, things starting, they'll continue to land, but some of them will land kind of over time and hence, some of that shape. The second bucket is then those things which are more new levers, and they're -- the larger basket, Grocery and what we've talked about today in terms of Retail. Clearly, for both of those things, we're starting from a kind of standing start. And so you'd expect that they will contribute more in, if you like, the outer years of that shape rather than the kind of earlier years of that shape. So that's kind of CVP new things. And then the final piece then in terms of that overall shape as you rightly called out macro, and I covered it on the slide. Implicitly, to get to that kind of mid-teens growth, we are assuming that there will be some improvement in macro from where we are today. I mean we all know the numbers in the last couple of years in terms of what the impact has been on GTV growth. And again, that's part of the reason for that shape and guiding it, guiding towards those mid-teens in the medium term because I'm not anticipating that it's going to rebound really quickly in '24. So that's kind of broad shape. The second part then was in the questions on mix. So what am I assuming in terms of that mid-teens, the composition between Restaurant, between Grocery and between Retail. So as you know, sort of sitting here today, restaurants is broadly 90% -- just under 90% of the GTV of our business. So of course, the restaurant business has got to be a large contributor to that growth and trajectory. And I do anticipate that even in a kind of 5-year view, restaurants will continue to be the majority of the GTV of the business. So if I was to try to break down -- that if you like the 15% total growth, I would expect the sort of high single-digit contribution to that is coming from Restaurants. We talked today about the contribution we've already had from Grocery, so very strong growth from 0 to 800 million of GTV and that's contributed 4 points of growth in the history. And I think it's some reasonable guide, clearly bigger base, but we still see lots of opportunity to go for there. And then Retail, we've given you some sizing in terms of what we think we can get to in GTV growth by 2028.

Question, someone who didn't ask before. Let's go to the front over here.

James from [indiscernible]. There's a question for Adam. In your presentation, you gave a robust figure of 6x for ads [indiscernible], which I guess slightly speaks to both the efficacy of ads on your platform, but also the price you charge. So I was wondering sort of how does your -- how does the pricing model for your ads work? Is that 6x kind of illustrative or exceptionally good because it's a very high number? And to what extent do you plan on pulling the pricing lever going forward versus leaving a lot of value on the table for the advertisers?

So I'd say the 6x is pretty illustrative. In terms of how our pricing works, it's predominantly bid auctions for both the restaurant and FMCG and all the brands. And we -- in terms of how we measure [indiscernible] we're pretty comfortable the way we do that is in line with the industry. So there's no kind of funny in that, which sometimes you do see with certain platforms. In terms of pulling the pricing lever, look, the way I think about that is we just continuously and relentlessly just focus on building ROAS. The higher the ROAS, the higher the pricing potential, the more powerful the ad platform is to an advertiser. So that's how we focus on it, for sure.

Should we go to Andrew then?

Maybe just a follow-up on the ad question. Thank you for giving the 2% target for '26. I guess beyond that, do we think 2% is the limit? Or where do we think you can get to -- I know you guys have always been quite aware of a trade-off to the merchant when you think about pricing and commission and adds, et cetera. So how do you kind of think about that longer term?

So we're not going to guide beyond 2026, but we don't think 2% is the limit, either.

So why don't we go to William?

William Woods from Bernstein. So given -- I think this is for Scilla, given that mid-teens isn't a CAGR, how are you thinking about the marketing and overhead line in absolute pound terms, particularly, obviously, you're trying to accelerate growth while bringing down marketing as a percentage of GTV? And then how important is this kind of top line for operating leverage on your overheads effectively?

