Deliveroo plc (ROO) Earnings Call Transcript & Summary

Hello, everyone. Good morning. Thank you for joining. I'm Will Shu, I'm the Founder and CEO of Deliveroo.

And I'm Adam Miller, I'm the CFO here at Deliveroo.

It's great to have you all on the line for this morning's presentation of our half year 2021 results. We look forward to talking through the results in detail. But first, let's talk about what we're going to discuss. So I'm going to start with a summary of the key business highlights from the first half, and Adam and I will walk through the H1 financial results in more detail. After that, I'll go into what we're seeing in terms of initial consumer trends as lockdown restrictions begin to ease across some of our markets. We'll then deep dive into grocery. Specifically, we'll share more about how we progress in the first half but also provide further color on the economics of on-demand grocery. Then we'll provide a snapshot of our rider proposition from the first half. And finally, Adam will provide a reminder, of our full year guidance. So let's get started. At a high level, we've reported strong performance in the first half of the year, and we've continued to make good progress in executing our strategy. In terms of key takeaways from these results, I want to highlight 4. First, consumer engagement post-reopening has been encouraging with H1 orders and average order values trending above expectations. We're continuing to see lockdowns restriction ease, but our markets are experiencing different stages of restrictions and reopenings. So we'll spend some time this morning talking through what we've initially observed from different markets reopening. Second, we've gained further traction in on-demand grocery. We've continued our rollout with leading brands such as Carrefour, Waitrose, Co-op, Morrisons and Sainsbury's, we've added many more as well. In addition, we'll go into more detail on grocery unit economics. We continue to be very excited about both the incrementality and the synergies with our restaurant channel. Alongside this, we are proud of our execution on the food merchant side. Deliveroo has the most food merchants of any food delivery platform in the U.K. We increased our U.K. restaurant selection by just under 30% in the second quarter from the first quarter. Third, we have also continued to see strong rider retention and satisfaction. We have 85% satisfaction reported across our global rider fleet. What I think is most interesting though is that applications remain at similar levels as we have experienced throughout COVID, and the retention of riders has been extremely strong post-reopening, despite significant increases in job vacancies, particularly in retail and hospitality. And then fourth, we continue to pursue our strategy of investing in long-term growth. These investments we make are meant to drive durable long-term differentiated improvements for our consumer value proposition. We're still very early in the maturity of the online food industry, and we will continue to invest in driving new consumers to the category and our platform. So what's been the output of this in the first half? We are operating a complex 3-sided on-demand marketplace. And we think of each side, our riders, our restaurants, grocers and end consumers as customers of Deliveroo. We need to balance the interest of all 3 sides and Deliveroo, but ultimately, the strength of our performance is driven by how well our proposition is working for each group. Throughout the first half of this year, we've seen strong engagement from each side of the marketplace. We've grown our consumer base 7.8 million average monthly active consumers in Q2. We've achieved 85% rider satisfaction across our global network of over 100,000 riders. And despite reopenings, we've continued to see very strong retention amongst existing riders and application rates from new riders. We've also added great restaurant selection. We have over 136,000 partner sites now live on the platform globally, and we've also rapidly expanded grocery with over 9,000 partner sites now live. And what does this add up to? We've experienced strong group financial performance in H1. We've seen 100% order growth in H1. Gross transaction value in the first half was up 102% to GBP 3.4 billion, showing continued strong momentum despite reopening effects at an increasingly tough comparison base. Gross profit is up 75% to GBP 264 million. Gross margin as a percentage of GTV was 7.8%, down from 8.8% in H1 in 2020. This is due to an increase in growth investments including acquisition of new customers plus consumers, and investments in differentiated content. Adjusted EBITDA was a loss of negative GBP 27 million compared to negative GBP 30 million in H1 2020. Leverage of overheads from increasing scale was largely offset by increased investments to support future growth. We'll go into more of these details later on. And just for clarity, all of the growth rates that we're reporting in this presentation this morning are in constant currency. So overall, what we've seen in the first half has been a strong engagement from each side of our marketplace, leading in turn to strong financial performance and continued strong momentum despite initial reopening effects. All right. Let's talk about these results in more detail. So overall, we're pleased with these results, which have been driven in large part by the larger consumer base that we've built through both attracting new consumers and continuing to build strong engagement from existing consumers over time. This has resulted in continued strong growth both year-on-year and sequentially despite the initial easing of lockdown restrictions across many markets. I'd like to thank the Deliveroo team for their continued hard work, relentlessness and excellent execution. Just a quick note on this page. Order growth has outpaced GTV in the second quarter for the first time in a long time. And the reason for this is AOVs have receded slightly versus last year's peak, although they remain well above pre-COVID levels. Let's look at consumer engagement. This comprises monthly active consumers on the form and monthly order frequency. There are 2 major things to call out on this slide. Number one, our biggest driver of growth continues to be the increase in monthly active consumers on the platform. Q2 was no exception to that. We grew MACs year-on-year 81%. And as of Q2, we have more than doubled the number of MACs on the platform since the beginning of 2020. The second thing to call out is that you can see frequency has remained flat, essentially since the beginning of COVID. However, the overall frequency of 3.3 is the blended average of frequency across all consumer cohorts, both old and new. And as we've shared before, consumers become more engaged with the platform over time. So the more mature the cohort, the higher the monthly order frequency. So it's very encouraging that average monthly order frequency has remained flat whilst we've added so many new consumers that are just getting started on their journey with Deliveroo. In summary, at a group level when compared to Q2 2020, we grew MACs year-on-year 81% and maintain frequency. Now let's look at the U.K. and Ireland. When we look at the U.K. and Ireland performance in H1, we're proud of the results this quarter, both year-on-year and sequentially. I think there's 2 major takeaways. First one, as the U.K. has reopened, and as a reminder, the U.K. experienced 3 stages of restrictions easing between April and July this year, we have not seen any material impact on consumer engagement. We're going to go into more details about this later on. And secondly, we are pleased with the overall growth, but also the broad based geographic momentum. We drove consistent growth throughout all of our geographic regions. London, the London suburbs, other major U.K. cities, other parts of the U.K., Ireland during this period. Our market position in London continues to be extremely strong and remains unaffected by ongoing competitive intensity. Across the U.K. and Ireland, we are now live with major grocery partners in 1,800 sites. That's up from 1,200 sites in December. These are partners such as Aldi, Co-op, Morrisons, Sainsbury's and Waitrose. And now we have more food merchants available than any food delivery platform in the U.K. In particular, I'm very proud of the business' ability to execute during a difficult time working remotely. The team has added over 10,000 new restaurants in Q2 alone, extending our restaurant base by almost 30% in 1 quarter. Ultimately, I'm very proud of the incredible content we have driven in the second quarter. Alongside the strong growth in restaurant and grocery selection, we've also expanded very fast geographically. At the beginning of the year, we aim to increase population coverage from 53% to 67% by the end of the year. But in H1, we've already gotten to 72% population coverage. Also, though, I think it's important to note that in H1, the vast, vast, vast majority of our growth came from existing locations rather than expansion. So this expansion makes us very excited for years to come as we think about planting the seeds for future growth. So we've had strong momentum in the U.K. and Ireland in H1. Now let's move on to international. All right. So in the International segment, we've seen strong growth in Q2 and for the half as a whole. I want to give you some further context here. Growth in our international segment reflects a number of different lockdown and reopening scenarios. Going to deep dive into examples from 4 markets later in the presentation. But let me give you a high level of the dynamics. Let's start with Asia and the Middle East. These markets went into lockdown first and came out first. We've seen very limited impact of reopenings in our business there. In Europe, vis–à–vis the U.K., Asia and the Middle East, the European restrictions were introduced later. They've been eased most recently as well. And we've seen a more pronounced impact from reopenings versus the U.K. The easing, however, has coincided with summer seasonality, which we see each year. And so we'll have a clear read on the impact of restrictions easing in our European markets as we get towards the end of Q3. And in some markets such as Australia and Singapore are now back in lockdown, and we're seeing continued very strong growth. So we'll talk about this in more granular detail in a bit, but you can see that the international segment is characterized by a number of different lockdown and easing scenarios. A few other points to call out in terms of international performance is the strength in selection in both restaurant and grocery. We added 10,000 new restaurant partners in Q2. We continued our rollout of major grocery partners. We've more than doubled our sites with major partners across the International segment. We further strengthened our consumer value prop because of that. Now I'm going to hand this over to Adam, who's going to talk about our financial performance.

