Afternoon, everybody. My name is Stephen Daintith. I'm the group CFO of Ocado, joined in March last year, and this particular session is something I've been looking forward to do for quite a while, actually. It's probably best to explain why that is by going through the objectives for the session this afternoon. Welcome to those who are joining us virtually, watching us on the screen. What we're trying to achieve in the next hour or so is provide a better understanding of the characteristics of the underlying business models in Ocado Group. A more intuitive way of looking at our three different business models. Provide a framework, how you might assess revenues and returns for each of these businesses, and how they might develop. For Ocado Group as a whole, over the next four to six years, we define that as midterm.
Just to be clear, I know there's all sorts of definitions around medium term and midterm. We're using four-six years as a midterm for the basis of our view of the trajectory of Ocado into that midterm period. We're gonna take a deeper dive into the workings of our Technology Solutions business. Now, that's the key driver of value for the Ocado Group going forward over the next five, 10, 20 years. And it'll explore why, based on the progress that we've seen so far and the visibility that we have ahead, that the investments that we're making today will drive very attractive returns in the future, tomorrow. Finally, we're gonna set out a clear path to in excess of GBP 750 million of EBITDA in the midterm, and these are based just on existing contractual customer commitments.
It's a really important point. There's nothing speculative here in this. There's no new business to be won. This is based on existing commitments that we have today with our customers and the roadmaps, that's GBP 750 million. At the same time, as part of that, we're gonna show a roadmap to cash flow generation that is sufficient at a steady state to build 10 live CFC sites going live every year, five module sites on a run rate basis, and still break even or even move into positive cash flow territory. That's what we're aiming to achieve over the next half an hour or so. Here's the agenda.
We're gonna start off just by mapping the three business segments that we have today to these three business models that I've just run through around Ocado Retail, Ocado Logistics, and then Ocado Solutions, Technology Solutions, which is the combined U.K. and International Solutions business. In summary, we're gonna bring it all together. We're gonna show, by the way, what a pro forma numbers look like in fiscal 2021 for those three business models and the drivers of the performance in 2021 and trajectory through to that midterm. What else shall I say on this one? On these business segments and business models you're about to see, it may be, in fact, it's very likely that we will re-segment our formal reporting at some point soon along these lines, because it makes sense to do so.
Again, just going back to my earlier comment, it's a more intuitive way of looking at Ocado Group. Probably most likely to be fiscal 2024, but we're gonna see if we can do that in fiscal 2023 next year. Okay, so Ocado Group. This chart, this slide is a reminder, really, of what the group is all about, changing the way the world shops for good, technology-led. Three business segments down here that you'll be familiar with and our ambition to reimagine the shopping experience. We thought that these facts around the progress over the last five years are pretty helpful as well. I'll call out one or two of these. We now have eight additional partners from the three that we had five years ago. We've got 11 global partners, two here in the U.K. and nine in the rest of the world.
At the same time, those partners generate over GBP 210 billion of sales between them. This is not online sales, this is sales that they have across their markets, but it's sales for us to go for with those existing customers. What else? 57 new commitments, around 58 CFC announcements and customer commitments that have been made. That's up from one just five years ago. That's probably sufficient for now. A pretty impressive list of achievements over the last five years. Let's go through those reported segments as they stand today and the underlying business models. First of all, as a reminder, on the left-hand side, this is the business you're familiar with today, Ocado Retail.
We consolidate those numbers in their entirety, 100% consolidated, and that is a joint venture that we have with Marks & Spencer, 50% JV. We have our U.K. Solutions and Logistics business. An industry-leading logistics business, not a simple 3PL, much more than that. It's providing services in the U.K. for Ocado Retail, the JV, and for Morrisons. We have International Solutions, and it also, sorry, includes in U.K. Solutions and Logistics, the U.K. part of our solutions business. That's what it does today in the way we report at the moment. The way we're gonna show the group over the next half an hour or so is to move the U.K. part of this business solutions in here, along with international solutions to give you that global perspective. Therefore, this business will be a pure logistics business that we will look at.
Sorry, let me just go through this one here. Ocado Retail, a breakdown of the revenues, pure-play online grocery business. The costs of Ocado Retail, cost of sales, distribution and fulfillment, marketing and head office costs. They also include the recharges of logistics costs from this business to provide their shopping baskets to their customers. This is the business model for Ocado Retail, the retail business. Our pure-play logistics business is the recharge of the costs incurred, that's in the revenue line, to execute those logistics services, and those are recharged to Ocado Retail and to Morrisons, two customers. The costs incurred to do that are recharged, as I've just said. There's also, within the cost base, an allocated share of around 30% of the total group operations costs, and that is human resources, finance, and legal. Finally, Ocado Technology Solutions.
These are the fees from the global retail partners, that's those 11 partners that I referenced a little earlier. The revenue is two key components. Up-front fees, now this is the contribution from our partners towards the capital build, and they are recognized as revenue over time as the CFCs go live. There's a release every year of that cash, the receipts that sits on our balance sheet, and that's what goes to the revenue line over time. At the same time, there are recurring capacity fees. You'll see shortly, we talk about a mid-single-digit fee of sales that are the recurring capacity fees in Ocado Technology Solutions. Now, the cost base for this includes an OSP cost, as well as a full allocation of the group technology costs.
That's the non-capitalized part of the GBP 250 million or so of tech cash spend. It also includes 70%, there's the other 30%, of the human resources, finance, and legal costs. That's an overview of what the businesses are. Here are the underlying business models and how the 21 actual reported numbers, that's this column on the left-hand side, tracks through to the pro forma numbers for the new business, the way of looking at the business on the right-hand side. Just gonna have a glass of water. As you might expect, the Ocado Retail number doesn't change. It stays as it is.