Okay. So let me start with the marketing piece. And actually, I might ask Adam to build on that, given he is sitting right next to me. So I'm not going to give you a kind of [indiscernible] target, but maybe to think about some of the levers that think we've got available. So as I described, I think we've made some kind of good first steps in improving our -- effectively our return through our performance marketing, and some more that we think that we can do there. And again, in CRM, some good first steps in terms of enhancing, targeting and the efficacy of [indiscernible] program, so more to do. But the piece, I'm probably very excited about because it's something I haven't been able to have in other worlds. And I'm sure Adam will have built here is what I was alluding to. So we're now looking effectively to build the models behind being able to make it really targeted at an individual customer level. And as I was describing, not just that I might get one promotion and Adam gets another one. It could be that I get that promotion over these particular restaurants, and Adam gets that same promotion over some other different restaurants. So that's exciting in and of itself. And then importantly, the way that we're building that kind of going to my other point on kind of broader OpEx on automation is that yes, that's being built with [indiscernible] but we're also looking to build the tooling and the automation so that the marketing team can then themselves deploy that rather than sometimes you build these things and [indiscernible] data science becomes a bit of a kind of bottleneck in that structure [indiscernible] if there's anything else that you would add.

That covers it. I think of it in 2 buckets, continually, we're going to look at I guess, our spend choices across media channels and there's always optimization to find there, and we've had a good track record of doing that to date, and we're going to just continue doing that, both using our own data science, but also using partner agencies like a lot of companies would. And then as Scilla says, this kind of moves from being able to target, I guess, at the kind of cohort level, through to being very specific and targeted at an individual level. We believe that unlocks a ton of efficiency for us and suits quite well with the -- I guess, the journey we're going to go on, on personalization that we showed you some kind of thoughts on to.

I think we -- just on Andrew's question, I think we can follow up a little bit on that ads question, which I think the way -- and please jump in. But the way I think we sort of think about it is on the FMCG side, we know what peers are doing out there. We know what Ocado is doing. We know obviously Instacart and all of those comps. So there's a pretty -- there's a large number of actually [indiscernible] all these, right? So there's actually a pretty decent number of comps out there like what's possible. I guess the way we sort of think about it though is our business obviously is a marketplace. We are in partnership with our grocery partners, right? And if there's cannibalization sort of from in-grocer promotions, obviously, people are going to care about that. Question is, if there's not, I've heard kind of different things from different people. I also think on the FMCG side, the other thing we think about and I think Eric alluded to this, to the extent that we can push down some of the markups to increase affordability, that is something we will invest in. And I think some of that could come from ad revenues. And of course, I think the other thing is what percentage of our grocery business is going to be fresh versus CPG/FMCG because they're monetizable in pretty different ways. But I think we're very excited about that FMCG opportunity. And then, of course, Adam talked about the non-endemic opportunity, the stuff we've been doing with ITV, the stuff we've been doing with Sky, which is just such a compelling proposition for these advertisers. And finally, on restaurants, the thing that we are spending a lot of time thinking about is not just sort of enhance the consumer experience by not over-monetizing ads, but also what is the trade-off if you're a restaurant between your commission rate, your ad load, your -- the discounts you offer on platforms, the discounts you offer maybe in restaurant. And sort of that interplay for us, I think, isn't crystal clear yet, right? But I will say this, if you're a restaurant group, you're spending maybe 5% to 10% of revenues on marketing, right? If you're a CPG company, you're spending, I don't know, 8% to 15%, 8% to 28%. So we are talking about very large pools of spend. But I guess 2% plus, I think is a number that by '26, we feel comfortable with. Do we think it's the ultimate limit? Probably not, but there's just a lot of interplays between all of this stuff.

All right. I'm going to circle around quickly. Your second question was basically how much are we reliant on GTV growth in order to hit the EBITDA margin target in '26, on paraphrasing, but basically [indiscernible] and so a couple of things there. So we have got a number of levers that are in our own control, both in gross margin, you've heard a lot about it from Camilla. But importantly, when we're thinking about marketing, I think we're in a much better place now in terms of optimization. And then with the overheads piece, what I was trying to describe was that we've made significant strides already. How do I think about that? I think about that in terms of us putting in place our offshore office in India that will continue to further leverage. It's about these tooling, the tooling and automation things where I was giving you that example in marketing. So not wholly reliant clearly on GTV. But if we were in a place where the macro in '26 was as difficult as it kind of is today and the consequences in terms of frequency that we've seen in GTV, I'm assuming that macro is going to get -- to be getting a bit better to the 4%.