Thanks, Will. We'll now move on to our H1 gross profit and bottom line results. But I want to start by setting some context first on how we think about investments and where we make them in our P&L. First, our philosophy on investment is to invest through the lens of our consumer value proposition, the 5 pillars of availability, selection, experience, value and brand. We understand which of these matter most to our consumers in any given market where we stand relative to competition, and we invest to build the best consumer value proposition in the market on a hyper local neighborhood by neighborhood basis. Importantly, we also invest for the long term. Our priority is delivering durable long-term differentiation. We're still extremely early in the maturity of the online food industry, and we're laying the foundations now to drive long-term value for our shareholders. That means that the primary impact of many of our investments will be mainly in future quarters rather than in-period. We're talking, for example, about investment in selection to create differentiated content for our consumers; expansion into new geographic areas; in brand building to drive up awareness or in new consumers that will significantly increase their engagement over a succession of future periods. Second, I also want to share some context on the timing of how we've invested in the business. We had a relatively slow investment pace through most of 2020. We face capital constraints for much of the year due to the CMA antitrust investigation, and long-term consumer behavior was uncertain during the initial lockdowns back in early 2020. As such, we took a very conservative approach to how we deployed our capital during that period of time. Starting in Q4 2020, we've been much better positioned to increase our investments in the category. We have a more secured capital position now due to the conclusion of the CMA investigation in Q3 2020, our Series H raise in January and the IPO at the end of March. We also have very strong conviction that the pandemic has accelerated the secular shift in consumer behavior, moving demand and food online. Third, I want to provide some context on where different types of investments hit our P&L. This may be obvious for some, but I think it's worth providing some examples before we look at the results. Starting with above gross profit, you primarily have 2 types of investments: commission and consumer fees. Investments in commission enable us to increase and differentiate our restaurant and grocery selection. This is one of the most important areas of the consumer proposition to create and maintain differentiated content. This includes, for example, strategic restaurant partnerships or the continued scaling of our grocery business. Consumer fee investments primarily come in 2 forms: first, 14-day free delivery to attract new consumers to the platform, which is an offset against delivery fees; and our Plus subscription program, which is dilutive to our unit economics on a per order basis but accretive to consumer lifetime value and to aggregate gross profit due to increased engagement on the platform. Now if you look between gross profit and EBITDA, you have 2 main areas of investment. So our philosophy is once we have a consumer value proposition that we think is differentiated, we then want to shout about it and investments in marketing, both offline such as TV media and other above-the-line spend, and also online or digital marketing, so performance marketing, app downloads, help us do just that. These types of investments increase our brand awareness with consumers, which helps to drive engagement and retention with our existing users and also drives new consumers to the category to try the platform. Investments in our overheads also drive the consumer value proposition. The primary area we've called out in the past is ramping investment in our engineering team as this drives a ton of leverage over time, but this also includes things like investments in our sales force to attract and work closely with restaurant and grocery partners to maintain and improve our differentiated content. Finally, you have CapEx and CapDev below EBITDA. CapEx almost entirely relates to our investments rolling out new Editions kitchens, which provide differentiated content with fantastic brands that you can't find anywhere else and the best delivery experience that you can get. And capitalized development relates to platform development by our engineering team, which we've already talked about. We wanted to give you this color on how we invest for the long term and also where things sit in our P&L, so you can see that our philosophy is not in driving a femoral value via short term discounts. Let me now move on and discuss more of the financial results. So starting with gross profit. The strong operating performance of our business in the first half that Will just took us through has also resulted in strong gross profit growth with gross profit up 75% year-on-year in the first half of 2021. As I just mentioned, starting in Q4 2020, we accelerated the pace of our investments in many of the areas above gross profit, which is driving some of the growth in the half, but more importantly, is planning the seeds for future growth. This includes incentivization for new consumers to join the platform via our news or experience program, incentivizing consumers to move to our Plus program and continuing to build and maintain differentiated content. Now let's move on to marketing and overheads, and adjusted EBITDA. As this slide shows, we've stepped up our investments in marketing and overheads in the first half of 2021 to capitalize on the momentum in the business. This is mostly about supporting future growth as we've invested in building out our engineering teams and invested in building brand awareness on the marketing side. It's important to note again that we also had a slower pace of investment for the first 3 quarters of 2020. This was due to both the uncertainty around the impact of COVID as well as the ongoing CMA antitrust investigation that resulted in us needing to exercise a high level of conservatism around our deployment of capital until Q4 2020. We're now in a very strong position to continue to invest to capitalize on the positive momentum. And finally, let me close by talking about our balance sheet and liquidity position. Following the IPO, we now have a very healthy balance sheet and liquidity and are well capitalized to go after the opportunity in front of us. In addition to our P&L performance, we also saw a positive impact from net working capital. However, approximately GBP 130 million of the GBP 184 million you see here in the first half is related to the timing of employee tax and social security payments related to the IPO, which will unwind in the second half.

