Good morning. Welcome to you all to the half year results for Ocado and financial year 2023. As you will have seen from the RNS this morning, we have made good progress against all of our strategic objectives over the last 6 months. Tim and Stephen will focus this morning on the real tangible results that we've made and concrete progress against those strategic objectives. In particular, we're going to focus on 4 areas where there's been particular discussion and debate in the markets. We'll look at the operational and financial results, and in particular, driving EBITDA, driving cash flow, driving down costs.
Second, really, I know a lot of you have interest in the Partner Success Program that we've been driving to make sure that we can help our partners improve their results and what they get out of OSP. Then, of course, Ocado Intelligent Automation prospect of deals in the non-grocery sector, and then the recovery plans and the program for the Ocado Retail results. Tim and Stephen will cover those in a few minutes. You know, I've often spoken about really, Ocado being a cauldron of creativity, and I say that it still remains very much that.
I think as we work with our partners, help them deliver for their customers a gold standard online grocery, I mean, it really much is part of that, just as is a culture of co-collaboration. As I look around the company, it's very present, very necessary, particularly as we really think about how to draw on this pure play of online grocery, where we have an enormous amount of experience. When we look back at that, Tim's going to focus on this a little bit from the outset, I mean, we are in the business of solving complex problems. We're in a high volume, low margin sector. We just got to do this in a very smart way.
You know, 20 years of knowledge and experience in this sector is coming through in particularly in this Partner Success program. I think that's going to be very key to taking the results further and delivering what we promised you. With that, I'm going to turn over to Tim and Stephen. Over to you.
Thanks, Rick. Okay, I'm going to first of all, just talk very, very briefly about our technology. As Rick says, we're in the business of solving complex problems for the largest grocery retailers in the world and beyond. Our solution today is without doubt, the industry leading end-to-end solution for grocery online. In terms of its levels of productivity, in terms of its end customer results, the NPS scores that we get from our clients' customers are outstanding. Our technology is applicable to all sectors outside grocery, to anybody that needs to receive, store, sort, and ship goods. It is driven by extremely advanced technology, and R&D in many areas, including artificial intelligence, machine learning, robotics, and vision systems.
Today, we are around 2,700 people, above 2,700 people, who are constantly working on this innovation and helping our ability to innovate at pace, and their productivity is accelerating at the moment. We talk a lot about IP and a lot about patents in this business, we have a lot of unique and proprietary technology. Just to give you an update on those types of stats, our technology is currently protected by over 900 granted patents. Almost, there's another almost 1,500 filed patents behind those 900. Those are quite staggering numbers of IP protection out there. Almost 300 of those are in the areas that I know is a great area of focus at the moment of artificial intelligence and machine learning.
Almost 300 of those, 2,400 are in that area, and it's obviously a constant, constantly growing, and we're very successful in this space. Just talking about the overall H1 highlights on financial progress, as we've talked about already this morning, we're seeing growing revenue, positive EBITDA, and most importantly, improving cash flow. Stephen is so strict with us at the moment, which is great, and it's a discipline that he has my 110% support on. I've been delighted that he's picked up the mantle so much.
He's already informed me that I have to get all of you out by 11:00 A.M., because that's how long the room has been paid for, and every minute you stay after that, it's going to cost me more money, and he's actually already told me that. Our operational progress is great, and we are driving Partner Success. This is the one area of center that has been growing this year. It's because we have realized that our partners need more help to create the success that they can from our solution. We're seeing strong growth in live modules, we're seeing improved productivity, and we're seeing cost efficiencies.
In terms of the high-level strategic objectives, Ocado Intelligent Automation, the proposition, they're getting out there with that proposition and creating the business model, and pursuing the client opportunities that we've already seen, and progressing those towards signed contracts is making great progress, and we see a very large and very attractive market opportunity for Ocado Intelligent Automation, and Mark and his team are exceptionally busy. It is a growing global team. Our H1 performance gives us confidence for the full year, so our financial guidance is unchanged. Just looking at some of the numbers, obviously, Stephen will take you through them in more detail. Our revenue in the H1 is up 9%, to GBP 1.371 billion.
Our EBITDA has gone from a loss of GBP 14 million to a profit of GBP 17 million, an improvement of GBP 31 million. Loss before tax is at GBP 212 million, down GBP 8 million. The underlying cash flow, which I think is an important number that I know everyone's been focused on, is an improvement of GBP 108 million, to negative GBP 288 million. The revenue up 9% is really coming from the underlying growth in Technology Solutions. There we've seen revenue grow by 59%, in Logistics up 2%, and Ocado Retail up 5%. In terms of EBITDA, what you've seen in Technology Solutions is an incredible conversion. We'll have a look at that in more detail in a moment.
The increased contributing margin and going EBITDA positive for the first time. It's worth remembering, it's a relatively new business. Technology Solutions opened its first international client, CFC, only in those first months of COVID in 2020. It's a business with an enormous amount of development behind it, 2,700 developers working on code. We, on average, put about a third of that straight through the EBITDA in terms of operational costs. To see that business turning EBITDA positive is good news. Ocado Retail is delivering improved profitability through the period, and Ocado Logistics is continuing to deliver stable returns. As I mentioned before, we've got great financial discipline, a very strong focus on costs.
We're being very good at delivering tangible results in this efficiency space, and we are running this business on cash. We're very much not-- we know we are looking at EBITDA, we are looking at profitability, but there is an incredible focus on cash that this business did not have before, and was probably over-focused in the last few years on EBITDA and not on cash. As a result, we've got GBP 108 million improvement in cash flow, which is both because of cost reductions, but also because of lower CapEx. Our liquidity remains strong. We have cash of GBP 1 billion, and liquidity of GBP 1.3 billion.
If we look at the main business sector of Technology Solutions, we've got an increasing in live modules, is what's driving the recurring fee growth. We can now have 25 sites live around the world. We now have more sites live internationally than we do in the UK. We now have 105 modules are now live and being paid for around the world. That's a 35% increase in the average live modules. I think the key there is what that's doing in terms of revenue growth. Revenue growth going from GBP 125 million, you can see on the right-hand side, they're up to GBP 198 million, 59% increase.
If we look at that absolute numbers, that is GBP 73 million in absolute numbers of revenue growth, but that GBP 73 million is dropping down to a GBP 65 million increase in EBITDA in that segment. You're seeing an incredible conversion of that increased revenue into EBITDA, which is both because the business naturally has a high conversion rate. Once you get to zero, it has a high conversion rate. Secondly, because of the improvements in operating costs of those of those 105 modules around the world. We've seen a drop-off from, I think it was 2.1%-1.8% from memory of those modules.
High operating leverage as the top line grows and the reduction in operating costs and lower support costs. This is really reinforcing the business model and hopefully giving you a guide as to what happens as we continue to grow, as we continue to roll out more modules for clients, and as clients continue to draw down modules in their existing facilities as they grow. That really highlights why it's so important for us to help our clients to achieve the most from those sites so that they can grow their volumes and then draw down more modules in the live sites.