U.K. Solutions and Logistics from revenue, we take out the Solutions part of that business, and we have a pure logistics business, GBP 595 million, 90% of which are the recharges in respect to the costs of providing those services to Ocado Retail and Morrisons. The remainder is a management fee of 4% of total costs, and then capital recharge fees, and that's in respect of Hatfield, Dordon, and Erith. That's a fee, by the way, that's gonna wind down quite significantly over time. Five years from now, it'll be less than half the number that it is today. It's an important point when it comes to margin for the logistics business. Technology Solutions, fees received from global retail partners. That's currently, globally, GBP 183 million.
As a reminder, that's a mixture of capacity fees and the release to the P&L account from cash that's already in the bank that's being recognized over time, that initial contribution up front to the CapEx build. If that makes sense. You'll see shortly that this number is gonna grow to over GBP 900 million over the midterm, and similarly, will the U.K. logistics business as well. It's coincidental that those two numbers are the same, by the way. We're also gonna show the progression of what we expect to be the progression of the retail business over the midterm as it sells into the capacity that we're building today and investing in today. EBITDA, GBP 150 million. Again, no change.
Logistics business is a GBP 31 million EBITDA business, and that margin, that 5.2% margin, reflects that 4% management fee on the costs that we charge, plus, importantly, the capital recharges of GBP 40 million in 2021 that I referenced earlier. Our technology solutions business is currently loss-making at EBITDA level, but it includes an attractive positive contribution offset by the full allocation of upfront investment in technology and head office costs. We've got that full P&L impact of the non-capitalized technology costs, and a significant contribution, 70% of group operations costs, leading to that GBP 81 million number. We're gonna show later on the trajectory, because we get asked this question a lot, for what is going to happen to those central costs over time. We're gonna share that shortly. Hopefully that all makes sense and it all stacks together.
That's how the math works. We still have GBP 61 million of EBITDA at the bottom of this chart, but that's the mapping of the business. You know, we believe that this is a much more intuitive way of looking at our business, and it allows you to get a proper understanding and interpretation of the Technology Solutions business, which is the most significant driver of value creation for Ocado Group. Having got through there, let's go through our underlying businesses, business models, and their key drivers one by one. Ocado Retail. This slide looks at, first of all, what's happening in the online market, where Ocado has done very well. The online market itself has gone from 7% to 12%. Market research suggests that that will grow to 18% by fiscal 2025.
We are outperforming the market today in a declining online market as we get into that normalization phase post-COVID. Our retail revenue, as a reminder, in Q1 of this fiscal year was at 32% versus the same time two years ago. Despite the trading statement that was announced today, as a reminder how strong the business is performing compared to where it was two years ago. Net promoter scores are 25 percentage points better than other grocers. We get very, very good feedback from our customers about the quality and accuracy of the delivery, the limited substitutions, the time of the delivery. We get very good customer reaction, and indeed, if there's a standout number for our international partners, it's the scores that they're seeing as well on NPS for those customers overseas.
We've kept customer penetration stable at 26% despite the volume declines and they're driven by the cost of living and normalization of shopping behaviors. The smaller basket in particular, which is trending at the moment around 10 items lower this year than it was last year. 47 items versus 57 items. Growth, revenue. The headline here, we see a route for Ocado Retail to GBP 4 .5 billion of revenue over the midterm, bridging from GBP 2.3 billion with the modules at site maturity that we have today. That's off the back of, at the end of this year, these six sites, Hatfield, Dordon, Erith, Bristol, Andover, and Purfleet. We really should have included Acton, the Zoom site in here to make it seven to get a complete like-for-like basis.
Going forward, this is how the sites progress on the back of recently announced. Well, all of these have been announced. We haven't named specifically the locations in the Northwest and the Southeast. But we've got Bicester and Luton over fiscal 2022 and fiscal 2023, respectively. All of these will give us the capacity as we add these modules at sites at the end of each year to a run rate by fiscal 2025 of 70 modules at site capacity at site maturity. As a reminder, the rule of thumb, one module is equivalent to around GBP 70 million of sales at capacity. This provides the runway to revenues in excess of GBP 4.5 billion. Now, clearly, there are factors that might affect the shape of that progression and timing around customer acquisition and retention.
We're growing customer numbers very well in Ocado Retail right now and retaining customers. The e-com shop, the basket, as we know, we see today, is an important variable, as indeed is the average sales price for each. Just to flag the facts, take into account as we consider the progression. We feel confident around the GBP 4.5 billion revenue that this you know, we shouldn't just assume that because you build it, they will come. The net promoter scores, we have never had excess capacity in Ocado. It's something that over the years I hear that we would have loved to have more capacity than we've always had, particularly during 2020 and 2021. We have good confidence around the visibility of filling this capacity as it's added and extends our geographical reach.
As a reminder, we only actually have access to 75% of U.K. customers today in the U.K., so there's plenty of room still to go for. The margin. Fiscal 2021, we delivered a 6.6% EBITDA margin and a particularly strong first half performance in excess of 8% EBITDA. We did the same before, by the way, as a reminder, in fiscal 2020. Now, clearly, there was a COVID benefit in those years. Notwithstanding that, margins of between 6%-7% in each of those two years. We just highlighted this morning that we expect a low single-digit margin in fiscal 2022. Over 90% of that margin pressure is driven by what we consider to be three temporary effects.
Number one, around 40% of the impact, operating leverage as Ocado Retail builds into additional capacity that's been added, but smaller immature volumes relative to the fixed costs across either overheads, capacity fees to Ocado Solutions, of course, and site costs. 30% of marketing costs. It is an incredibly competitive space right now in the marketing arena for online retail, especially in the immediacy market. We're also investing to grow into the capacity that we're adding, particularly as Bicester goes live. There's a spike in marketing spend that's another key driver of that low single-digit margin. Then finally, around 20% of the impact is the utility cost inflation that we're seeing across the country, particularly diesel and electricity for Ocado Retail.