So as we've discussed, given our geographic footprint, markets have experienced different timings of lockdowns and easings. And although it's too early to make definitive statements on post-lockdown behavior in general, we wanted to provide some more color as to what we're seeing on a market-by-market level. So we're in a deep dive into the U.K., France and Italy, the UAE and Hong Kong. So let's start with the U.K. So what we've seen in the U.K. in terms of easing lockdowns, there were really 3 stages of restrictions lifting. First in mid-April, the nonessential retail and outdoor dining reopened. You can see that as the first line on the left there. And in mid-May, restaurants reopened for indoor dining with social distancing. That's the second line. And then in mid-July, all social distancing restrictions were lifted. And as a result of these lockdowns being lifted, we have seen no material impact on consumer engagement at all. Order volume has remained robust throughout all 3 reopenings. AOV has declined slightly as people move towards smaller party sizes, but it still remains above 2019 levels. We do expect any AOV benefit from larger party sizes that we've seen during COVID to trend towards pre-COVID levels over time. Okay. I thought it would be interesting to really dive deeper into the cohorts themselves. So we'll talk about the underlying consumer behavior. We showed you monthly average frequency before. But as I said, that is a combination of new and existing user behavior. And so cohorts give you a lot more color. And what we've seen is that recent cohort behavior continues to be consistent with pre-COVID cohorts. In other words, as cohorts mature, the existing consumers transact more and more. And that's been the case ever since we started the company. So on this chart, what you see here is we have laid out 3 different cohorts. You have users that we've acquired in Jan '16, users that we've acquired in Jan '18 and then users we've acquired in Jan '20 and we wanted to compare the frequency across these. So if you look here between June '19 and June '20, so this is relevant for the 2 left most cohorts. As COVID took off, you can see that the frequency increased. Yet, if you look at June '21, you can see that frequency has still continued to increase even as restrictions have eased. And this frequency also applies to cohorts that we've acquired during COVID. So that's the bar on the right, the Jan '20 cohort. So yes, I think this is really encouraging because what we're actually seeing is, obviously, you see a booster in COVID of frequency, but then you still see the frequency improving even post-restrictions easing. So what we're seeing at this stage of reopening is in the U.K., we're not seeing any material impact on consumer engagement. Now let's go to some of the European markets. So France and Italy were about 1 month behind the U.K. in easing lockdown restrictions. So what we've seen in France is a moderate decline in orders through reopening. But over the past 18 months, France has seen a very significant amount of growth. So we entered lockdown reopening from quite an elevated position. And when outdoor dining reopened in May, we saw that -- and more so than in other markets, we saw that consumers flock back to restaurants. And once indoor reopening happened in June, indoor dining reopening happened in June, we remained flattish. I'd say in addition, there's been a number of different national and regional curfews that did apply both to indoor dining as well as food delivery. And this restricted operating hours pretty significantly alongside the reopening of restaurant outdoor dining in May. And then from late June to July, now we're getting the stage of lapping our typical summer seasonality in France. And in our European markets, we typically see this happen around the holiday season. So -- and this has been every single year since we've been in these markets. So I think for us, it's a little difficult to disaggregate the reopening impact from the seasonality, the summer seasonality completely. We expect the picture to become clearer towards the end of Q3 as normal summer seasonality ends, and that will give us a better picture of how consumer behavior begins to settle. But the initial trend in France suggest a moderate decline due to factors that we think apply industry-wide in France. Now let's go on to Italy. So in Italy, we've seen reopenings having some slowdown in order growth, but to a lesser extent than France. And you have the same sort of seasonal factors as well in Italy. And taken together, you can see that the initial trends for Europe is that you had lockdown restrictions being lifted later than the U.K., and these have coincided with our normal summer holiday seasonality period in Europe, but it is difficult to disaggregate the 2 at this point. Again, at the end of Q3, once the European summer holiday is passed, we'll have more information, but we think that the impact has been moderate in France and small in Italy, which is encouraging, but we hesitate to make any firm conclusions yet about the reopening impact in Europe. All right. Now let's talk about Asia and the Middle East. So compared to the European examples, we've seen a very different situation play out in the UAE and Hong Kong for many reasons. First of all, restaurants began reopening in October 2020, so much, much, much earlier than in Europe, and there's also been a number of events that have happened in both markets. We've highlighted them on the chart. I'm not going to go into all the details, but take a read some of them are important. And so I guess one thing to call out is seasonality in these markets are just very, very different than Europe, right? You -- Ramadan and Eid in the UAE, you've Chinese New Year in Hong Kong, these are very prominent points in the calendar where you can see order volume shift. But the key takeaway is this. These markets have been reopened for a long period of time. Yes, you might have a localized lockdown here and there, but overall, pretty open. What we've seen in the UAE is we've seen a strong acceleration in growth post-lockdown. And you can kind of see the market's been growing despite all of the restrictions and just continues to grow. You've a few dips there when it comes to the Ramadan and Eid holidays, but those happen every single year. And then in Hong Kong, we're seeing positive growth through the different stages of reopening as well. So look, if I was to summarize everything that we've learned through these 5 markets, looking at the U.K., France, Italy, UAE and Hong Kong, you see a number of very interesting and somewhat different scenarios. So how to summarize. First, no material impact in the U.K., and we've had now been through 3 stages of reopening. We've seen a small to moderate initial impact experience in Europe, but to different extents and combined with the summer seasonality, it's a little difficult to disaggregate the 2 because that is when demand typically drops in Europe. Then we've seen accelerated growth in the Middle East and Asia. These are the markets which have been out of lockdown for the longest period, which I think is very encouraging. And then we have strong growth continuing in the markets that are still in lockdown. So that's Australia and Singapore. So for Europe, I think we'll have a clear idea of the overall picture in Q3, but we are encouraged about what we've seen from the markets where we have the most data points. All right. I want to take some time to discuss the grocery business in some more detail. Just as a reminder for everyone, we operate an on-demand grocery model. Our consumers get products in about 25 minutes. We partner with major and independent grocery brands. So those would be Aldi, Carrefour to your local green grocer. They pick-and-pack from their stores offering about 1,500 to 3,000 SKUs and the consumer and rider base for grocery and restaurant are the same. And that has driven very positive effects to the delivery flywheel. All right. So in the first half, we continue to rapidly develop our on-demand grocery offering. We first began offering this in late '19. And since then, we've seen rapid growth in what we call ODG. This has been driven by COVID factors initially. But as you can see, the growth has been very, very sustained through the first half of '21. So on-demand grocery now represents 7% of total GTV. And what's been interesting, I've said this a number of times before, more than half of people don't want to preplan food in advance. And before we launched ODG, the options available typically were the fastest for next day, right? And so we do think this represents just a completely different use case in a completely different set of consumers. And so on-demand grocery is uniquely positioned to fulfill this, I'd call it more spontaneous consumer demand. So we've continued to roll out this ODG proposition, this grocery proposition. So in the U.K., we continue to roll out with Waitrose, with Morrisons, with Co-op, with Aldi, with Sainsbury's. In the International segment, we continue to roll out with key partners and new partners, Carrefour in France, in Italy, in Belgium, Casino in France, Conad in Italy, Park N Shop in Hong Kong, 7-Eleven in the UAE as a few examples. And if we look at the U.K., I'm really proud of the team here again for the rapid expansion of our sites. So we now have over 1,800 sites live with what we consider major partners. And so we've extended our grocery population coverage with these major brands to about 70%. So that's up from 36% in the first quarter. So it's been a quite a rapid pace of expansion. All right. We're now going to talk a little bit about our longer term view on on-demand grocery and why we think it's such an appealing model. So on-demand grocery for us provides, we think, very powerful synergies to our core restaurant marketplace. Why? Well, it represents 100% incremental demand to our restaurant channel, and it's very, very helpful in driving new customer acquisition as consumers can find the brands they know and love now on Deliveroo. So overall, it drives this flywheel much faster. Why is that? Well, first, almost all grocery consumers start transacting on the restaurant marketplace eventually. And if you think about restaurant consumers, once they start purchasing groceries their overall frequency goes up significantly. And the output of this is we end up increasing our network density much higher on a hyper local basis, where, obviously, we have a lot of grocery partners and restaurant partners. And this allows us to operate more efficiently, one. But secondly, it also provides more work for riders. And grocery is a less peaky demand business than restaurants, and you can spread that overall demand throughout the day. It's also an attractive economic proposition for Deliveroo. In the U.K. and Ireland, gross profit per order for grocery stands at approximately GBP 2.10 compared to GBP 2.40 for our restaurant business. And as we've described before, the basket size is slightly larger, the take rate is a bit lower than restaurants, our consumer fees are a little bit higher. So today, net-net, there's an approximate 30p difference in unit economics. But we also think there's further building blocks available to us over the long term in regards to economics. These are things that we haven't done yet, but we think may be significant in the future. So examples of this: improving selection and inventory management to improve the overall consumer experience. Improving that inventory management is, in our view, a precursor to be able to add more SKUs and move into medium-sized baskets ultimately driving higher AOVs. We can also continue to realize efficiencies through our rider network. We believe there is still room to run here via things like increased stacking, which is theoretically more appropriate for things that are not immediately perishable such as grocery. And then we can also continue to improve in our operational efficiency via just improvements in our logistics algorithms. We also see the potential to generate non-commission revenue. So I think an example of that is when you look at what's happening now with CPG companies and FMCG companies is more and more of their sales or online that creates an excellent opportunity to target consumers in a very different way. It allows people to understand the results of their marketing spend much quicker. And as more and more sales go online, we expect those ad revenues to follow. And you're seeing that with Instacart and Amazon and companies like that, where this type of marketing dollar has become a material part of their revenues. All right. I'm going to give an overview and snapshot of the rider proposition in the first half now. All right. So I wanted to first just outline again in detail what our rider proposition is and really highlight the true 2-way flexibility of this. So flexibility is the #1 thing that riders tell us they value the most when working with Deliveroo. The key, though, is true 2-way flexibility. What does that actually mean? It means first, you can choose when and where you want to work. You can work with multiple apps at any time. You have unlimited freedom to reject any order. You get paid per delivery complete. You get full visibility of all your fees for each delivery upfront, which means you can reject them, as we said. And we also give you a clear picture of what the consumer demand looks like in each neighborhood. And so you don't have to preplan anything. You can show up whenever you want. And so we've been focusing on that proposition that delivers ultimately on the things that riders care about because ultimately, they are customers on our platform, and it is a competitive industry. And this is ultimately what riders want. And what we've seen in the first half is that we've seen 85% global rider satisfaction. We've continued to see high levels of demand for this work, over 14,000 applications per week in the U.K. And it's also worth us not just looking at kind of the proposition abstract, but let's actually look at what's happened in the U.K. during COVID with riders. And actually, sort of as the economies reopen, I think it's really interesting. This half, we've seen a clear continuation of the evidence of the popularity of the work that we offer. So if you look on the chart here, you've seen that the U.K. economy has reopened. You see this massive growth in available employment vacancies. You see job vacancies were estimated to grow 40% quarter-on-quarter, huge spikes in retail and hospitality. So there are a lot of jobs available out there, yet you can see that Deliveroo's monthly active U.K. riders was up 12%. And so our rider supply has not been impacted despite all of these vacancies being out there. So we see this continued high demand on both the application side, and we see a very, very healthy retention rate for the self-employed work that we offer. And let me just go into detail on retention as well. 90% of the riders working with us in May or June were also working in July. And obviously, this is as employment vacancies have surged. So what -- I think the picture that I'm trying to paint here and I think that is happening is during COVID, we certainly had a lot of riders join us. And I think what's happened is that they're staying with us. Well, clearly, they're staying with us. Clearly, they're telling their friends about it and get more and more applications. When there are a lot of alternatives out there, right? And you see that in the retail space. You see that in the hospitality space. You see all of these newspaper articles and saying restaurants can't staff people. We're not having that problem. And I think this is evidence that our proposition is really, really appealing. That this flexible nature of riding with Deliveroo is the #1 thing that riders value. Now I wanted to talk about some of the key regulatory developments concerning rider status, giving you an update on that. So like all platforms, we're engaged with courts and regulators across markets on issues regarding rider status. And where there's a challenge to our model, we robustly defend our riders ability to work flexibly. And I talked about the appeal of that, and I think that the numbers bear out the appeal of this type of work. Now we're happy to take any questions on any market when we come to Q&A. We want to be transparent on this issue. We're going to highlight what we think are significant developments in H1. Let me start with the U.K. first. So we've said previously that our model has been tested 3 times in the courts, including twice in the high court. And these cases have demonstrated that Deliveroo offer self-employed work. In June, we had another case, this is in the Court of Appeal. The Court of Appeal, again, confirmed that Deliveroo offers self-employed work. And this was a unanimous decision. Now in France, we had a similar judgment in April, where the Paris Court of Appeal found riders working under our current model to be self-employed. I think there are furthermore in France. The French government is consulting on how self-employed platform workers can be given more protections. That's something we support, but they are not considering reclassification employment as an option. And ultimately, this is what we've been arguing for a long time. That riders should be able to have flexibility and security, not just this binary black and white choice between the 2. And it's really important, I think, that a major European economy is thinking this way. So we've been engaging with the French government on these issues. A few words on Spain. So later this month, a new law comes into force in Spain, which means any platform worker is presumed to be an employee. Now we've announced proposals to exit that market, subject to consultation. But I want to be very clear that we would have proposed to exit this market regardless of the law. The timing of this royal decree coming into force serves to accelerate the timing of our decision but doesn't change it, and we don't see any read across from the action taken by the Spanish government to our other markets. In Italy, the sector-wide labor inspection, which looks at a historical operating model from 2016 through 2020 and which is subject to appeal is ongoing. This process remains at an early stage, and we are engaging with the relevant authorities. We call out these events as they are, in our view, the most significant this half. We will continue to disclose any developments in relation to rider matters that we believe to be material in the context of both the relevant country and the business as a whole. And we're happy to take questions on this to provide more detail in the Q&A.