The one area of business that we've been putting more money in, and I think the number was GBP 3 million in the last half alone, in terms of the increase in operating costs, is Partner Success. I think it was a mistake early on to believe that if we brought these solutions to our partners, our partners were going to be brilliant at running manual, you know, running facilities with hourly paid staff inside them, at driving vehicles around, at marketing, retention, supply chains and stuff like that. They are amazing at doing that in their existing businesses, but it's all subtly and slightly different when you try and apply it to online, and therefore, we're out now helping our clients in these areas.
We're working closely with our partners. We're embedding our 20-plus years of experience and know-how into our partner operations across a number of aspects of OSP to help them get the very best out of the platform. We have a combination of on-site, in-region, and central teams in Partner Success. We've got global excellence teams based here in the UK that are supporting on-site and regional teams working directly with our clients, and sometimes those centers of excellence teams are actually traveling to our clients as well. I got on a connecting flight a couple of months ago in the US.
I was sitting, they were on a small regional jet, so I'm sitting there like this, and I can see this group of people coming in, and they were looking at me, and I'm thinking, They've all got British passports. I think I recognize a couple of them, and it was a supply chain team from Ocado on their way out to Kroger. Who, who did extraordinary things in helping reduce the waste that we're going to talk about in terms of Kroger's, both high, you know, high levels of availability, low levels of waste, by just teaching them how we operate things and helping their supply chain teams to create better results. It was a coincidence to bump into them on a flight out to Cincinnati. The main areas of support are in fulfilment.
That can be in areas such as labor planning and warehouse processes, in last mile, in how you know how many sites you have, where your sites are, how you draw up zones, how many vehicles to put in, when to release those vehicles and not release those vehicles, how you plan it, how you route it, the proposition that you might offer in those geographies. Lots of area in last mile. Last mile is the biggest operating cost for all our clients after the COGS. And after the costs of the groceries that they sell, last mile is the big number.
In e-commerce, whether that's in monetisation, i.e., how do we get the media platform moneys that we get in, about 4% of sales, about GBP 100 million a year at Ocado Retail. How do we generate that money? Whether that's about how do you do marketing and acquisition, how do you do nursery journey, taking those first-time customers and turning them into loyal customers, different acquisition channels, and whether it's about merchandising and various other things there as well. What's interesting when I look at our clients is they don't all need as much help everywhere, right? Some need more help in customer acquisition. Some are doing great in customer acquisition and need more help in their operations.
Some operate the warehouses brilliantly but need some help in the last mile. It's slightly different everywhere, which is how we're developing. This will improve the economics for our partners, and as a result of improving their economics, will support our long-term growth. I think a few people have asked the question, which is: Is this just another big cost line that we now need to think about in the business? I think the way that we're thinking about it is this is an area that's going to operate very similarly to most enterprise software and enterprise hardware businesses, right? With an amount of sales, you get some free credits. Your free credits, you can use them wherever you need to use them.
If you need more help in the warehouse, use your credits there. If you need more help in understanding customer acquisition, use them there. Beyond that, you can buy more help. Today, our largest client, who is receiving the most help and has, at times, a dozen or 2 dozen people actually live on the site, helping them, they pay for that assistance. Don't expect some kind of large, new expense area within inside the business. Just to talk a little bit about an example of this, we've been running some programs with Kroger that are known between ourselves as Project Endeavor. Working very closely with Kroger and their teams to drive operational performance. We set out initially a single 12-week program.
We are currently in the 2nd 12-week program. The initial focus was on CFCs, what they call one and three, that were actually at the 1st two that opened in 2021. That's Ohio and Florida. We completed phase 1. We're on to the 2nd phase. We're cementing the initial progress and driving it forward. We're addressing other additional opportunities, and Kroger now, and we're supporting Kroger in working out and working out how they can successfully apply those learnings across the other live sites, and how they can make sure that those learnings are taken into account in the sites that aren't yet live. There's two Krogers in build due to go live.
Just to give you a highlight of what that first program achieved, units per hour improved by more than 25%. That's the key metric in the warehouse. How many units do I ship from for how many labor hours I pay for? That's an extraordinary achievement in a 12-week program. Drops per van improved by 10%. That falls straight down to your bottom line. Food waste, as I mentioned before, whilst improving availability, food waste went down by over 30%, and operational cost reduction of over 15%. There's a lot more to go after, some of those stats are starting to get pretty good. Overall, with growth, they will show and help Kroger to get to a very attractive operating metrics.
That was the first Project Endeavour. Looking at our own operational progress here in the UK, we're seeing improving productivity in warehouses and in our delivery stats. Let me just talk you through those one at a time. If we start off with the OSP UPH, that's the UPH in the OSP warehouses, up to 206, up 14%, continues to improve. I think some of the most interesting stat I've seen there, actually, was we started getting Bristol, which is our smallest warehouse.
It's operating at about 20-something thousand orders a week at the moment, low 20s, and we've got it up at 215, 216 UPH, which is really something to show our partners what is possible to do, not just in a 6-module warehouse running at 80% of productivity. What's possible to do in a 2-to-3-module warehouse or a 6-module warehouse, frankly, when it's operating only using, you know, kind of 2 modules or less, that it's possible to achieve these levels of productivity, and that's before Ocado Reimagined. That's before we take the benefits of AFL, automatic frame loading, of On-Grid Robotic Pick, and the other enhancements. There's a lot more enhancements to come.
There's a lot of things we can do to drive this productivity even higher in those warehouses, but we're seeing a strong continuing trend. Now, what you can see on the other side is a bit more of a mixed picture there, where you can see that we hit 200 drops per van per week, coming into COVID, which then dropped down to 175, and has only improved now to 182. It's running at about 189 at the moment. The exit rate there where, there we are, is 194.
What's happened there is actually that stat is made up of 2 numbers, but publicly, for the last, 15 years or whatever it is that we've been a public company, decided to try and condense it and just give 1 number, the drops per van per week. The more important number actually is what we call internally in our business DP8, which is the drops per. The drops that we make for every 8 hours of driver time that we pay for, every full shift of driver time. The other stat there that's critical is how many shifts and how many times the vans go out every week. What you're actually seeing there is we are at a record DP8, so the underlying productivity is at a record.
We have more vans than we need at current volumes. That means that the van turner is not as fast as it would be, 'cause we have vans originally for intended higher volumes. We will see that number continue to rise and exceed the 200 peak that we've seen before. As I say, we're up at well over 22 in terms of the numbers of drops that a driver does on their shifts. That together with overall high-level tight cost control on the other costs in the warehouses and the central costs and the operating teams, is driving improved performance. Moving on to Ocado Retail. We're making really strong progress.
That's not to say it hasn't been challenging, 'cause obviously Ocado Retail had a huge surge over COVID. It was extremely profitable over COVID. A number of its suppliers started operating at service levels that would never have been acceptable pre-COVID. A lot of experience got lost in the whole industry generally. We came out of COVID not in as strong a position as we wanted to be in. I think the key thing that Hannah has been doing with the team at Ocado Retail is a back to basics, is perfect execution. We are seeing really positive results from that already. We can see it in the customer perception. We're seeing markedly marked improvements in our customer's perception across value, service, and range.
Our, our value NPSs have improved materially since Hannah reintroduced a price-matching scheme that we had successfully ran for over 10 years before it was turned off. Turning it back on again and pushing it to customers has really been reinforcing our value credentials as well as the actual hard investments in margin, in pricing. You can see our inflation numbers at in the 8%, you know, kind of totally different to what we read in the newspaper about, you know, high teens inflation in food. We're running at less than half the stated inflation in food in terms of our average selling price, and only a small amount of that is to do with product shifting.