Why don't we take questions right at the back there where we haven't had questions before so?

[indiscernible]. That was a fantastic presentation. I just wondered on a sort of 5-year view, what does the property portfolio look like? So where are the technologists going to be [indiscernible]?

Sorry, is your question where our engineers are based? Is that the question? Yes. I mean I think that we launched our India Tech Center about 18 months ago. The commitment of that team has been really, really great. We've had a lot of people go and spend time there. But I also think our team here has been really, really great, right? And I think we'll figure out sort of that ideal mix over time. I think what we're sort of balancing -- what we're balancing is, obviously, you have a team that is based in a country where we don't operate, right? And I'm a huge believer of primary research, doesn't matter what you do at the company, right? But at the same time, clearly, there are certain tasks in certain areas that can work really well in a place like this, but there's also not going to be a substitute for product people and developers actually working in areas where we operate. So I don't know what that long-term kind of split precisely is, but certainly, we've made that investment in India. We certainly see the rewards of it, but we balance that against primary research for everyone at the company.

Should we come back to the front to Joe?

It's Joe Barnet-Lamb from UBS. I'm looking at Slide 126, which is your capital allocation slide, and I'm squinting at the bar on the right, so if there's any more of that lovely light blue color, which is the structural surplus capital left within the business. So that's sort of question one. But then building off that, the headroom that you have within the business, which you say including provisions and contingent liabilities, what could change your view of how much headroom is required in the business and therefore, could bring more surplus cash?

Thanks, Joe. So I wouldn't try and squint too hard at it because it's exactly the same slide that we showed at the half year. But kind of more seriously or kind of going back through our capital allocation approach, so as you know, and you heard us describe at the interim, as we think about it kind of in 3 parts. So firstly, and kind of most importantly, it's about making sure that we've got the strategic and operational capital within the business to go after the growth opportunities that we see ahead of us. And that's both making sure that we've got cash to invest behind new opportunities, and it's also making sure that we've got sufficient cash available to continue to invest behind the very strong market positions that we've got across several markets. So that's kind of bucket number one. Bucket number two is, as you rightly call out, what I'm calling headroom. So a couple of things there. It's what any business would have in terms of headroom for kind of unforeseen circumstances. And it's also there partly for the potential crystallization of kind of [indiscernible] if that's the right way to phrase it. And then any excess we've returned, and I think we've now got a good track record of doing that. So what will change? Well, clearly, I said I'm going to continue to keep the balance sheet and capital allocation under review as we go forward. So what will change. We will continue to move forward and generate more cash as we become further on our journey of sustainable free cash flow, those things will change. And as certain things become clearer in relation to some of our kind of [indiscernible] then that may also change the position. But I think the structure and the framework is pretty clear and that will continue to apply, and we'll continue to keep the balance sheet under review.

Let's go to [indiscernible] in front of you.

[indiscernible]. So looking at the cohort charts, first cohort was 2015. So that cohort is now 8 years old. What I'm interested in is that as those consumer's age, they should have some life event things, that things will change and so on. Have you got any information about how sticky the delivery behavior is across life events like marriage, moving house, all of those things? I'm just trying to figure out exactly at what point sort of maybe that starts to tail off a little bit for some of the older cohorts, might not be now, might be 10 years, might be 20? Anything on that would be appreciated.