Before we close, let me just touch on guidance again quickly. As a reminder, at our trading update in early July, we revised our guidance after taking into account the strong trading performance in the first half and the positive consumer engagement that we've been seeing post reopenings that Will has now talked about. At that time, we revised our GTV guidance upwards from an original range of 30% to 40% growth, up to 50% to 60% growth now for the full year. And we reconfirmed our gross profit margin guidance while narrowing our expectations to the lower half of the 7.5% to 8% range that we guided to at IPO. We have no changes to our revised July guidance today.

All right. So to summarize the key highlights for this half. Some of the things I want to remind you of. First of all, consumer engagement post reopening has been encouraging. We've seen H1 orders and average order values trending above expectations. We've gained further traction in on-demand grocery, rapidly rolling out further sites with leading brands, and we continue to be excited about the synergies with our restaurant channel, but also their incrementality. And alongside grocery, we're proud of our execution on the restaurant side. We've had strong selection expansion across the U.K. and international segments. We've continued to see strong rider retention and satisfaction despite a sharp increase in job vacancies as economies reopen, especially in retail and hospitality. And finally, we continue to pursue our strategy of investing in long-term growth. These investments are meant to drive durable long-term differentiated improvements to the consumer value prop. And at the same time, we are extremely early in the online food industry, and we will continue to invest in driving new consumers to the category and our platform. So thanks for joining us this morning. I hope this presentation has been useful in framing up the performance and providing more context on post-COVID trends, grocery economics and our rider proposition. So let's go to Q&A. Thank you.

Thank you.

[Operator Instructions] And your first telephone question today is from the line of Andrew Gwynn of Exane BNP Paribas.