The majority of that is that our prices are not rising at that level. We've got very, very strong customer metrics, but at the same time, we're improving the economics. At the same time, as we are increasing our volumes QoQ, and we have got our eaches up, and we have got all the customer numbers in terms of retention up, retention in terms of conversion from first shop to fifth shop is at record numbers at the moment. We have improved the operational leverage. We have brought down the break-even point per week, and we brought that down through strict cost discipline that was not applied as it could have been during COVID, when it really was just trying, just in.
Take on any cost you can, as long as it will improve volumes, because it was just all about volumes, volumes with unlimited demand in that period. We've had to kind of go back in and imply a lot of cost discipline into that business, which has driven down the weekly break-even points and leaves us in a good position for the rest of the year. Productivity across the board is up, and up to 212 in Q2. We have a very clear pathway to full-year positive EBITDA, and we also have a very clear pathway to getting back to the kind of, high single-digit EBITDA margins that we expect that business to generate in the long term. It's a plan that we've all bought into. It's a clear strategy. We agree it.
We agree with the management, we agree with our, with our joint venture partners. I think everybody knows where this business is going. Just looking at some of the impacts of COVID and the operational progress, what you can see here, just to kind of understand the numbers that we've been talking about, you can see the basket size. It was running in at GBP 46. Into COVID, it went up to GBP 57 for a half. It actually peaked at over GBP 70, and that was basically some of a subset of our existing customer base that wanted to spend a lot of money, increase frequency, and increase basket size from us.
What you could see with those increases in sales and the eaches shipped, if you see eaches shipped as the green line here, that went from here and shot up to there, in theory, above our capacity, was driven by a combination of increase in frequency and basket size, but actually meant that our active customers had to drop that far. Right? We saw a steep drop in active customers. We couldn't let them in the shop. We made a conscious decision to give the capacity that our best clients wanted from us. You can see the challenge that we've been in now, which is basically saying that we've lost three halves of growth, right?
We lost this half, this half, and this half of growth. If it hadn't been for COVID, and you took those out, and you shifted all this back, and you kept going, we would be at the numbers that we wanted to be at now to generate the profitability we want to be at now. It's just a question of keeping going at the moment, and I think that's the critical thing to understand. Management made decisions. They were probably the right decisions to make at the time, otherwise, you disappoint everybody. We just basically had a problem that our most loyal customers wanted to spend between 50% and 100% more money with us than they did before. They didn't want to shop anywhere else. They didn't go on holiday. They didn't eat out.
They had nothing else to spend their money on, they wanted to buy food, buy lots of it. They wanted to bake and not even eat the cakes, in some instances. You know, so we had to serve them, but it has caused us some pain coming out. What you can see on the far right is, in the purple bars, is the underlying theoretical capacity of our warehouses. We came into 2019 with a bit of capacity. By 2020, in the early parts of COVID, we were actually shipping out more each than the theoretical capacity of our warehouses.
You'll remember I explained how we did that at the time because we were able to ship on a Tuesday what we had the capacity for to do on a Friday, and normally Tuesday is smaller than a Friday. By doing seven Fridays in a week, we were able to ship more capacity than we theoretically had. We built an added capacity. You can see that in the purple bars going up. At the same time, as we were seeing this effect and the slowdown in frequency, meant that overall we were shipping less each than we were at peak COVID, but more than we came into COVID with. That's the green line over here.
Now with the closure of Hatfield and the opening of Luton combined, we end up on the blue line and will be over 75% of capacity utilization at the end of the H2, of course, that significantly helps ORL's profitability. It is still bearing costs. It is still bearing cost to group for the infrastructure in its current contract on Hatfield. It is still bearing costs in terms of property costs for Hatfield, nevertheless, it still is a material improvement in its profitability, will be better once it's kind of disposed of that site and bought some more capacity to replace the current contract it has with group. Moving on. Great progress in Ocado Intelligent Automation.
Not yet, signed contracts with names of clients that I know that everybody's waiting for, but that will come. Nevertheless, let's just be clear what the model is. The model is a capital-light model. Fees up front to offset the cash outflows. Obviously, those fees are not recognised until the sites are delivered, so you won't see recognised revenue until sites are delivered. Thanks to my friends at the IFRS board, but you will actually see the cash coming in advancing as the contracts are signed and as we build out those facilities. We don't expect it to need to utilise balance sheet. It leverages our existing OSP technology. There is a limited extra R&D investment that's ongoing at the moment.
Obviously, based on the success of that business and the requirements of the customers, we'll decide how much that investment should be into the future. As I say, the profit is only recognised at the go live of those at those contracts. We are talking, as I said before, to several prospective clients in a large range of industries. As I say, it's really anywhere where people need to receive, store, or even manufacture and then store, sort, and ship goods. Our OSP Hives are extremely space efficient. They are capable of extremely high throughput, and they are capable of being extremely efficient from a pick, sortation, and storage perspective. We think unmatched in their efficiency. There are a lot of markets outside non-food.
A lot of people always say, "Oh, it's general merchandise." I think the view of general merchandise online retail is a small, or not small, but is a fraction of that market. There are lots of other areas of store-based replenishment, of spare parts, of storage of other goods, of all kinds of things that, this is very interesting for. It is becoming increasingly interesting to. The market for this kind of automation is really growing very rapidly, based on global labor shortages in developed markets and the increasing scarcity of people to do these jobs manually. It's no longer just a question of trying to understand, do I want to put the capital in, and how long does it take me to get the capital back?
It is also, how do I make sure my business can function, given the fact that I am finding it harder and harder to hire people to do what was just something that I assumed got done in the back? Sometimes these businesses, it's not the most important thing. They don't think about it that much. It's not like our warehouses that have always been center to our business. It's the thing that replenishes the stores. It's the thing that gets the spare parts to the manufacturing facilities, and actually, if they don't function, it's a major problem for these businesses. We have very high productivity and very small warehouse footprints.
That, if I just thought about one of the projects that we're working on closely, they found us quite late in the game, so they've already commissioned a site, built a site to put someone else's solution in, which they're considering putting ours in instead. We just, to do the throughputs that they needed compared to another very high, very well-known brand name, kind of automation solution, we need a fraction of the space. It's now kind of we're not going to use the whole building if they put our solution in, which just go to show significant savings for clients going forward, as they can right-size those buildings to what they actually need with our very dense and high throughput solutions.
In summary, we have a H1 of clear progress where we're delivering results across all our key objectives. Our KPIs continue to progress across all segments. We're seeing improved profitability, we're seeing our cash outflows reducing. I think our H1 performance reinforces the outlook for the full year. Our priorities for the H2 remain operational execution for us and our partners, remain financial discipline, including getting you all out in 59 minutes, cost reductions and focus on cash flows, obviously, mark strategic growth on capitalising the non-grocery opportunities. Obviously, the Partner Success is underlying all of this. Partner Success, Partner Success, Partner Success.