In the midterm, though, we see a return, or I should say, we see the route to a high mid-single digit EBITDA margin. If you'd like me to get specific about that, 6% is our number. May go beyond that, but 6% is the number that we've used for our math for the whole flow through. Underlying trends are encouraging. Customer acquisition continues. This goes back to my point that I raised a little earlier. 12% growth in customer base year to date, and we are well invested for that future growth, that over GBP 4.5 billion of revenues with investments that are already made and were now flowing through as sites go mature and we grow the customer numbers. This is an important point around the efficiency of our operation.
Andover and Purfleet are now operating at over 200 units per hour picking. That's per labor hour, 200 units picked. More than 200 in a shopping basket of 55 items. Four family weekly shopping baskets in an hour, one every 15 minutes. In fact, we are doing 220 in Andover, and Purfleet is getting very close, if not to that number as well, when we look at day-to-day trends. We think we can take that number even higher, particularly as the Re:Imagined products go into place. More on that later. That's our ambition and intent for Ocado Retail. We have strong conviction around this, around the revenue and the EBITDA margin. Ocado Logistics. Here on the left-hand side, a 2020 versus 2021 comparison.
Each is, that's a, an Ocado term that I've only in the last year become familiar with, but that is individual picks of stock keeping units in the shopping basket. At Ocado, we call them an each. You can see they've grown 3.6%, 2021 over 2020. Here are the cost recharges that we are charging our customers, growing in line with that growth in cost per each. The logistics business is very much a cost pass-through business. The volume that's going through the logistics business is the biggest driver of revenues and costs.
Going back to that earlier comment around the capacity that we're adding in the U.K., the capacity that we're adding by the midterm, we'll be doing well over one million orders per week compared to around 750 or 1,000 or so at the moment. Growing significant capacity in the U.K. with those investments on that slide that I showed a little earlier around Luton, Bicester, the Northwest, and the Southeast. The capital charge that I mentioned a little earlier will reduce by more than a half in the midterm, but the margin will trend towards 3%, and that pretty much reflects that underlying cost-plus business model. A pretty simple business model. Gets you to GBP 35 million of EBITDA in the midterm with that 3% number. Technology Solutions.
Reminder of this business, 'cause it's the first time really you've seen this business in the round as a global solutions business. It's a leading solutions provider, world-class customer experience. As we will see shortly, proven operational economics, which we think can only get better, and I'll share with you in this presentation some detail around the Purfleet CFC, return on capital that we're seeing at this stage based on the existing cost base and with more to come. Solutions is enabled through the Ocado Smart Platform. This is a suite of solutions, end-to-end software, but it's combined with our physical activities in the warehouses, which we then sell as a managed service to our customers. 11 of those customers around the world, over GBP 210 billion of sales, and we have announced committed capacity rollout plans equivalent to 58 CFCs.
That's using that proxy of 5 modules per CFC to arrive at that 58 number. That would represent over GBP 20 billion in client sales for that 58 CFCs. Capacity rollout is picking up pace. We went from 5 CFCs at the end of 2020 to an expected 19 by the end of this current fiscal year. We expect a 10+ sites per annum run rate in the years beyond. For the purposes of our model, we've used 10 5-module sites going live each year. Shortly, we're gonna look at the CapEx phasing, but clearly, if it's a constant run rate, that CapEx number pretty much stays flat through that plan based on that assumption.
As the assets we are currently installing our clients go live and ramp, we then expect strong returns for both Ocado Group and our partners, importantly. Those returns will get even better as we iteratively continue to improve our operations and the Re:Imagined innovations are still to come. There's no impacts of Re:Imagined benefits in the operating cost that I'm about to show you. Indeed, when we get to the direct operating costs for Ocado Solutions in the CFCs, there's a revised improved target guidance that we're putting out today, and that is despite or before Re:Imagined benefits. Still under discussion, the extent to which we share those benefits with our partners. Just reinforcing the point. Okay, revenue, how does that progress? Well, strong ramp as the CFCs roll out. A familiar story.
As those CFCs go live and the modules go live around the world, the revenue potential for this business, based on that mid-single digit fee from our customers, is over GBP 900 million of revenue potential by the midterm. Important point again, to emphasize, this is based on existing customer contracts. There's no speculative business in here. This is based on what we know and have today. We're gonna show how we turn that revenue into both an attractive contribution margin and EBITDA margin and even at EBIT level as well. We're already one third of the way to the 300+ midterm module target at the end of 2022. The modules that are either in sites already or already ordered and in build are around 8% of the target of 300. We're at this gray box here, already at this stage.
Having established the route to revenue and GBP 900 million of revenue, we thought we'd go into the costs that deliver that revenue. First of all, the direct operating cost, which will be familiar. This is made up of two components, the engineering team that are local to the CFC on-site, and then the cloud costs to run the OSP platform. In fiscal 2021, we reported this as a KPI for the first time at 2.7% of site capacity sales. Per fleet, current fiscal 2022 run rate is now 1.9%. We had a target of 2% for this cost base. Today, we're reducing that to 1.5%. Largely based on the real experience that we're seeing today in our warehouses and the Purfleet performance at the moment.
Revising the overall target to 1.5%, we actually do believe that number can go even lower than that. We could get it as low as 1%, for example. There is a line of sight to that sort of number, but at this stage, we're comfortable with the 1.5%. That's the key direct cost within the warehouse, the technology, the cloud costs, and the engineering teams. The other cost, which is a non-cash cost, is the amortization of the 10% of client sales, sorry, that are funded by the client as a capital contribution towards the CapEx build.
We amortize that over a 10-year horizon, so a cost of 1% per annum. Which is that amortization coming through, which is important when we get to the return on capital employed calculation. What else would I say on that? That probably does the job. We invested in our teams. These are our technology cost side of the Technology Solutions business model. This is the 2,600 headcount, grown significantly by 1,500 as we've over the last five years, as we've invested in the platform for the products that we have today and the products that we will have over the next two or three years as Re:Imagined rolls out. As a reminder, 35% of this cash spend makes its way to the profit and loss account, and 65% of it is capitalized on the balance sheet.