I mean it's a great question. I don't think we have precise information of when people get married or have kids. But what we do have is very granular socioeconomic data. So we're able to -- using Experian, we're able to box people in the 250-meter by 250-meter squares with very, very accurate demographic data. So that's something we use. In terms of what does it look like for frequency going forward? It's -- you've seen the numbers. Obviously, we grew frequency CAGR at 10% to 15% up until '19. Then it really accelerated, and we've had a bit of a step back, obviously, with cost of living. I guess a few things I'd say. One is when you look at our more affluent cohorts and usually those tend to be older a bit, maybe some of the early London cohorts, which were mostly in affluent neighborhoods. The engagement on those cohorts still remains really, really high and has decreased less than maybe some of the less affluent cohorts. So I'd say for me, the sort of slowdown is less around COVID unwind and more about affordability, right? And so let's see what happens on that. The other thing I'd say is if you look at our 2014 cohorts, which I think was all London at that point. I don't think -- maybe it's Cambridge and Brighton, but early days. The last 12 months, that frequency is something like 120x, right, maybe a little higher. So -- and those -- look, those are power users that early adopt. I'm not saying that's representative, although in the UAE, it looks like it might be. But what I would say is this, it sounds like a huge number, right? Oh, 10 times a month, 120 times a year. But if you think about how many times you eat, right, it's not that huge. And as you add on retail, as Grocery becomes a bigger part of our business, I'm quite confident we can be a bigger part of people's lives, even for those people that have joined a long time ago. But yes, we would love to know when people get married, we'd love to know when they have kids, when they send their kids to school, but I'm afraid we don't have that data.

It's Andrew Gwynn from BNP. So apologies, David, but it's [indiscernible] sort of kind of one question. But you guys are sort of the gorillas of food delivery, you built the business very much bottom up. And therefore, obviously, it's one of the advantages have been sort of relatively small is you've been able to skim off a very profitable part of the industry. As you move to wanting to become a bit more mass market, how do you retain that DNA? Can you retain that DNA and retain, I suppose, some of this sort of very profitable volume that you do? And then the cheeky Part B, and apologies for it, but I'll ask it. Are you happy remaining an independent business?

Did you say we're a gorilla but we're small, so we are like a small gorilla? Are we -- I wasn't sure what you meant by that, but I think I know what your question is.

We're also thinking about Guerilla Warfare. Apologies to confuse you.

Sorry. Yes. So let me just start by saying we are a mainstream product, right? Our areas of high market share and dominance, I think, are clearly areas that have higher population density and affluence and independent merchant mix. But that is an intentional strategy where our belief is when you take the U.K., you've got 3 players that are kind of getting to the same GTV market share. But when we look at our -- the areas that we're really strong in, take London, the Southeast suburbs, the Southwest, Manchester, places like that, our belief is that we'll drive a disproportionate amount of free cash flow as a percentage of GTV from those places. And ultimately, it is about free cash flow per share. That's what we care about. Now I think in terms of making progress in less affluent areas, I think we've done a bunch of those things in certain areas. I look at the progress we've made in [indiscernible]. I've looked at some of the progress that we've made in some of the outer parts of Manchester. So we can absolutely do that. But I think for us, how do we generate more cash over time and be efficient, I think, a focus on an affluent consumer base has paid off for us. But this is not some sort of rich person's product. This is a highly mainstream product offering, right? You can get all kinds of things from expensive sushi restaurant, to McDonald's, to whatever. Now increasingly, you can get all different types of things, right? So we think we play a big part in many people's lives, and we're going to play an increasingly bigger part in people's lives, but also I do believe that this is a very mainstream product. Your question about, I guess, how do we feel about this being an independent business? Well, I'd say great. First of all, right? We just came up here. I think you guys have been here for 4 hours, so thank you for listening to us, but we're incredibly proud of what we've built here. We're incredibly optimistic about the future. You've heard the plans for all of this. And there's no question that this is a competitive market. But what are we focused on, right? We're focusing on executing every single neighborhood and building that CVP the best we can. At the end of the day, this business -- oh, and by the way, your cheeky question is a question people have been asking me for 5 or 6 years. It's not really a new question, that's fine. I think the way I sort of think about it, though, is what is the industrial logic for a combination. There are no network effects that are national or international in this business. These are neighborhood businesses, right? And I think any time you have a sort of neighborhood combination, there might be some antitrust concerns as well, right? And so I think we're a public listed company. You've got our chair over there as well. If people decide that we're interesting in that way or vice versa, that is something we have to consider as a Board. But overall, we are incredibly enthusiastic about the future. This team that is sitting up here with me, like I said, is the best team I've worked with. And I just feel that we can drive a better experience than other people over time. That's all that matters.