I spend an inordinate amount of my time at the moment working with our Partner Success teams and working with our partners to help them to get the best out of OSP, to drive long-term, profitable business growth from their OSP sites, so they need more of them, so they fill them, and so we can grow our business. On that note, I will hand over to Stephen.
Thank you, Tim.
You've got less than 48 minutes.
Yeah, I'm on it. Thank you, everybody. Good morning. Right, before I get into the numbers, I just want to sort of say a couple of words about the new segmentation, because I think it's, there are three key layers, I think, of benefits to the way we're going to be looking at our, these, the business over the next half an hour or so. I think, and one more is well, one is a, for, I think for you, visibility around three distinct businesses with three distinct business models and the financial economics and drivers of each of those three businesses. I think previously, our accounts may have been a little more difficult to understand, given we kind of joined together one or two businesses, and we hadn't presented solutions business on its own.
I think, first of all, visibility, really important. I think secondly, and this is an internal one, management. We now have three distinct businesses. We have three distinct management teams now. We did not previously have a distinct management team for Technology Solutions. We now meet every Monday afternoon for between two and three hours, and we go through the business from top to bottom. Each management team has its very clear goals and priorities, and that helps enormously. In fact, it's a big deliverer of the third aspect, for me, which is important, is that cost discipline approach that Tim mentioned. Costs are owned in the right place by the right management teams. They're accountable for it. They have their budgets, they have their targets. It's a much clearer way of working within the organisation and already delivering results.
You can see those results in the half-year numbers. There's more to come as well. We've put in the appendix, a reconciliation between the pro forma numbers for fiscal 2022 that you saw in the various events that we held last year with the full-year results in March of this year, and now the reported numbers that you're seeing today. That's in the appendix, if you need to go through that when you see the comparatives. The financial review. I won't go through this slide in great detail because Tim's talked about many of the numbers up at the top there. The revenue growth, the EBITDA growth, 9% revenue growth, EBITDA improvement by GBP 31 million.
Below the line, depreciation and amortisation is probably the key number that gets you down to that group loss before tax, post exceptional. That number will settle down, I think, at around, sort of on a full year basis, between GBP 450 to GBP 500. That'll be its peak level as it starts to decline, as we gradually fully depreciate certain assets, depending on course, on CapEx rollout over the next few years. Exceptional items, just two items, really, in the exceptional items to talk about. There is a breakdown in the appendix.
First of all, the UK network capacity review that's taken place following the closure of Hatfield, designed to, first of all, introduce more OSP into the model in the UK, replacing the old Hatfield site through Luton, and a big driver of the profitability, the growth in profitability of Ocado Retail as we cover those fixed costs from that excess capacity that we've been running with over the last 12, 18 months or so. There's been a small reduction in the deferred consideration due from Marks and Spencer as well, of GBP 17 million. Those are the two key numbers in exceptional items, getting us to our group loss number. Technology Solutions. We're going to start with Technology Solutions. Revenue grown by 59%. The key driver there, as Tim has explained, that growth in average live modules.
That's the key driver of revenue growth. The contribution margin improving from 65% to 71%. The direct operating costs, two key components, the local engineering costs, that's Ocado people supporting our partners overseas, maintaining the bots, replacing parts that need replacing, and so on. That's improving as a cost as a percentage of sales capacity. Tim mentioned that earlier. A very high contribution margin of 71%. Technology costs of GBP 46 million in the H1 and now at peak levels. You will see technology costs flatten, if not decline, into 2024. Support costs already reducing and expected to reduce further.
I think going back to my earlier comments around the visibility of the cost base, putting costs in the right place, strict financial discipline, adherence to budgets, these are behaviours that are now pretty much across Ocado and very helpful in delivering these outcomes. I've pretty much gone through those statements down there, and then delivering that positive EBITDA flow through that we talked about. Very high profit flow through. KPIs are improving across the Technology Solutions business. Ordered site modules capacity now at 232 modules. What this represents is that sites that have been ordered at full capacity, so sites ordered by our partners that you will have read about, where those announcements have been made across our various international partners, the most recently being the Lotte order.
Live modules are up by 22% since the H1 of 2022, at 105 live modules. Here's the direct operating costs, down to 1.8% from a peak of 2.7% of sales capacity. We expect this cost to reduce further, as we've said previously. Technology Solutions. What's the swing in profitability for EBITDA of Technology Solutions? The key number here first, of course, is the revenue growth and the contribution margin flow through. That's 71% number. We then have a small increase in technology costs that you saw on the previous slide, and to Tim's earlier comments, we are investing in sales and Partner Success. The return on capital from that investment, there are strong arguments for that number to grow, so don't be surprised if it does.
We know we'll see great returns from that investment with extra orders, new orders from our existing partners. There's been a GBP 12 million net reduction in support costs, GBP 10 million reduction in support costs. Again, I believe there's more to come. We benefit from the sale of the Dartford spoke, a GBP 5 million benefit there, getting us to that GBP 6 million H1 EBITDA number. I've mentioned this in my earlier comments, technology investment is now at peak levels in fiscal 2023. A lot of our technology investment over the last couple of years has been on these innovations here on the right-hand side of this slide. We announced these in January of last year. Auto Frame Load is now live in two sites. On-Grid Robotic Pick is progressing well.
It will cover over half of our range, covering ambient and chill. It's on track for target of 300+ units per hour. More on this later as we get ready to go live. The 600 Series Bots prototypes are ready now for live testing, we're aiming for full production by the end of the year. The lighter bot on the lighter grid, that's on track for installation in fiscal 2024. My point here is that having invested a lot in technology over the last couple of years, the outcomes are close to delivery, if not being delivered already. There'll be a natural wind down therefore of our technology spend, unless, of course, we find other exciting things to do. At this stage, that's the guidance that we're giving on technology cost.
Just want to highlight here the hosting cost optimization, that's the single largest contributor to the client contribution margin improvement. Hosting costs, that's cloud cost per module, have reduced by over 30% since the end of fiscal 2022. Support costs, GBP 101 million last year, GBP 89 million this year in the H1, and we put here on the right-hand side, the breakdown of those costs, what's included. A key point to make here, these costs are fully costed. There are no costs outside of the divisional spend, group and other. Last time around, it was a distinct cost. The Technology Solutions business largely takes on those group and other costs.
Sales and account management, Partner Success, you know about those teams, supply chain operations, finance, legal, HR, Group IT, real estate, and then finally, board costs are all fully costed within Technology Solutions. The key drivers, I'm going to talk about these actually. We have invested in Partner Success and account management, but costs now are directly owned by the appropriate owners. Really important discipline. It sounds obvious, but it's proving incredibly valuable and wasn't always as it should be. Continuous improvement of our operating model. What do I mean by this? Ocado grew a lot over 2020, 2021, bit of 2019 as well, as we expanded internationally.
We set up teams overseas, a payroll team, a finance team, engineering teams, we've learned a lot about our operating model as our business has evolved over the last couple of years, we've identified ways of doing things less expensively, more efficiently, effectively than our initial foray into the wider world. Okay? We're getting leaner and tighter on cost management. Efficiencies from investments. You'll recall we put in an example from my patch, from finance, we put in Oracle Fusion in October 2021. That has proved to be a big deliverer of a reduction in cost base for us, not just in giving numbers that we can rely upon more easily, and a better platform from the R12 version we have for Oracle, but at the same time, delivering headcount efficiencies.