Two questions if I can. So firstly, I mean, certainly for me, a much smaller loss in the first half than I was expecting. I think the market is looking for loosely GBP 160 million loss, EBITDA loss for the full year. Presumably, loss will increase pretty materially in the second half, but is there any help you can give us there as to where we may land for the full year? Second one, grocery. It's quite a surprising difference actually. I think the gross profit per order might be much closer to 0 for the grocery business. So are you able to help us out a little bit, just maybe on the gap in say, consumer fees or even the commission. Maybe a bit optimistic, but any help on it would be very gratefully received.

Okay. Thanks for the question. Will and Adam here. Adam, would you mind answering these questions for Andrew?

Yes, happy to. So I think on the first question on kind of the full year EBITDA guidance. So we provided formal guidance on GTV growth and gross profit margin. So as a reminder, we said 50% to 60% GTV growth for the full year and then coming in at the lower half of the gross profit margin range to 7.5% to 7.75%. So maybe you can infer a gross profit range in the back half of the year from that. I guess to help with modeling below gross profit in terms of marketing and overheads, I'd point you back to Slide 12 in the presentation. But as I said in the comments earlier, we felt like that we were at a slower pace of investment, and we were very conservative in our deployment of capital. Certainly, in the first half of 2020 and into the second half given the ongoing CMA investigation, which did not conclude until Q3 2020 and also just the early uncertainty around COVID in the first half of the year. So you saw marketing and overhead step-up sequentially in H1 2021 from H2 2020 by around GBP 100 million. So we would anticipate a further sequential increase in H2 this year relative to H1 this year, but by less than the GBP 100 million increase you saw in H1. So hopefully, that's helpful on that front. And then on grocery, just so I make sure I understand the question. I think it was just around some more color on the difference in unit economics in restaurant and grocery is that right?

Yes. it's just surprising that it's so profitable for grocery. So any extra help would be very greatly received.

Yes. Happy to. So look, I think if you just think about kind of comparative impacts down the P&L across both channel. So relative to the restaurant channel, grocery, we see a higher basket size, and that's a higher AOV. The take rate from the grocer relative to the restauranteur is a bit lower and the consumer fees are a bit higher. But as you can see, the difference in overall unit economics is 30p. So -- and we also talked about where we think there are opportunities that we haven't yet explored to drive more kind of unit economics over time there. So as a reminder, these are the things around...

The only thing I would say, Adam, to that is, so that's sort of on the revenue side. As we add more order density to the network on a hyper local basis, we're driving more and more efficiency. And because grocery, that demand curve is not as peaky as restaurants. We're providing more earning opportunities for our riders at the same time. But we're very, very encouraged with how grocery is going, both from a potential profitability standpoint, but I would say the consumer value proposition has expanded pretty tremendously in the first half.

Yes. Great. And then just for the international for grocery, any sort of big takeaways there? I know it's much smaller, but any extra help?

Yes. I mean we, we're pursuing the same strategy. Again, it's 1,500 SKUs to 3,000 SKUs, the grocer pick-and-pack, you get it in kind of 25 minutes or so. And I think the focus on the international side has been on the rollout with some of these major partners. I think, in particular, calling out Carrefour, calling out Park N Shop in Hong Kong, really kind of partnering with these major players and offering their private label products is a really important area of what we're doing. So the international market as compared to the U.K. is slightly behind on the ODG side, but the signs are very encouraging.

The next question is from the line of Giles Thorne of Jefferies International.

My first question was on Editions. I'm picking up on some of the commentary from the head of your GCC region talking about virtual brands outstripping physical restaurants within 3 to 5 years. I'm actually also noting in the U.K., the comments around having the highest -- I think I forget the exact words, but essentially the highest restaurant count. It feels like Editions is really turbocharging this entire conversation, and it feels like a 5-year window for virtual brands to outstrip physical is a bit more aggressive than what we were hearing through the IPO process. So Will, an updated commentary on how strategic Editions and the concept is overall would be useful.

Yes. I'll take...

Secondly on...

Sorry.

Yes, let me -- yes, I'll get them all out, and then over to you. Secondly, I wanted to pick up on the hire of Devesh Mishra who I've not heard of before, of course. Evidently reading about them on LinkedIn, and world-class talent around supply chain management from one of the world's best companies. But I'm curious why a fundamentally marketplace business like yours suddenly needs to build competence around physical inventory. I'm assuming, it had to do with growth rate, but just some commentary around the implications of that hire? And then finally, still on grocery and I've asked this before, and I'll keep probing, just your latest commentary on the dark store model. And it was interesting to note to lead or question to see Carrefour investing in [indiscernible], They're obviously a big partner of yours in France, and they're now putting money to work in the dark store model. So your thoughts there would be useful.