The 35% of it that goes to the P&L account is allocated completely to our Technology Solutions business. There are no dangling costs in this analysis outside of the three business models. We expect that to decline gradually from probably the end of 2023, from a GBP200 million down to a GBP 200 million per annum run rate. Of course, you can never predict what else might be on the horizon for new innovations or new products. As it stands, based on this business that we have today, that's the progression that we are expecting to deliver on and get to by the midterm. That lower cash cost for technology, a reminder of which 35% of that will go through the profit and loss account.
I'm gonna show what this business looks like when you put all the numbers together to get to both the contribution margin and an EBITDA margin. The other key cost line in the Technology Solutions profit and loss account is its allocation of group support costs. I'll start with group support functions down the bottom here. As a reminder, Logistics get 45% of this number. 55% of the number of GBP 75 million goes to Technology Solutions, and that's the finance team, the legal team, and the human resources team, and all their activities and so on. The client services and platform implementation costs are allocated to Technology Solutions. In very simple terms, these are the teams that are setting up the CFCs as they go live, helping with that rollout plan of live CFCs and supporting, particularly in the early years of a CFC going live.
Other is Kindred and other venture-related costs. Finally, the blue row here, the blue bar, is a solutions team headed by Luke Jensen that basically interacts with all of our customers around the world and prospective customers, either winning new business or ensuring that existing customers are getting the service and the returns from our businesses, particularly, clearly, of course, the technology solutions business. FY 2021 context, I'll let you just read that, but I think I've pretty much run through all those points. This is the key column, though, in respects of the guidance that we're giving as to how these numbers might change over the midterm. A question we get asked a lot, "How might central cost develop?" In very simple terms, we expect to pretty much keep these flat in real terms.
The one thing I would say on this one is that there is real conviction at Ocado, and if I say at the highest level, I think you will understand the point that I'm making to reduce these costs. My own view is that I think these are pretty lean as they are, but there is still a runway to go for here, particularly as, for example, in finance, we get the full benefit of new software and platforms we put in place. Last year, we invested in Oracle Fusion, for example. There's still an over-reliance on Excel spreadsheets, but we know that there is a route through a much better way of working, and that will enable a reduction in finance costs accordingly. At the moment, we compensate for the systems that we have or have had through headcount, as a mini example.
Real conviction around cost management and what we can achieve from these cost bases. Putting it all together, what does our profit and loss account look like for Technology Solutions? Well, again, here's how fiscal 2021 maps onto pro forma numbers from the actual fiscal 2021 numbers, a GBP 81 million loss, but a contribution margin already at 56%. By the midterm, we expect this to get to 70% of that revenue number of GBP 900 million that I shared a little earlier. There's the GBP 90 million of group support costs staying flat again. As a reminder, what was in there, platform implementation, group operations, and so on. Then technology costs. That's that piece that's non-capitalized from that total cash spend of 200. You get down to an EBITDA margin of around 50% for our Technology Solutions.
Of this number, around GBP 450 million of EBITDA by the midterm. A significant evolution in margin, and we believe the economics of the underlying economics of the business model really coming through. Just to stress again that these numbers, this number in particular, is before the benefits of Re:Imagined CapEx investments. On the left-hand side of this chart, we go through a very simple example for a 5-module site. Gross CapEx cost of circa GBP 50 million. Again, sorry to labor this, but that's the pre-Re:Imagined CapEx cost. We believe there's a number there that's at least 10% lower than that as we put through the benefits from those innovations. The next CapEx cost, that's after the upfront fees from the clients, GBP 36 million.
Most of the CapEx is spent in the two years prior to the warehouse going live. This chart on the right shows that for modules that go live of three modules and two modules, respectively, we spend three-quarters of the spend if we're building a three-module site at going live in the first two years, before go live. Go live is here on this chart. There's the point that I made earlier. At a steady run rate, phasing is smoothed, and we've assumed for the purposes of this model, 10 five-module sites going live per annum. Gross CapEx using that GBP 50 million number of GBP 500 million. Net CapEx of GBP 3,360 million using the GBP 36 million number. Making the point, of course, that not all CFCs are equal. They can vary in size, shape.
They can have specific features to a location, for example, seismic, particularly important in Japan, we've found, and cost that goes in there which makes the cost different. We believe that there's a rule of thumb. This is a good average for the purposes of this model. We go live with two-three modules of capacity, then we add further modules as the client builds into that capacity. We expect attractive returns. Purfleet is on track and at the moment, a 22% return on capital employed. That's at an EBIT level. That's after that amortization of that client contribution up front that I talked about a little earlier. Excluding the benefits of Re:Imagined, this number, we have a clear line of sight, though, to a number in excess of 30% return on capital employed.
Much of that is around those CapEx cost reductions, but also operating cost improvements. As the sites ramp, these returns will become evident in our financial numbers. Purfleet. Here's a photograph of Purfleet on the right-hand side. Just a reminder of the key characteristics of Purfleet. Just over six modules doing 85,000 orders per week. That's the capacity at maturity. An average basket of 45 inches. It took 22 months to build and ramp to November 19, 2019 to September 21, 2021. 42,000 orders per week run rate right now. It's the fastest ever ramp that we've seen for our warehouses. GBP 55 million gross CapEx and net of GBP 39 million after the client contribution, those upfront fees. As a reminder, over 200+ units per hour.
In fact, very close to that 220 number for Andover that we reported earlier, and expected to get better. A 22% return on capital employed. The return on capital employed is a combination of all sorts of things. Clearly, the upfront fees, the capacity fees on a run rate basis, the upfront fees from the customer. They can vary from CFC to CFC. The average that we've guided to is a sensible average. Just to make the point that not every CFC is the same as the other. In respect to Purfleet, though, a 22% return on capital employed. The CapEx summary. Here's how the CapEx looks for Technology Solutions. GBP 500 million in CFCs. Going back to that, 10 five-module sites per annum.