Headcount reductions, I've talked about generally, going back to my comment around how our operating model has evolved and certain teams that we believe we can do either without or reduce numbers. Key driver of the reduction in costs. Tight management of discretionary spend. Again, simple example, a review of contracts that were in place on a retainer basis, often advisory, opportunities to terminate those because maybe their value has run the course. Key part, again, of cost reduction. Just highlighting that the reduction does benefit from that one-off benefit from the sale of the Dartford spoke. Whoops! That's Technology Solutions. Our logistics business remains a safe, reliable generator of EBITDA and cash flow. As a reminder, it's a cost-plus business. The important line here is cost, because there's always a markup on costs.
We recharge pretty much all of our costs to our two clients in the U.K., Ocado Retail and Morrisons. Costs grew by 2%. Orders were at 4%, but each is with 4.3%, each is down 2.6%. The impacts of wage inflation was offset by those productivity improvements that Tim talked about, units per hour improving, and also a reduction in unit utility costs that are off the highs that we saw in fiscal 2022, principally for electricity. Productivity, I've just talked about. EBITDA of GBP 50 million in the H1 of the year, and in line with the number we saw last year. Ocado Retail. Again, we've pretty much gone through the key messages here. Revenue growing by 5%, gross margin declining somewhat as we've invested in price.
Our price inflation has been well below that of published national inflation. We've seen, therefore, that reduction in gross margin, but also impacting gross margin by an increase in our vouchering spend activities. You'll see that in this line here around cost optimization, in marketing spend, GBP 7 million lower, but we've pretty much reinvested that number back into vouchering to grow customer numbers. Strong customer growth up 11%, orders per week up 4%, and the basket value is at 1.5% to GBP 121. Gross margin, I've talked about. I've talked about fulfillment and delivery costs. These are pretty much passed through from our logistics business. Marketing costs, I've mentioned. Our support costs are up GBP 18 million.
It should highlight that last year's numbers benefited from a one-off GBP 16 million release of a management incentive accrual. That is, really, one could look at that as a net GBP 2 million increase in support costs. Finally, EBITDA is in line with guidance. We've had improving EBITDA as the half has progressed, and we've been positive throughout Q2. The H2 of the year, we'll see a growth in customer numbers, a growth in orders, and increasing profitability as we cover that excess capacity that Tim mentioned in his slide a little while ago. Okay, I think you're all familiar with those drivers of how we get to our revenue number, orders per week up, average basket value up, average selling price up 8.4%.
NielsenIQ's number is around 13% grocery inflation in the UK over the same period, well below national grocery price inflation. Ocado Group cash flow. Tim mentioned it, and thank you, Tim, for highlighting that point, that moving the business on to a cash flow understanding. Cash flow is not a particularly well understood financial term. You know, people are used to revenue and EBITDA, and it still surprises me. You'll still hear occasionally, "Well, we don't need to worry about that number. It goes through CapEx. It doesn't affect EBITDA." You may have heard it from other businesses that you've talked to, but that's, it does get so repeated. Moving on to cash flow is critical.
You know, cash, true economics, an understanding and appreciation of cash flow. Cost management is around cash management at the same time. Underlying cash flow has improved by GBP 108 million the H1 of last year versus the H1 of this year, driven by that EBITDA growth. Also, CapEx reductions, fewer sites under construction at the moment versus where we were this time last year. You'll see that in the next slide. We remain in a very healthy liquidity position of GBP 1.3 billion. That's GBP 1 billion of cash and the GBP 300 million undrawn revolving credit facility. Here's our debt profile in the bottom left-hand corner here. The convertible bond, still over two years away from maturity, the unsecured note of GBP 500 million, and another convertible in January 2027.
Just reiterating, our guidance is unchanged on cash flow to improve by around GBP 200 million year-over-year. CapEx. CapEx materially down from GBP 367 million to GBP 284 million, H1 versus H1.
Key driver of that, the site CapEx is lower. Again, we highlight at the top there, that a year ago, H1 2022, we had 12 sites under construction. We've got 5 under construction at the moment. Key driver of the site CFC CapEx. As a reminder, most of the in-year CapEx is in respect of sites going live the following year, if not the year after. Having that visibility of those outer years is a key driver of a capital expenditure. There's our H1 summary. KPIs improving across all our operating divisions. Technology Solutions, turning EBITDA positive for the first time. We're continuing to improve productivity and logistics. Retail, that improving profitability on track for that full year positive EBITDA. Cost control and financial discipline in place.
Still room for opportunity there, and we expect to deliver more. Cash flow improving and liquidity remains strong. On the guidance, our guidance remains unchanged, the one that we gave six months or so ago. You can go through there. Technology Solutions, Luton's going to open in the H2. The two Kroger CFCs scheduled for 2023. We're now working closely with Kroger on the revised timing for those. Further efficiencies and cost reductions I've talked about. Similarly, in the logistics, EBITDA outturn of around GBP 25 million. Again, you're familiar now with these messages for Ocado Retail. Growth, revenue growth, customer growth, order growth, plus efficiencies to deliver a positive EBITDA. Just to highlight that the customer growth will be slower in Q3 of the H2 when we give you our trading update for Ocado Retail for Q3.
We had a big marketing drive in the comparative quarter of the previous year, so look out for that one. The volume-driven growth is expected to accelerate in Q4. Positive EBITDA for the full year, reiterating that earlier guidance. There you are. There's a summary of it, just as a reminder of what we said with our full year results back in March. All of these on track. That's it. Got us back on time and ready for your Q&A. Thank you.
Oh, I might lie somewhere else at the same time. Hi, Nick Coulter from Citi. You somewhat teased us on Intelligent Automation. You talked about a fraction of the space, but could you perhaps help us quantify how much better the throughput is in your solution? The business model that you're going to adopt. Is it kind of a lease model, or are you actually selling the kit? Do you see the cash coming through faster? If you could just comment on the customer acquisition environment in the UK and the kind of combination between-
Let's do one at a time.
Yeah, sorry.
Sorry. That's all right. Yeah, the current anticipated model and all the customers we're talking to in OIA is the traditional material handling robotic supplier kind of sales basis. From a cash flow perspective, it is a question of making it, marking it up, selling it, and receiving the cash during the period that you're making it. It's a cash flow positive from day one, substantial margin on sale, and it looks very similar to some other businesses in that space. Okay? The scale of the opportunity, you can see that from all the other listed providers of storage and, you know, automatic storage and retrieval systems, and robotic providers.
The key normally is how much you need to spend on R&D to develop a product and onwardly develop a product, and at what point do you hit scale, where you therefore, your attractive margins on each incremental installation can cover that R&D? The attractive thing here is the grocery business has paid for 95% of the R&D it ever needs to do. Therefore, it can start very quick. You know, the first sale can actually be genuinely a cash flow positive, not just on a contribution basis, but on absolute basis, 'cause there's very little incremental R&D for it.
Just asking about the, kind of the operational side of the solution. As the customer sees it, why do they come to you? What are the throughput advantages?