Okay. Georgios?

Mid-teens growth. We've broken it segmented in many ways. Can the U.K. grow at mid-teens? And then in International, there was front in Italy data, which is underpenetrated, but International is currently weighing on group growth. So just kind of, has anything changed recently? You've clearly been doing a lot of work and analysis on all of the different divisions, the profit pools, new verticals? How does that change the growth confidence for the International segment?

I'm presuming I'll take that one. So stand back. So obviously, we think that we can grow in that level in both of our reported geographies is the kind of quick answer to it. Kind of why do we think that? It really goes back to everything that I was sort of really answering in relation to kind of Lisa's question, it's all of those levers that we now have at our disposal in terms of CVP, in terms of use cases in terms of verticals. So we think that they will apply in the UKI as much as they will apply to kind of other markets. So yes, to the UKI. Secondly, on International. Well, I mean to state the obvious, International is a kind of bunch of 8 countries that to grow at different rates. And yes, at an aggregate level, it's been growing less rapidly than the UKI recently. But actually, if you look at Q3, for example, Italy and the UAE both grew at a faster rate than the UKI. So there's always a kind of mix impact of that. What's been a little bit of a headwind on growth in international recently, as you all know, has been -- the headwind there has been what's been happening in France, which, again, as you've heard us say, is a kind of market-wide point rather than something specifically to delivery, which is all linked to consumer confidence. So we believe that that's a temporal thing. And if you stand back and you look at our international portfolio, deliberate choices to operate in attractive profit pools with some of those characteristics that I've described in terms of what makes the hyperlocal pieces attractive and you are spot on, Georgios. So in addition to those local characteristics, if you stand back and you look at it on a market basis, there are lots of very underpenetrated markets still there to go for. So mid-teens growth, UKI and International.

Okay. I think we have time for one last question. So I'm going to Monique, you've had your hand up for a while, so I'm going to go to Monique.

Thanks for taking the question. It's a question, I guess, mostly for Scilla. Slide 120 of the pack, you've got the markets by that EBITDA. And the chart on the right, you've got the sort of GTV versus the EBITDA and how it's grown over the past 4 to 5 years. I guess what I'm really interested in is there's a country there. I'm assuming it's UAE, maybe I'm wrong, where you've got an EBITDA level that is I guess, the same as the U.K. despite about half the GTV growth. We're looking at it the other way, you've got 3 of the top 4 markets that have got very similar GTV and one's got pretty astronomical EBITDA. So just wondering, are a lot of the factors that got that market such high EBITDA locally specific? Or are there lots of things that you can do with those other 2 markets to drive an EBITDA closer to that level?

Thanks, Monique. I was wondering if I was going to be asked [indiscernible] the country. So I'm not going to say which country is which. But there are -- what I would say is that we apply the same hyperlocal approach across all markets that we operate within. And one of the lovely things about the business is that we do have that approach sort of building wants and then rolling that out kind of internationally. So we're very consistent. So the things which have driven some of that growth are very consistent. It's about all the things you've heard me talk about in terms of CVP and so on. Obviously, different markets have slightly different characteristics, particularly in respect of frequency and respective drop density. So some of those things will have skewed those numbers. That's probably all I'm going to say.

Just to be clear, that slide is just for the International business. So UKI is not on that slide. Okay. So I think we're unfortunately out of time for further Q&A now, given that we need to move to the breakout sessions. But I'm going to just wrap up this part by saying thank you to everybody for joining us in person and for joining on the webcast. [indiscernible] I'd also just like to thank everybody in the delivery team who've been demonstrating the, I suppose, the characteristic of relentlessness in putting this all together. It is a big lift on top of everybody's regular day job. So thank you to everybody in the team. And we'd just like to echo our appreciation to everybody for joining us and for the effort of putting this together. Okay. So we're going to close the webcast.