This is the capitalized part of that GBP 200 million of technology cash spend that we expect to get to in the midterm. Then there's a small balance of GBP 45 million. It was GBP 47 million in fiscal 2021 around supply chain, the capitalized cost pre-go live, and a variety across cloud platform implementation. Retrofit of CapEx again. That can happen going, for example, from the 500 bot to the 600 bot. That goes in there. We're guiding the midterm GBP 675 million.
With reimagined, we can bring that number down to 600, but still carry out that level of activity around those 10 5-module sites per annum. An important slide, because what this reinforces is that with our fiscal 2021 cost and contribution profile, and assuming on target cost base that we have, going back to those earlier numbers that I highlighted, whereas we required 80 mature warehouses to be cash flow breakeven for this business and still build those 10 live sites per annum. The targets that we've indicated today mean that that number is now just 50 mature warehouses. Which is consistent of course, with where we expect to be in the midterm, given those 58 contractual commitments. In other words, at that point, we will be cash flow breakeven/positive despite while still funding that CapEx rollout program. Really important point.
Clearly, there's a piece in here as well. There's still a piece of cash as the upfront fees from the ongoing capacity build, warehouse build that we receive from our customers. In summary, bring it all together. Ocado Retail, Logistics, and Solutions, the three business models, not reported business segments today, I would hope they will be very shortly. The three business models. In the medium term, a GBP 4.5 Billion revenue business, GBP 900 million revenue, and GBP 900 million in Tech Solutions. All built, we think, on solid assumptions and existing customer commitments. EBITDA, Ocado Retail, high mid-single digit EBITDA margin. We've done this before. We think we can do it again.
We see a line of sight through there, particularly as the business scales up and many of the costs in the cost line remain flat in absolute terms and then come down in relative terms. U.K. Logistics, a 3% margin reflecting the cost plus business model, and then Tech Solutions, that 50% margin that I highlighted earlier, and that's after all of those cost contributions that I highlighted, direct operating cost, but also that share of the group operations costs and the non-capitalized technology spend. You put all of those together and we get to an EBITDA of over GBP 750 million, and you can run the math on the percentages and the revenues to see how we get to those numbers. Cash flow, talked about. CapEx is GBP 600 million in Tech Solutions.
Plus, we've allocated another GBP 300 million to the Retail and Logistics business, which is pretty much in line with the sorts of numbers that we've seen in previous years for CapEx across those two businesses. Summary. The roadmap for growth. Sorry it's taken a little while. Hopefully the last 45 minutes has given you a better understanding of the underlying business models in Ocado Group and a clearer roadmap for growth. Principal driver, our technology solutions business, currently delivering OSP, the Ocado Smart Platform, to 11 clients globally. We expect to grow that number. This model is not built on speculative business wins. This is existing customer commitments. We are always in live talks with potential customers, several going on at the moment, and hopeful of getting one or two of those over the line.
Revenue growth from Technology Solutions, it's secure and visible. Existing customer commitments deliver a clear path to in excess of GBP 900 million in revenue. We made good progress in our target operating model. We believe the Purfleet example reflects that and gives us real conviction the investments we are currently making will produce attractive returns in the future. Ocado Retail is well-positioned to grow profitably in the U.K. We continue to grow customers, continue to get great customer feedback. We're extending our reach around the country with the investments that we've announced. Now we have a clear path to in excess of GBP 750 million of EBITDA in the medium term, that four to six years that I referenced, and to positive cash flows and strong returns while still investing in a rollout of CFCs. We're very excited about the future. Thank you very much.
Just gonna have a glass of water now. Do you mind if I sit down whilst I take questions, actually? Thank you.
Okay.
Sorry, give me a second.
Yeah. No, I'll speak up for the webcast. Absolutely. Andrew Gwynn from Exane. Thank you very much, Stephen Daintith, for the presentation. First question, just on the 58 commitments, I have sometimes trouble sort of reconciling exactly who they are, where the timeline of when they'll land. Is that something you can help us out with, or is some of it a little bit sort of top secret?
No, it's not at all.
Yeah.
In our fiscal 21 results pack, Chandler, perhaps you can share that.
Yeah.
Afterwards, we have a clear summary by partner of how those numbers add up. I think I'd just highlight that 40 of those 58 are across Kroger and Aeon in Japan. Then the balance across the other partners that you should be familiar with. That provides a sort of timeline chart for by what point we expect or and/or the customer has committed to roll out those CFCs.
Okay. Second question. I'm being a bit boring, unfortunately, but is EBITDA the right measure for Ocado? I've asked this question before to Duncan, unfortunately.
Oh, yeah. How did he answer?
The second question then is sort of related to that, which is what's run rate depreciation? It's a very boring question, I know, but unfortunately kind of relevant if we're thinking about underlying free cash flow generation.
I mean, I'm a fan of free cash flow. I think what I like about free cash flow is it sort of takes away the noise in the accounting, which can be complicated. You can get working capital coming into it as a bit of a twist. There are certain games that people can play with that. As a rule, I quite like free cash flow. I think you sort of, you know, you hold your feet to the fire with free cash flow when you're generating cash and delivering on returns in a meaningful way. I think you might expect a bit more reporting around free cash flow and free cash flow evolution from Ocado in that respect.
EBITDA kind of does the job, but there are still a few puts and takes that get you to your free cash flow number. That's the one. I think then the other one is return on capital. I think, as we mature into our business model, particularly for Technology Solutions, I think having a formal cash return on invested capital number for the group to report against would make a lot of sense. We've got return on capital employed for individual investments, but I think a group cash ROIC would be a good move. There you go. Those are the two that I'd look at.
Okay. Last question before I hand over. One of the questions we get asked a lot at the moment is about funding. Obviously there's a profile which is a fairly large chunk of CapEx to come before the EBITDA comes a bit later, and there's potentially a bit of a gap in terms of liquidity. What are the thought process around that gap?