I think some of the reasons they come to us in particular are much deeper knowledge of what it is they're actually needing to achieve to most people that make this kind of equipment. If you look at the Swiss, German, Austrian, Norwegian, et cetera, manufacturers of storage and retrieval systems, that's largely what they are. They're manufacturers of storage systems. They don't operate two dozen warehouses globally, 24 hours a day, you know, 7 days a week, achieving very high levels of uptime and throughput, with very varied requirements on that kit in terms of seasonal patterns, weekly patterns, hourly patterns, and unpredictable customer kind of demand and stuff like that. Our knowledge is an enormous asset to us and something that we probably underestimated in terms of its importance to the potential client base.
Our solution itself, I believe, has the highest throughput capabilities in any space requirement of anybody. If you have a high throughput requirement rather than just a high storage requirement, we're extremely attractive, 'cause grocery is very high throughput, extremely high throughput. If you have a complex requirement, which grocery is, because, you know, we've got, on a given day, tens of thousands of customers, each taking 45 items from three temperature zones that are very varied, we're very, very good at handling that. Our solution is incredibly flexible, if you're not 100% sure of your operating metrics and what they're going to look like, not just now, but by the time the system goes live and in 10 years' time, our system is, and our software is very, very good at handling that.
Our onward development journey, meaning that you can combine best-in-class, you know, robotic pick with best-in-class storage and retrieval systems in a very efficient way, is very attractive. Our unfortunate experience with in a certain risk area, thermal risk events, means that we're now a world leader in understanding and mitigating thermal risk events in future sites. There are a number of areas where I would say we are ahead of the market. The historical product, the 500 Series, and the associated grids, were more expensive to achieve that ability to have such high throughput per square meter of space.
The innovations that we bring with the topology-optimized, additive manufactured, 600 Series, massive weight reductions, and therefore the associated new 600 Series grids that have got, and I outlined when we did Ocado Re:Imagined, massive cost and time reductions mean that we're no longer at a premium for that excess performance. Overall, it's very attractive.
Would you care to put a figure to that excess performance in terms of throughput?
I would rather not, but it's very meaningful because it's faster in a number of ways. Accelerates faster, dealer accelerates faster, hoists faster, can operate at much greater density, in particular, because of its single space footprint, as well as the very advanced algorithms that run it, that we constantly enhance. You know, our bots per module has come down by something like 20% in the last two years. That also means that our throughput is limited by bot congestion. If you can get your output from less bots, you can do more throughput, and that's why, again, we're going to go back to all our clients and explain to them that the sites that they built, that were, say, six modules, are becoming seven modules type of thing.
I think it's not 5% or 10%, it's probably more like 2x of what you're capable of doing in a square meter, but obviously I can't be 100% up-to-date. Funnily enough, the rest of the global industry don't enjoy sending Ocado stats on their machinery. We are very, very high throughput, and we've not seen it anywhere else. As I say, when we look at one of the, or more than one of the clients that we're talking to and the site that they had and the plan that they had before they saw our equipment, it's like: Okay, well, we can do that, but what was the other half of the building for?
They say, "Well, what do you mean, what's the other half of the building for?" I say, "Well, we're going to do that in half your building. What was the other half for?" You can see how much better density we're achieving than other people's solutions.
Great, thank you. Then on customer acquisition in the U.K., just to comment on how that environment is evolving and a mix between marketing and vouchering and the kind of combined cost. Thank you.
I would say that operational excellence is the official term of what Hannah and the team have been doing for the last kind of six to nine months. Back to the future or back to basics may be another term for it. Our marketing mix is looking more like it did in 2019 than what we were doing in 2020, 2021, 2022. It is a mix between getting the customer to try it, and part of that is the incentivisation. We are not going heavy on incentivisation. We've never gone heavy on incentivisation, as we saw, particularly in the q-commerce world, where people were doing 100% off your first order or something like that.
We normally try and limit it to below 25% off your first order. What we have been doing, is doing a little bit more incentivisation during the nursery journey to get that first-time customer through to their fifth shop, where we know they then become, you know, a kind of exceptionally loyal because they've got familiar with it, and the familiarity means that they, in their own mind, they can see all the benefits without seeing any losses. They shop forever with increasing frequency and growing basket size. It's just, there's always a balance.
Historically, if I looked at it outside of the for the period that I've run it and the period that Hannah's run it, I would look at it and say we were usually somewhere between 2% and 3% between the two, and neither ever went below 1, if that makes sense. You know, we were sometimes slightly more on the vouchers and slightly less on the media. Sometimes it went the other way. Media also got crazy expensive.
Media in early 2022 was, I don't know, it was double or something like that, for the same amount of voice that you had before, because of all that stupid, if I'm allowed to say that now, VC money, that was chasing the q-commerce world at, like, ridiculous levels, where they were trying to, you know, buy every... You know, you'd wake up, and if you looked at your Instagram, you got 15 adverts from different people trying to do an EDC grocery in London, and it was just stupid. As you can imagine, just trying to do your normal business was hard because they were buying all the space, trying to throw money at that.
You know, you'll recall, we sat here two years ago, and people were talking about those companies being bigger than us in the UK now, and I think we're 20x larger than Gorillas and Getir combined in the UK today.
Thank you. Thank you.
Hi there. William Woods from Bernstein. Three questions, and I'll go one by one. First,
Thank you, 'cause you know I can't remember them otherwise.
On Kroger, we've obviously seen some impressive improvements in terms of percentage changes. Could you give us some context in terms of how that benchmarks versus the UK today? For example, you talk about less than 1% of waste, getting up to 200 units per hour. How much more is there to go into H2?
There's lots more to go for. There are some attributes, for example, of their basket. That means that the prod that I'm aiming for there, we're aiming for an adjusted prod. We show them what we do in Purfleet, and we then adjust it for certain US-centric. They have more frozen in their baskets, and frozen at the moment in those original facilities is un-automated. If you swap out, say, chill for frozen, you know the productivity is going to be slightly lower. They do more manual food prep in a US supermarket than they do in the UK. Their produce is less packaged and barcoded, and stuff like that.
We work with them on an adjusted UPH target, sorry, and we're working towards it, and we're doing really, really well helping them. The key now is to make sure that what we're achieving in Monroe, where I've got the former manager of Erith living, have moved his whole family to Cincinnati, right? Is getting the help that he's delivering to the on-site management team, and that's Kroger's site, to be able to kind of do that across the eight live sites, if you see what I mean. The results are there, but we need to. In something like DPV, it's harder because comparing DPV of...
You can't even compare Cin, you know, the Cincinnati shed to the Florida shed, because the Cincinnati shed is in Kroger's most A, it's in a dense area, you know, relatively dense area. B, it's in their most mature geography as a, as a retailer. Their highest sales penetrated, you know, their highest market share. Whereas Florida was an entirely new market for them. You know, as you start, the aim in Florida was never to have one shed. The aim was to use that one shed as a landing post for a Florida business, and therefore it covers a wider geography, and so you can't get the same DPV there. We've got a lot of work going on in that space. We've improved the numbers already, and I can see a lot of significant improvement opportunities.
Got you. The second one was just on modules. It looks a little bit slower in H1. I think you opened six modules. One was in the UK, you got two new live sites. Is that a slowdown in the ramp-up of some of these CFCs? We used to talk about three years. Do you think that's now more like four?