Great. Thanks, Giles. I've got a bunch of questions here. So let me just -- I'm just taking a few notes here. So I think on the first question, which was regarding Editions, just to clarify the point we were making on the U.K. sort of food merchant side. So the U.K. food merchant side, we're #1 in overall site. That does include our grocery proposition as well. So that's restaurants and grocery, just to clarify. So hopefully, that clears that up a bit, but happy to chat more about that. Look, I think on the virtual brand side, I listened to [ Hani Weiss ] presentation. And I think that I agree with him in certain markets, I think there's definitely opportunity for virtual brands to outpace physical restaurants. And I think if you look at the UAE, Dubai places like that, they're sort of at the forefront of that. But I don't necessarily generalize that to every market. I think there are -- if you just look at the prevalence of delivery in the Gulf states, it's always sort of outpaced that of maybe the western economies. I think they're just naturally in a position where the maturity of that market when it comes to food delivery is a bit ahead of some of the other markets we're in. And as a consequence of that, you're seeing more and more virtual brands. I definitely think that virtual brands have a huge amount of potential, but I also think it's really important that people develop these brands with care. I think sometimes you see restaurant operators -- the creativity of some of these operators are amazing, but sometimes you'll just see them kind of without sort of 5, 7, 8 brands and sort of experiment. I think that is fine, but overall, the more thought that's put into these brands, the more planning that's put into these virtual brands, the higher level of success, which I guess is fairly obvious. Now on Editions. So I think that, we -- in the first half, we've added around 40 kitchens in the first half. But the vast majority of that was in the second quarter. So we are definitely ramping this up. I think one of the issues has been during COVID, it's been a little more challenging to secure sites. There's been a lot of issues with COVID. And as we sort of roll out of that, we expect an acceleration of Editions rollout. I would also say that you asked about Dubai, the UAE, we rolled out our largest site to date. So we rolled out fourth site in Dubai over 8,000 square feet with over a dozen restaurants. So I think there's a lot of very interesting things happening on Editions. I think the key point is out of COVID, people view this as a core part of their operating strategy as opposed to sort of a nice to have thing. So we're incredibly excited about this. On your second question, which is really around Devesh. So, Devesh is, I'm, first of all, so excited that he is joining. And I think for a few reasons. One, he has overseen just an incredible amount of scale at Amazon -- because I think if you take a step back and actually think about what sort of problem he was solving. Now you asked -- because I think the sort of VP of supply chain title may just be a little confusing. But actually, if you think about what he's done, it's basically this. Forecast the amount of demand for the millions of SKUs on the Amazon retail business and the marketplace business, make sure you source those products from both marketplace vendors and manufacturers across 185 countries, have a system in place to predict what warehouses those go to, what fulfillment centers they go to and also making sure that reaches the consumer on time and then also understanding the downstream impact of adding any level to the retail business to the marketplace business. So that business, if you think about it, it's mind-blowing to me that something like that can be automated to the degree that it has been, right? And when we think about that and we think about the problems that we're trying to solve now they're not directly applicable one for one, but many of the innovations that Devesh and his team has built in terms of machine learning. And really, machine learning and automation, I think, apply heavily to the businesses we're in. So I wouldn't read too much into the "supply chain side of this." Just the thought of someone operating at that scale, automating some of the most incredible challenges out there is -- I think, makes me super excited, and I think it gets our company, the engineers, the product people, the data scientists, incredibly excited about working with him. So really looking forward to that. And then finally, the sort of dark store model. Now look, I think that we we've been obviously -- we're in the epicenter of this dark store sort of backfill in London. I think there's got to be now 10 or 12 players out there, certainly getting a lot of discounted food products from all of them. So very happy about that. I think that there are a few elements of this model that we actually -- we find very appealing and we think are very interesting. And I've said it before but probably worth repeating. Number one, on the inventory management side, having a dark store where you know with high levels of precision and what's in stock between the 4 walls, is critical to really preventing issues such as stock outs, right? I think secondly, because you're in a dark store, the pick-and-pack is more streamlined and faster. Now, and so from that standpoint, I always think about it first from the consumer and then I think about it more from a Deliveroo perspective. I think from a consumer perspective, I think the proposition is really good. Now certainly, we don't know what the market clearing price of some of these products are because of the heavy amount of discounting. But I think overall, that stuff makes a lot of sense. And then, of course, compare it to our model, which I think is much more scalable because we're not building out sites. We also have the advantage of partnering with supermarket chains, right, the likes of Aldi and Sainsbury's and Co-op, et cetera. These are extremely valued partners to us. And I think that we can offer their private label products in a way that some of the dark store 1P operators can't do. So we absolutely think our model makes a lot of sense, but we do think elements of the dark store model also makes a lot of sense. And at the end of the day, I'm just going to be pretty open-minded. When I launched this company, people told me that it would that -- it was pointless and it would just never work. And so as a consequence of that, I'm just super open minded. Any time I see something that I think is potentially good for our consumer that a different sort of business model is producing. I'm going to look at it very, very closely. The other thing I'd say is, on your question on Carrefour and [indiscernible]. Look, I don't have the details on exactly what that arrangement is, But I think Carrefour is one of the most innovative forward-thinking grocers out there. I think they've got that big China business. They probably studied closely what that business has meant to them. And I'm sure they're exporting some of those ideas out there. And I'm just excited that a grocer is taking a big step like this. They're a great partner to us, and we're going to keep engaging with them. But I think it's a very interesting step.

The next question is from the line of Rob Joyce of Goldman Sachs.

Sorry, I've also got 3. First one, just on, within your monthly average consumer, customer data, sorry, can you give us an idea as to whether you're seeing, what sort of level of churn you're seeing in there and whether that's stepping up quite a lot? Just the reason I ask is because if I roll-forward the MACs, you've got at the end of the second half, and I think the full year guidance looks very well underpinned. Is that a fair conclusion? And then on the first point, it would be nice to hear. Second one, in terms of the Plus, can you just talk a little bit about the decision to cut the subscription fees there, the rationale there and whether the comments about this being lifetime accretive still stand at those lower fee levels? And thirdly, a quick one on monetizing the grocery platform. Can you confirm that you own the data of everything that sort of process through your platform?

Okay. Thanks, Rob. We've got a number of different questions in here. Hold on, I'm just taking a few notes. Okay. Data Okay. Adam, do you want to take this first question on MACs. And I think, I believe, Rob, you're basically saying, is there implied churn in MACs because if you roll it forward, then basically, you come up with a different top line number. Is that more or less what you're saying?

Yes, that's the second part. And the first one, just in terms of those churn, are you already seeing churn level sort of pick up versus the sort of as the reopening pickup is churn also picking up?

So maybe I'll answer that first part, and then Adam can kind of talk about our sort of GTV guidance for the second half -- sorry, for the full year. So we walked through a number of different countries and different reopening scenarios. So in a market like the U.K., we really haven't seen an impact. I think in markets in Asia and the Middle East, we've continued to see very strong growth post-reopening. I think in Europe, what we said was, and France and Italy are slightly different in sort of the impact. But as reopenings have occurred, we've seen people flock to restaurants, but that has also coincided with the beginning of the summer seasonality period, where we always see a drop in demand, right? So if you think about it from a sort of monthly retention basis, we always see a pretty significant drop during the summer period in places like France and Italy, in a way that we generally don't see in the U.K. historically. So it's a little hard for us to disaggregate those factors. That's why I think towards the end of Q3, we will have a much better view of some of the European markets. But sort of underlying consumer behavior in those cohorts in some of those markets from a frequency standpoint is unchanged. So it does feel like certain people are on holiday or what have you. So we'll know more of that in Q3. But in the U.K. and in the Asia and Middle East, we haven't really seen any change. And then I think Yes. So Adam, maybe you can talk a little bit about how that might affect our view towards GTV.

Yes. Thanks, Will. So I think on the revised GTV guidance, Rob, we try to take into account if we look back to Q1, the big piece of information we were missing at the end of Q1 was what was going to happen to consumer engagement post-reopening. I think on the whole, we've been encouraged about what we've seen. We've seen continued growth, as Will mentioned, even in the markets that have been out of lockdown the longest in Asia and the Middle East. I think that what we've seen in the U.K. across 3 different reopenings has been encouraging with no material impact to consumer behavior. And I think what we've seen in Europe is small to moderate impact, as we've talked about, but hard to disaggregate that, as Will said, from some of the summer seasonality. And so we tried to take all of that into account and our GTV guidance. And there still is some data that we'll have a better readout at the end of Q3, kind of early Q4, as Will mentioned, in our European markets.

Yes. And on the Plus question, so I think there are a few things happening at the Plus level. So we constantly run experimentation on what types of pricing we think our customers prefer. And what we found on the Plus side after running very lengthening experiments with huge amounts of participants was that overall, customers preferred a lower monthly fee, but that is offset by a slightly higher service fee, right? And so Adam, I think the comment on whether or not it is lifetime accretive has not changed.

Not changed.

Yes. So I think the unit economics net-net are I think either the same or very, very similar. But the presentation and the distribution of those fees is slightly different than it was before. And then finally, on the grocery side. The question was, do we own the data that goes through our marketplace, right? Adam, I don't know if you want to just take this one.

Yes. That remains the same, Rob, whether it's across our restaurant or grocery partners. So the data that comes through our marketplace is owned by Deliveroo. That doesn't change.

The next question is from Georgios Pilakoutas of Numis.

First one on MSCs. These have been on your platform in the U.K. for a while. I was just wondering if there was any learnings that you're seeing from that partnership? A quick follow-up is just could you clarify what you mean by inventory management when you talk about on-demand grocery? Editions, you've taken 40 kitchens in the first half, are you able to kind of share a number for the second half in terms of how that pipeline is growing? And then a couple on kind of your U.K. business. London, there's been a lot of investment coming from competitors, a lot of third-party data are also being put out by other players. I guess just wondering, looking for a comment in terms of how your London business is performing and how you're feeling about your London business right now? And then on the other side, you've expanded your population penetration quite significantly in the U.K. And I was interested to hear any update on thoughts of the model in some of those more rural or smaller towns? If you're seeing any differences in consumer behavior? And then how does that impact your view longer term of the Deliveroo model as in kind of the ability to reach, I think you used to speak about 80% of population coverage. Is that broadly where you still think you can get the business to?