First of all, we have healthy liquidity. We started the year with GBP 1.5 billion of liquidity. We've got liquidity today in excess of GBP 1 billion . I would also add to that, we have shown that in the past we can access the debt market successfully with the recent bond that we did in the autumn of last year, that GBP 500 million bond. We've had two good convertible bond issues in the last few years, and then we have a strong, supportive shareholder base as well. We don't believe that we're short of access to capital. I think what our shareholders find attractive about us are the returns that they see that we can generate going forward.
We're quietly confident around our ability to finance and put the capital behind the investment profile that I highlighted just now.
I think I'm quite clear about what you said about the OSP recurring fee as being a mid-single digit percentage of your partner's revenues. What I'm less clear about is how that might change, if it does change, as you become more efficient. You've talked about Re:Imagined, for example, or if the retail partner's basket changes in some way. Could you just explain how those efficiency benefits are shared with your partners over time?
It's a really important point. I mean, we acknowledge that the retail space generally is a high volume, low margin business, and that therefore our clients will have a very clear view of our returns and margins. As Tim highlighted with Re:Imagined, we expect to share those benefits with our partners and at the same time with our fees as well. If we, you know, when we start delivering those goals that I've just run through, I would expect us to have a grown-up conversation with our partners and because they would expect that of us as well, and that's what a partnership is all about.
I mean, we are going for volume, and it would be very easy just to stick with as we are and say, "Well, let's call it, you know, 58 CFCs does the job." We want hundreds of our CFCs around the world, and to do that, it may mean that we have to be leaner with the fees that we charge. Still attractive returns to win that volume, which where we think our product can play best around the world in the grocery space. So it's a very live debate. And as we really sort of, you know, we use reimagined to highlight for the first time our willingness to share with our partners, but with a goal to significantly grow volumes of warehouses around the world.
Thank you.
Hi, Nick Coulter from Citi. Three, if I may. Firstly, on the depreciation question, but from a free cash flow sort of angle. If you think about your maintenance CapEx as a percentage of GMV, for a 5 mod CFC in any given year, how would you kind of characterize that figure? 'Cause clearly with depreciation, you have the back end of asset life. You choose 10 years.
Yeah.
These things will run for probably more than 20. You include engineering costs already in your fee. Presumably maintenance CapEx is significantly lower.
It is. It's less than 1%. I mean, it's sort of, you know, 0.5% type number is what you should be thinking about. We'd only hope, for example, with the 600, with the new bot that we highlighted, that's gonna get even better, the durability of the bot that's 5 x lighter than the current bot. And, you know, we think it's, you know, we're gonna start to see a reduction there as well. Yes, a small number.
Okay. The line of sight on the 1% net cost rather than the 1.5-
Yes.
Sounds intriguing. What are the kind of the broad drivers? Is that kind of the modular nature of the robots? Are engineering costs come down? Or how is this?
It's a mixture of things. I mean, the big outperformer that I haven't talked about is, in fact, the cloud costs. Cloud costs have reduced most dramatically from that direct operating cost base. But I think the more we see of our engineering cost base and their activities from live warehouses today, the more we know we can do with smaller teams. I think the longer we see the durability and reliability of our products, the more we can dial down that engineering resource. That gives us the conviction around that 1% number.
Correct. Thank you.
I should stress, by the way, that's not formal guidance at this stage. It's the 1.5% that I've highlighted. My point was, we have clear line of sight to that even lower number. We'll be public and formal about it at the right time.
Great. We'll take that as an aspiration, then. On the logistics business and kind of the 3PL-
Yes.
Fade and the margin fade. Could you just run through that again, please, just in terms of how you kinda keep a flat profile as the revenue increases? Obviously, there are timing impacts.
Yeah. It's a 4.34% margin business, 3% margin business. The reason for an elevated margin currently in fiscal 2021 is because there is a recharge in there, a capital recharge to retail from logistics, for effectively the use of the Hatfield, Dordon area sites. In fact, Andover and Bristol are included there as well, but they're just less significant to the number of GBP 40 million per annum. That is effectively a wind down of the net book value of the capital base of those assets. Clearly, that gets to a point in time where you've exhausted that net book value through your capital recharge, and it reduces to the number that I've highlighted today of being much less than half of the 14 that's in.
You kind of lose the benefit of that margin over time.
The 4% management charge stays in place, does it, or do more of the kind of 3PL responsibilities fall to the JV over time?
No, the 4% stays in place.
Okay, great. Just one cheeky follow-up, if I may. On the margin for the JV, you've alluded to a 6% medium-term target. Presumably, that's excluding the kind of toward 200 basis points that would come from Reimagined for the proportion of the JV with Reimagined.
Yes, it. You're absolutely right. It does. You could say, well, Ocado is not dissimilar from sort of any customer. As those innovations go in and there is improvement in operating performance, looking to what extent we share that with our partners.
Great. Thank you. I'll pass the baton.
Stephen, thank you for the presentation. A couple of questions from me. The first one is, as you outlined, Profit is already doing 22% return on capital, and you're expecting that to grow to over 30% over time. What thought have you given to actually funding these projects with infrastructure finance? Because the spread is quite, you know, annualized returns on capital of an excess of 30%, you would think that you could finance each project individually or as a group for Kroger or whomever. The second question, given that your partners, I would imagine almost all of them, have a cost of capital meaningfully lower than yours, have you thought about vendor financing or partner financing, if you will?
Thank you, Alex. The short answer to both those questions is yes. First of all, project finance, I've looked at this in quite some detail, actually. It could be a very attractive thing for us to do, because effectively, we could secure a line of credit for the CFC rollout through project finance. There are just the considerations to take into account are that you would probably structure these on a client-by-client basis, and that the other side of the investor of the project finance would be looking at the financial security of the retailer. At the same time, the retailer would need to be comfortable with the third party investor involvement.