I think it's going to wildly vary CFC to CFC. You know, I think Luton, for example, we're going to see a very rapid ramp-up of Luton, 'cause we're going to migrate the volume from Hatfield into Luton, for example. You know, our clients are doing different things in different places. Some of our clients, it is a new market for them, and then it can be a slower growth. Some of our clients have got existing store-based businesses that they're going to just immediately flip that volume into, and therefore, we expect those to have faster growth towards full. There are some out there that I could see being full in 18 months, and then there are others going to take four or five years, because they've just entered new markets, and they're not very aggressive on customer acquisition, and they're happy growing it slowly.
I think it's really vary site by site. What we have done in the business is, going forwards, we have changed the, both the CapEx for new sites and the need to deploy that CapEx in those sites, such that our economic return as a, as, let's say, a percentage of deployed modules, doesn't mind whether you roll them out over five years or two years. Right? The fixed element of each warehouse is covered by the design upfront fees of the client, and therefore, whether we roll out 10 modules in one module in 10 sites or 10 modules in one site, our contribution per module will look the same. Does that make sense? We weren't there two years ago. We are there now in terms of our future.
Can I just add on that one as well, that you're right, that the rate of modules going live isn't what we would have expected for our international sites, but it does reinforce the importance of the Partner Success program. You know, unlocking that will accelerate the growth in live modules. That's why it's so important that we get that right, and that we invest in that space.
Got you. The final one was just on OIA or non-food. Obviously, one of the key benefits, I think, of Ocado, is not just the hardware, but the software as well. Do you see recurring revenue and profits in OIA coming from either software and analytics or from things like Partner Success consulting revenues?
I mean, there will be some, yes, but the overall You know, whereas the OSP module is a fully recurring revenue type model, the other model is more of a hardware sale and a recurring license type fees and maintenance and advisory and stuff like that. That is currently the two kind of different models there. There are opportunities that I'm not currently pursuing, but we have been approached, et cetera, to take some of the In the same way we're taking kind of the storage machine of OSP and saying, We can sell that as a standalone business, we're in the ASRS business now.
We do have some very advanced software in our stack as well, whether that's in supply chain, whether that's in routing, whether that's in warehouse management systems, that we could do some additional development on as well, and then move into one of those markets. We haven't decided to do that at this stage.
Got you. Thank you.
We have time for you. Is that okay? Thank you.
Thanks. It's Luke Holbrook from Morgan Stanley. I just wanted to ask on your CapEx guidance, you've reduced it half on half, you're expecting lower CapEx. I think on the last slide, you also said 5 sites were currently in construction. If just if we take out Coles and we take out, I guess, Luton, that would only imply potentially a couple of additional CFCs, which for next year. Is that the right way to be thinking about it?
Do you want to go?
Well, we expect as a rule of thumb, we expect 10 sites to go live over the next couple of years. It could be five, or it could be six, four, but that's the kind of ref guide, the visibility that we have today on sites under construction. It will be, there's Madrid, for example, that's going to go live. We've got Poland on the track. We've got the two in Australia that we've talked about.
Korea.
There are remain, I think, about sort of four or five Krogers still to go as well.
Okay, those are confirmed from the customer, they're in construction.
Those are confirmed from the customer, yes.
Okay. second question?
There's two Krogers in construction. There's a series of signed contracts, they're still looking for sites.
Okay, thank you.
We've... Sorry, there's some more Japans that will come as well.
Okay, that's clear. Then just second question, I know you've seen some ongoing delays with Coles at the moment in Australia. I just wondered if you can touch on what the reason behind that was, 'cause we have seen a couple of delays with other CFCs in the past. I didn't know if it was just a customer-specific issue.
Well, the main delay there was COVID. We have seen some construction delays beyond that, which is the first time we've seen construction delays of that type, and they're down to kind of early health and safety requirements, and the way that we were allowed to build the grids that we would normally have built in a certain way and being asked to build them in a different way, and just working our way through those issues. I think we're going to shortly have very clear guidance as to when those sites are going to go live now. They will have taken some months more than sites in other markets. It's kind of ironic because at the same time, the team in Asia were building in Tokyo.
Tokyo obviously is the most different environment for us to build in, and the grid in Tokyo is by far the most complicated grid we've ever built because of the whole seismic element to it. We built in a non-base isolated building, and that one went live early. It's a little bit ironic that Australia is where we've had the delays, but they are delays. We've built 20 something of these sites, and they're one-off. We'll get some learnings as to why they, you know, why we got delayed there and make sure that we learn from that, so we don't get delayed again. I don't think it's a material issue. Don't expect we're going to be talking about it for very long.
Thank you.
Hi, it's Marcus Diebel from J.P. Morgan. Also three questions, but maybe also one after the other. First one, again, on marketing spend. You're running at 75% utilization rates, and clearly it's about volume growth, more customers, then utilization rates improve, and then EBITDA improves as well. Which raises the question, can you grow customers even faster, and why is it right now the right mix? On the one hand, more marketing costs, but on the other hand, you would have benefits and utilization, and how do we actually manage that mix?
Marcus, you're absolutely right, and we are discussing this ourselves with the business, and it does come down to the marginal acquisition costs. If we could spend double in marketing and grow twice as fast, you would do that. It, the payoff is phenomenal if you can do that based on the capacity and the marginal contribution on incremental sales now is in the high teens, and so it's got a very attractive payoff. The issue is trying to. If you kind of think about it as like mining, I've always described the marketing as mining. It's finding other seams that you can mine at not massively increased cost.
If you could go out there and say, for a 50% incremental cost, I can go and acquire another equal amount, another 10,000, 12,000 customers a week, we should be doing that. It is a question of the teams there working out where to find those seams from and not, you can sometimes think, "Oh, I know what, I'll just throw a bigger voucher at it." You throw a bigger voucher at it, you acquire 20,000 customers, not 10 that week or 12 that week, but the problem is they don't stick with you because you acquired the wrong ones. It's kind of the ones that you want are the ones that are. You need to get some voucher to kind of make a small motivation.
You give too big a voucher, you get the wrong customers in, if you see what I mean. There's a balancing item, yeah, the teams there are working to find more seams to mine more aggressively. It's a very good point because we have very strong returns on it right now.
Yeah. Plus, majority of customers do stay, yeah? I mean, that what you said in the past, I mean, that the voucher hunters seems quite low this group.
The problem, and it's the problem that the q-commerce people have found, is that you can put very aggressive growth. You know, it's very easy to give away pound coins at the end of the street. It's very quick to find a queue.
Right.
Right? The problem is that when you now want to charge for your services, they've all disappeared. It is a balancing item, and as I say, the team have got back to a lot of the methods that we used historically. They're doing a very good job with it. The retention is very strong, and they are seeking to work out if there was the opportunity to invest more money to do it.
Perfect. Maybe just the second question on, on continental Europe or maybe even just Germany. Some interesting developments there, yeah. A large Norwegian player had a lot of ambitions, launching in Germany. They're gone. You mentioned ultra-fast delivery. They have the massive problems, and you called it obviously right. They're probably gone-ish, yeah, if I can say in this audience. What does it actually mean? Now, the only one that's really successful is someone like Picnic, who has a different business model. From a consumer perspective, not as strong as Ocado, but from a financial perspective, really strong, given their business model. What does it actually mean? Is that kind of like the, the setup for the industry? That is, say, "Okay, we don't best in class, super high customer experience.