Okay. Thanks. We got, I think, a lot of questions. I'm just taking a few notes, hold on 1 second, please. All right. I will try to make sure I got everything to -- Okay. So your first question was really around Dija, right? And yes, we have a partnership with Dija. For those of you on the call that haven't heard of them, they're one of the many dark store operators in London. Actually started by 2 ex Deliveroo colleagues of mine. So we definitely know them. Well, I think the learnings from Dija are still pretty early. I think that what's been interesting is they've been operating both a Deliveroo delivered model and also their own fleet and sort of -- and I think they've been trying to test out what actually works better, which I think is a pretty interesting experiment because I think almost all of these dark store operators have their own fleet, right? And I think they're trying to figure out at peak times, do I turn it off, do I turn it on. They're also trying to figure out how well does this do on a marketplace. I think that what we've done from our perspective is it's actually performed very well on our marketplace. But I think what we're kind of struggling a bit with is how to properly merchandise something like this, right, which is slightly different than our traditional grocery offer. And so we're digging in, we're learning more, but I think those are some of the initial conclusions, which may be an unsatisfactory answer to be honest, but it's pretty early. So we're trying to learn more. On the inventory management side, what I mean by that is if you think about a traditional grocer, right, they usually have a front of house and back of house. And so you've got a lot of things stored in the back. I think grocers generally have a very good sense of what is in the entire 4 walls of the building. But kind of differentiating between what is in the front and back usually is done by, I'd say, possibly machine learning models, although I would suspect probably something a little more -- maybe a little more just strict algo-driven, right? And so I think that if you combine that with the fact there's off-line shoppers coming in or the vast majority of the shoppers, right, coming in and taking that piece of salmon off the shelf, we may be in a position where that's unavailable. Whereas in a dark store, you just know right away, whether or not something is there, right? And I think that is something that we think about a lot that I think drives a better consumer experience in certain instances, right? So that's why I talk about it. On the Editions pipeline, I think that your question was is -- and just interrupt me if I've got it wrong, but I think your question was do we see that growth continuing in the second half? Was that it?

Well, I guess you spoke to 43 sites in the first half. Do you have kind of a rough figure for the second half? Is that 43 or that can accelerate?

Oh, yes, yes, yes. So I hesitate to give you a very specific answer on that, not because I don't want to. It's really because the process of securing a site during COVID has honestly been just a little all over the place. I think I would guide you in the direction that we are going to expand this more. But to give you sort of a breakdown of how much more than the first half, I think it's a little hard, But we definitely want to expand that proposition. So we will be. Whether or not that number is -- I can't commit to a very specific number at this point. But once we know more, we'll definitely give you more data. I think on your question, which I believe is a combination of a few different topics, which is sort of market share and how we're doing in London. I guess I would say a few things. First of all, when anyone talks about market share and growth rates, I think first, I always ask. Well, what metric are we talking about, right? Are we talking about app downloads? Are we talking about monthly active consumers. We're talking about orders, we talk about GTV. And then secondly, what source are we talking about? Are we talking about credit card panel data? Are we talking about published financials? Are we talking about an e-mail sort of receipt reader, right? So my view is this. Now when I think about metrics, I care most about GTV because ultimately, that is the total consumer spend that flowing through your platform. And then the source I care most about is published financials. And the good news is that you have that from 2 of the 3 major players now. And then secondly, I think the second best source of data is actually credit card panel data because I think the e-mail receipt data can be interesting from sort of -- it can give you a more rich picture, but the overall accuracy is actually much lower. So that's sort of my context. Now when I think about it, as to our performance, I'm very pleased with our performance in the U.K. We now have published data from ourselves and one of the other players. And if you compare our GTV growth rates to that published competitor data we've had consistently higher growth quarter after quarter. So let's just take the second quarter, for example. We grew GTV 87% in the U.K. and Ireland compared to a published figure of 40%. So we're very pleased with our U.K. outperformance, not on just this quarter, but also on a sustained basis. The other thing on London specifically, I would say, is this. There's been strong competition in London for a number of quarters. I think we talked about in the past that the Just Eat free delivery campaign, which I think was something like 9 months to 12 months. I think it might be over now. I don't really know. It's been intense. But like I said before as well, it's not as intense as it's been before, right? I think the height of the intensity for us in London was 2016. This I will say as well, a lot of the competition we're seeing is focused on the very, very value end of the market. We see a lot of discounting of sort of QSR meals. And so people fighting really hard discount in the QSR meals. Now our focus has been different. We have a very differentiated offer in London. We have brands like Dishoom and Shake Shack and Five Guys, Waitrose, Whole Foods, Wagamama, those are all exclusive on our platform. And we've been continuously improving our value proposition through better service and selection. And really, that's where we spend our time, and we've talked a little bit about Plus as well. And so I think the competition has impacted market share for some players but we have not been affected. If I look at our GTV, our London market share position has been unaffected this quarter and frankly, has increased over the last 12 months. That's what our data says at least. Final question, sorry, I know this is a long one here. So your question really was on population penetration, I think, of like what that could be over time. And as I said, we're over 70% today. By the way, we're kind of repeating again the growth that we drove in the first and second quarters. The vast, vast, vast majority of that came from areas that we were in over a year ago, right? So what we think is great is we're planting the seeds for future growth. And we hope over the coming years that these customers will develop into the healthy cohorts that we've seen historically in other parts of the U.K. The metric we really look at is what is the population density of the built-up urban area, right? So we don't look at overall population density. The U.K., I think, has some advantages, which is even these smaller towns and cities of, call it, 30,000, 40,000, 50,000 people, actually a quite high population density in the built-up urban areas. And those will be the areas that we continue to target. We -- just to kind of finish a bit. We're not targeting these rural villages, places where people are super spread out, that really doesn't work for our model as well. The good news is, in the U.K., any place where there's people generally implies a pretty high population density in those built-up urban areas. I hope that helps.

The next question is from the line of Sarah Simon of Berenberg.

I've got a couple of questions. With the Editions rollout, can you give us a bit more color in terms of how you'd expect CapEx to evolve over the next, say, 12 to 18 months? That's the first one. Second one, you talked about reinvesting in the proposition and so on and how that's impacting gross profit. Can you give us a sense of how the year-on-year change kind of splits between more discounts on delivery versus slightly more generous take rate or less aggressive take rate? So let's say, the split between investment in the consumer and investments in getting the restaurant on board. On the third one, that was helpful to get the size of the grocery business. Can you give us an idea because you said at the time of the IPO that the U.K., I think, was around 10% GTV? Can you give us a sense for how that it was at the end of the first half?

Thanks, Sarah. Adam, I think these are all questions for you, but let me just repeat them to make sure we're all on the same page here. So I think that's number one, what is the CapEx profile of Editions looks like over the next 12 to 18 months. Second, if we look at sort of the investments we've made on that gross profit line, could we split out sort of investments in differentiated content from consumer acquisition, right? I think discounts, I think she mean new user incentives, right? And then finally, just where we are at in grocery as a percentage of GTV in the U.K.?