These are considerations that can be accommodated, but I'm just making the point that it's not simple and straightforward. It requires the cooperation of the retail partner as well. Indeed, it may be even commitment around the ongoing operation of the warehouse. No reason why that shouldn't happen, but I guess the point that I'm making is that it is not straightforward financing. There are issues attached to it, but it's very doable, and I believe it is an attractive thing for us to look very hard at. It's a very live project for me and has been.
In the instance that most of us, I suspect, anticipate that you will sign increasingly larger orders with your existing customers, including, as you pointed out, both Kroger and Aeon are responsible for 40 out of 58 facilities that you have on contract. The next tranche, let's say Kroger comes out and says, "I want another 20, 30, 40 facilities going forward," you can actually attach that.
Totally agreed. That's exactly the point because.
You can attach the infrastructure financing to that specific 40 CFC project.
Completely correct. You'll really, you know.
Pointing to the logic for doing it because.
You become a spread business.
Of course.
You're collecting 33% against it.
Of course you do. Because the alternative route, of course, is you're in a constant world of refinancing and either debt or equity calls, and that's probably not a place you wanna be for a growth business. Project finance has its appeal. I think you can understand.
Thank you.
Good morning, everybody.
Hi, Stephen. Rob Joyce, Goldman Sachs. I've got-
Hi, Rob.
How are you doing? I'll go through three. First one, just to clarify in terms of that definition of medium term or midterm, sorry, you said you're expecting GBP 750 million EBITDA and free cash flow breakeven in 2026-2028. Is that the range we're thinking?
No. Well, yeah, I mean, 4-6 years, that's the sort of range that I'm thinking. Yes, that's a good guide.
Okay. In terms of the, I know you said you definitely wanna build beyond the 58, but just to build on Nick's question about the sort of maintenance CapEx levels. If we take the GBP 100 million or so on the retail business and then take, I think you said 0.5% of the 900. We're looking at GBP 150 million, maybe GBP 200 million CapEx as a sort of steady state at that basis. Is that a fair way to think about it? I know it's not the way you're looking at the business, but would that be a fair way?
No, it wouldn't be that high, actually. I mean, it's sort of I think it you know probably I mean, it was certainly not. We haven't hit over GBP 100 million, for example, on maintenance CapEx so far. So, I feel pretty comfortable that we're talking about a certain number that's sort of between GBP 50 million-GBP 100 million of maintenance CapEx.
Oh, okay.
Clearly that number varies according to the number of warehouses.
Okay. The third one would just be on inflation. Haven't really talked about that, how that feeds into both the payment you get from your partners. Is it volume linked or is it linked actually to the value of sales?
No, there are inflation clauses in all of our customer contracts that protect us from inflation increases in all of our customer contracts.
Okay. It's both on the revenue side and the CapEx side.
Yes.
Okay. Thank you very much.
Hi. It's Simon Bowler from Numis. Two hopefully quite quick ones. I think you said at the start that you're thinking about moving towards this kind of new structure of reporting, maybe by fiscal 2023, definitely by fiscal 2024. It seems a sensible way to kind of resplit the group. Why does it take that amount of time to move the group in that direction?
I'm smiling at the one or two of my finance colleagues in the room here at the moment. The finance machine at Ocado is not what I would want it to be. I don't think anyone who works in finance would want it to be. I mean, it's probably an area that could have been invested harder in over the years. I think the Oracle Fusion implementation last year, we went live in September, is a big step forward for us. We're now bedding it in. We're putting in place a reporting tool around it, and we're expecting significant benefits.
I'd love to do it in fiscal 2023, but I just want to be sure that our finance team work extremely hard as it is just to make sure that we're in the right place to be able to do it in fiscal 2023. Let's see what we can achieve.
Okay, cool. Then the second one was, I think you've kind of set out very clearly kind of your side of things from a CFC perspective, but I wonder whether you'd give any additional color on, say, the Purfleet CFC and what kind of contribution margin you think from that facility with its UPH you're gonna be able to achieve relative to a Hatfield or a Dordon. Like I say, looking at contribution margin level rather than some of the other KPIs you give.
I mean, I'm not gonna be completely explicit about the contribution margin. I just think we can do better in Perfect than the numbers that I highlighted, given where we are today and the trajectory that we're expecting around direct operating cost. Again, it goes back to my comment around that 1% number that we believe that we can get to, in the relatively near term. Hope that answers your question, probably doesn't.
That's why I meant more with, like, your Ocado Retail hat on, to the extent that you have one.
Oh.
In terms of how their contribution margin fits through rather than your side of it as Ocado Group.
Well, we would hope that, you know, that's really one of the key enablers to get into that sort of mid-high single digit EBITDA margin, that I talked about. I think, you know, that's one of the areas today that is a significant opportunity for us in the efficiency of our platform and bringing down their cost base 'cause they're right now I think, you know, they could benefit from that in light of the inflation they're seeing in the utility cost base. Does that answer your question? I'm not sure it necessarily is, so.
That's okay. I'll follow up. I'll take it with your finance.
Fine. Okay.
Cheers.
Okay. We've got 2 more in the room before I go to.....
Stephen, thank you for the presentation. Sreedhar Mahamkali from UBS. If you're not gonna move to this way of reporting, which clearly is very helpful for us and for a lot of investors, I'm sure, why are you doing this now, as opposed to at a later date?
Well, the reason why we're doing this now is that, I thought it was a piece that was missing in the Ocado narrative around the midterm trajectory. I think, a lot of our reporting is currently on sort of, you know, actual reporting of current numbers, and I think. For a business like ourselves, which is an investing for growth business, I think it's important for our partners and our investors to have a clearer understanding of why we're investing and what the returns can look like. As I thought through that, I thought the most simple and effective way of demonstrating that was this resegmented look at, this new way of looking at the group.
Intuitively, when I arrived new and started looking at the accounts, it was actually quite difficult, I found, to understand the business, each of the businesses. The U.K. solutions and logistics business in particular, was complicated in respect to having those two different business models together. There's probably very good reason at the time why we did it that way, but I think this new way of looking at is much more intuitive and iterative as well. We hope to get to this way of looking at as soon as we fiscal 2023 if we can, but certainly by fiscal 2024.
A couple of other short questions, hopefully. In terms of the moving parts to driving the contribution up from 56% to 70%, what would you say are two or three of the key drivers? I think you've referred to one at least.
I think, the ongoing evolution of cloud costs that I've just highlighted, we expect that number to come down as, you know, that commodity to reduce as a cost for us. The engineering piece is clearly a big driver. We haven't yet talked around or even really looked at the extent to which we can have virtual engineering, monitoring of performance rather than on site. We expect the 600 bot to be more reliable and faster and cost less to maintain than the 500 bot, which was indeed improvement from the 400 bot. Those are all the. It's a variety of things that get us there. Again, that's before Re:Imagined on-grid robotic pick will reduce labor cost in a warehouse by over 30%.
That will go live within the next sort of. We'll start testing it live in our warehouses in the next sort of six-nine months, and picking around 70% or so of the shopping basket in a robotic way, and drawing on the Haddington and the Kindred technologies, those two businesses we acquired at the back end of 2020. That's probably a simple summary of the biggest drivers of that reduction in direct operating cost.
You, you've captured the 600 series bot in this?
No, sorry. That's not in that 1.5% number.
Okay. The last one then, in terms of trajectory, just back to Rob's question, 4-6 years to get to that EBITDA. Do you see this more as a straight line through to that cross? Are there any big sort of ups and downs that we should be aware of?
I think we're probably gonna have some years when there's gonna be more CFCs going live than other years. I don't think it will be necessarily linear. I think the economics and the progression and improvements in the economics will be pretty linear. It's really down to site by site, location by location, evolution and how that might phase. You shouldn't expect big. You know, it's relatively linear, but it, you know, but it's a channel.
Thank you.
Thank you. William Woods from Bernstein. The first question is just on the modules. Obviously, you've got some very large Kroger CFCs coming live and some much smaller ones later down the pipeline. What's the scalability of the economics in your mind based on modules, and how does ROCE change?
In very simple terms, the larger the warehouse, the better, because there is a degree of fixed cost here and the bigger the site. On the other hand, though, I think the versatility of the model that we have, the fact that we can put our the OSP in Zooms, which are half a module and 10 module sites, is terrific for our partners because, for example, well, Kroger are very keen to explore a variety of different models. The immediacy market is an important market in the U.S. This is almost certainly gonna be in the next, you know, short period of time, Zoom's going live in the U.S. Many CFCs as well.
We've experimented with those and worked those well in the U.K. with Bristol, a smaller CFC for a smaller density, population for smaller populations. That's the attraction of the model. Where there's still attractive returns, albeit not as great returns as the large CFCs, they're still very attractive returns to go for. They're a sort of, you know, market, you know, highly competitive in the market in terms of capital investment and returns on investment. I think, you know, going for volume for us is incredibly important with high returns and being prepared to accept a slightly lower return, but high volume of many CFCs is not a bad thing to do.
Just on the 10 CFCs per annum that you've kind of given as midterm guidance, that seems feasible until FY 2025, when you kind of drop down in the number that you've given to probably 4 or 5 per year. How confident are you that you're gonna get some more orders in the next six months to make FY 2025?
Okay. I think given the talks that we are in, I feel quietly confident that we will be announcing new client wins this year, not just the one. Watch this space on that one. I also feel quietly confident that we're going to win additional orders from existing clients. Again, there's work for us to do, the work for our clients to do. They're getting used to the CFC experience, but the customer relationships are strong and we think we can grow customer orders from existing clients as well.
Great. Just one final one. Just on the last accounting seminar that you did, I think you talked about EBITDA margins at a single CFC getting to 60%-70%, and you've obviously guided to 50% midterm. Is that just because of the mix effect of immature CFCs at midpoint?
Yeah. I'm not familiar with those materials at that previous seminar because that was before my time. I would. That sounds a plausible reason. Chandler, do you have a-
This was the 2020 seminar?
More than the 2018 one.
Oh yeah, that would be it, probably a nice question since 2018.
I think that's the final question from the room. Yeah.
Yes. Some of these have been answered already, but try to do the ones that haven't yet been. Given the slowdown in online sales and the rapid increases in inflation, are any of your partners slowing down their commitments?
Short answer is no. I mean, they're all rolling out on their commitments at this stage.
Where do we see the difference between gross and net CapEx come through the cash flow?
Gross CapEx is what we include within our reported CapEx guidance. The net CapEx, the cash receipts from that, you'll see that cash coming in every year as fees received, and that sits on our balance sheet as a contract creditor, although I prefer to call it accrued income. It's accrued income on the balance sheet waiting to flow through the profit and loss account. Once the CFCs go live, it starts, and then it's effectively amortized to the P&L account as a credit, as revenue through the life of the CFC.
I think some of these have already been asked. Yeah. How are the discussions going with existing customers as it pertains to the Re:Imagined tech? Any thoughts on how these improved economics get shared?
I'm not gonna go into explicit detail. They're on a customer by customer basis, as you might imagine. I think I've probably said enough to say that we acknowledge that there is benefit in sharing those returns with the benefit there with future volumes, additional volumes.
Just final one, which I think was kind of on Aeon, which is just the number of CFCs that Aeon have a firm commitment and the breakdown by year, and when they're expected to be rolled out.
Crikey. I don't have that breakdown by year, but I will let you know that the 20 needs to be built by 2035. That gives you a profile, quite a long time horizon to get those done. We've got the one, and I think there's another one, but we haven't announced the other one yet that's close to being secured.
Yeah.
Okay.
I think that's a little over time.
Sorry it's taken so long. Thank you very much, everybody, and I hope that's been useful. Hopefully we'll be reporting formally this way, in the not too distant future. Thank you.