Yes. Yes, happy to take those. So thanks, Sarah, for the question. So on the CapEx evolution, I think you said over the next 12 to 18 months. As Will said about when he was talking about the Editions pipeline, we are planning on ramping up investment. And at the sites that we added in H1 and kitchens we added in H1. The majority of those were in Q2. So we have been ramping. We would expect that to continue to ramp over the next 12 to 18 months. But I think at this point, just hesitate to give a firm number given we, there's some part of that, that we, with the process of securing the sites is something that we're still working through, but the intention is certainly to ramp that spend over the next 12 to 18 months. I think on your second question on reinvesting in the consumer value proposition on the gross profit investments, And you asked about the split between how much of that is on differentiated content versus on things and on the consumer fees. So look, I think if you think about kind of the primary areas of investment for us, it's really -- in those 2 lines, it's really kind of 2 things. So on the differentiated content, this is about continued rollout of grocery and expansion of that. But also continuing to maintain and secure and expand our partnerships that are strategic with restaurants. If I think about kind of just the U.K., we've expanded the number of partners that are -- the number of exclusive sites with us in kind of that independent segment by over 60% in the last year. And globally, that's just under 50%. And so it is something that we are investing in, and I think we feel like there is good value there for us, where it makes sense to do that. On the consumer fee side, again, just I think you mentioned discounts, but really just to clarify, the examples of the things that we kind of are most excited about in that investing in there. One is on the new user piece. This is about giving people that haven't yet tried the Deliveroo product or ever been on our platform. Their first 14 days, free delivery. Now that is something that we have a very clear view as to what the return is on that. When you think about kind of payback on a program like that, that's about a year payback for that 14-day free delivery acquiring that customer and getting them to pay back on our platform. So we're super excited about that. We started ramping that in Q4 last year, kind of with the hindsight of what was happening with COVID in Q2 of last year and wishing that with hindsight that we could have invested more aggressively back then, and we've continued investing behind that.

Well, I don't know if we could have. We were in the middle of an antitrust case. I mean it was not I don't think -- I mean, it was tough situation.

Totally. Yes. I'm saying more in terms of dynamics of what was happening. And then the other one, Sarah, is around our Plus program. And so this is another one where if you just look at from the end of last year until kind of the end of the first half, we've grown our Plus subscribers by about 1/3 from where they were there. And that's really been, it took us some time to get to a place where the economics of the program were something that we wanted to invest in because you do take away the actual per order delivery fee. As Will talked about some of the -- pricing model is something we're looking at. We are excited to roll out a program called Plus Lite in Q1 of this year, which offers an even lower subscription fee in exchange for a higher minimum order value targeted at segments like families. But we're investing not only in moving people to Plus, which is dilutive on a per order unit economic basis, but still accretive on an overall aggregate lifetime value or gross profit basis, but also on incentivizing them to move to Plus and letting them trial that for a period of time. And then I think that -- sorry, the last question was just on grocery and U.K, the U.K. GTV kind of as a proportion of that. So I think what we've said is at the tail end of 2020, U.K. was just about 10% of the GTV was grocery. At the end of Q1, we said it was more than 10%, And we're now saying that at a group level for H1 that was 7%. I think as Will mentioned, the international markets are a bit behind the U.K. in terms of where they are in the rollout, but it is something we are expanding aggressively.

And we have time for one final question, which comes from the line of Marcus Diebel of JPMorgan.

It's Marcus Diebel. Just 2 questions, very short. Well, when you commented on the GTV growth, could you give us any indication, I know it was [ anticipated ] in the past, any indication about the split food and grocery orders or at least, yes, in some areas, any comments would be quite helpful. I don't know if you can do it for the U.K. entirely, but anything you can comment on? And then secondly, very quickly for Adam. So you're talking about the GBP 100 million kind of like investments below gross profit in the second half. Did I get that correctly, yes, GBP 100 million additional investments in the second half below gross profit. And also, is there a bit roll over of this GBP 100 million also into H1 '22? Is there basically a similar amount to H1, next year, should we think about it this way?

Marcus, Will here. So your question on I think the split of grocery and restaurants. So I'll get into it in a second. But I guess from my perspective, I think the company's perspective, we always, I think about the world. We think about ourselves as the definitive online food company. We think of the 21 meal occasions out there for each individual. And ultimately, whether that comes from the restaurant side or the grocery side, we want to fulfill those occasions for the consumer. And they're the same customer, grocery is 100% incremental. And us sort of embedding ourselves deeper in people's food occasions is ultimately what we're trying to do. In terms of the sort of split, Adam. I don't know -- we've said that it's 7% of...

The total GTV.

The total -- I guess, Marcus is looking for whether that growth rate is different. I mean I certainly would say that, yes, the grocery side is going faster than the restaurant side, but I don't think we've disclosed any more details that, right?

No, exactly.

Okay. And then, Adam, do you want to take the second one?

Yes. So just to clarify on the marketing and overhead piece of the market, what I was saying before. So again, just as a reminder, on the context. So if you think back to 2020, and we just talked about this briefly, but there were kind of 2 things happening in the first half of 2020 and into the second half that caused us to be very conservative in our deployment of capital. One was we were still going through the CMA antitrust investigation, which did not wrap up until Q3 2020. And the second was the uncertainty around consumer behavior at the beginning of COVID. If you recall, in Europe and the U.K., many restaurants were closed, not just for dine-in but also for delivery, and it was uncertain what was going to happen. So us, along with many other players pulled back spend at that time. So, if you think about the sequential half-on-half growth in the marketing and overheads line from H2 2020 to H1 2021, that's about GBP 100 million up to now GBP 291 million in H1 2021. And what we're saying is that we would expect a further sequential increase in H2 of this year on top of the GBP 291 million in H1, but less than the GBP 100 million sequential increase you saw in the last half. Does that make sense, Marcus?

Yes, that makes sense. That I understand. Perfect. But I have to ask on the first question on GTV. I mean, some investors want to know. I understand the argument that you look at it more from a view like kind of overall consumer orders, but some investors just want to know and why can't we hear the share of food orders in the U.K.

I think we definitely think about it as the same set of food occasions. And as I've said, the grocery GTV growth rate is indeed faster than the restaurant one in the U.K. Is that -- can I just make a closing remark? Is that okay, operator?

Sure, please go ahead.

Yes. I just wanted to say thanks to everyone. It's been, obviously, we're 18 months into what has been a very difficult time for a lot of people. And I just wanted to give the word of thanks. So I want to thank the analysts for your interest in our company and the hard work you guys have been doing. I want to thank our investors for your support. I think in particular, though, there's 2 groups I want to thank, I want to thank the 3 sides of our marketplace. I think about our restaurant that have struggled so much for the last 18 months and kind of seeing your businesses recover and also in pricing delivery, of course, has definitely got a smile on my face. I think for our riders, we're here to make sure that you guys have really great flexible work. And I thank for our consumers, we're really proud of the fact we're playing a bigger and bigger part of your daily lives. And then finally, to the employees at Deliveroo, I just want to thank you for your extremely hard work for this half year. I think we're incredibly happy with the results, but also acknowledge that it's pretty tough working remotely, and I really hope to see everyone back in the office soon. So thanks very much, everyone, and have a great day.