We actually try to be the best alternative, but we need to have a happy CFO, basically, yeah, at the end of the day.
Sure.
Is that the new model?
I wouldn't say it's the new model. I mean, look, Germany's always been a very, very, very price-conscious market. It's got the highest share of discounter, market share of anywhere, both the big, successful global discounters are from Germany. Average item prices in Germany are low compared to the rest of Europe. The Germans have not chased high quality groceries, particularly, and so their average item prices are lower. It is a market that we know a number of incumbents in, and are continuing to talk to a number of incumbents. Our software allows you to do, you know, to kind of set the dials and operate more like a Picnic or operate, you know, more aggressively than Ocado Retail does in terms of slot availability.
Depending on how you want to set it up and how much you want to spend on delivery costs, if you see what I mean, you can make sure that every time, you know, 95% of the time a customer comes in, they see 15 slots the next day, or you can set it up so you look like a Picnic, and you're doing a milk round, right? You can do all of that in our software. Obviously, that's up to a retailer to decide what their proposition is and how they want to run it. I think, you know, that Germany is an opportunity, but yes, the few people that have gone in and tried to crack it externally, being the player that you say from Norway, Oda, and the q-commerce players have struggled.
Perfect. Then maybe the last question for Stephen, just on DNA. Obviously, had a big impact on the results, on cash. When does DNA actually peak, given the pipeline that you have today? Just as an assumption.
Yeah, given the pipeline we have today, I expect it to peak around fiscal 2025.
Okay, thank you.
Yeah.
Appreciation on the time.
Thank you. Sreedhar Mahamkali from UBS. Excuse me. You've talked quite a lot about the Partner Success program, and rightly so, but just give us a sense... I mean, clearly, this wasn't something that you were expecting to put money into, nor probably your partners. When does it really start to move the dial in terms of additional orders of modules or CFCs? What sort of time line should we-
It's moving the dial now. Of course, if a client took down a CFC with four modules, and they were using one and a half, and now as a result of it, they're using two and a half and three and a half and three and three quarters, soon we'll see the impact because they'll now need to buy the fifth one, right? You know, what you can't see is the underlying growth of our partners utilising what they've already drawn down. I think in some of our partners, probably both because of our, what we were doing before in terms of what William was asking, in terms of the costs, you know, we were putting more CapEx in up front.
We were saying to them, "You need to draw down this much to start with, and you can't draw down more modules for 12 months." Clients went in there, sometimes with 50% or more on day one, and they're filling that. I think in future, partners will come to us and say, "Can I start with 25%?" You'll see these sites go live, and then you'll see them starting to draw down modules more quickly. It is having a impact already at helping our clients to grow their underlying businesses. That will result in the drawdown of more modules as they start, as they grow.
Got you. Couple of other ones. Ocado Retail, you start to get some options next year as you approach the fifth anniversary with the Ocado Retail JV. What are your thoughts now? How do you look at it? What are the sort of parameters you look at when you get to that fifth year?
I don't think the fifth anniversary triggers any, you know, rethinking about our ownership of the 50% of the joint venture. I certainly wouldn't even, you know, contemplate any change of ownership before that business was humming in exceptionally good shape. You know, in high mid digit, high, mid-single digit, EBITDA-type margins, and, you know, above mid-single digit, PBT-type margins. We see that business having a very exciting future in the UK, and at the moment we're very happy with it being part of the group.
What I am happy with is that we have this clarity now that we possibly lacked before in terms of understanding what are the costs attributable to Retail and Retail, what are the costs attributable to Technology Solutions and Technology Solutions, Logistics, et cetera, not kind of having this bit of confusion sometimes internally with people saying, "Oh, we're a, we're a technology business with 20,000 people." I'm like, "We're not a technology business, 20,000 people. We're a technology business with 5,000 or 6,000 people in technology solutions with 13,000 drivers and warehouse operators," right? you know, you can't think of yourself as a technology business, 20,000 people.
Yeah.
We're very comfortable at the moment with it, and I don't see any need to do anything different.
The one practical change is that from fiscal 2024 onwards, M&S will have the right to elect to consolidate the numbers within their numbers. Next year could be the final year that we consolidate the Ocado Retail numbers.
The last one. You've talked quite a lot about non-grocery opportunities. What are your thoughts on grocery, growing the grocery business further, new clients? How is the sort of cadence, intensity of those conversations going? Are there much conversations.
Yeah, strong at the moment, but there is an element of ebbs and flows to it. It's got stronger again because it was kind of... The first part of COVID, everyone was too frantic to talk. you know, just literally trying to run their businesses and get the throughputs out that they were kind of being demanded of and operate in that difficult environment. There was a lot of interest, but people couldn't travel, and people couldn't look at things. We went through the kind of the COVID slowdown, then people suddenly went, "Oh, I thought I was really interested, but I want to wait and see what's happening now," in terms of the kind of post-COVID slowdown.
Now things are stabilizing and growing again, and a lot of people are coming back to the table to talk. We've got a lot. I think we've probably got as many live conversations going on now as we've ever had. Obviously, we're already in a number of the largest markets in exclusive contracts, but we have a lot of conversations going on in other parts of the world. I expect to see growth of customers as well as existing customers grow their businesses. As I said, we've got a lot of focus on that at the moment because of the fact that we're already in a number of, you know, we're already in the U.S., and we're already in Japan, we're already in Korea, we're already in Australia, we're in the U.K., whatever.
There is definitely opportunity for growth in grocery, and there's also opportunity in grocery for growth outside of single pick e-commerce.
Hi, Andrew Gwynn from BNP Paribas. Let's go for three questions, Annika. Firstly, just clarification. David, you mentioned earlier, sorry, the PBT guidance for the year. I think it's GBP 450-GBP 500. Was that including one-offs? Sorry.
Sorry.
The, you said PBT.
We're talking about depreciation, amortization.
That's for the year?
Yeah.
Sorry.
Yeah.
Yeah. Okay.
That's for the year.
Second question, I suppose to be very blunt, is Ocado or one of its advisors looking to sell the group?
Definitely not.
Okay, fair enough. The final question, just on the non-grocery business, you think about obviously when the Ocado Retail sort of kicked off, and you found partners, obviously a couple of catalysts for that. Is there a particular catalyst we should have in mind for signing those first non-grocery contracts? For instance, when maybe when Luton opens, and we begin to see reimagined.
No, I just think that these things are normally bubbling along in the background, and then, you know, you get to a point where you're the chosen supplier, and then you go into to contract negotiations, and people need approvals and stuff like that, and they just take some time. You know, when our starting point just meant that the those earliest ones are going through those motions. So I don't think there's a catalyst that suddenly we're going to go, "Oh, that was the catalyst, and now we're going to sign loads of them." I just think it's an ongoing process and expect that we will start to sign some.
Okay. Sorry, Stephen, just back on the PBT point, GBP 450-500 million, it was GBP 192 million in the H1, it's a very, very big step up in the second.
Of depreciation?
Yeah.
That is sort of a guide when it will peak. I'm not saying it will get there immediately. It will take time to get there. GBP 192 H1, you know, double that, and that's your full year number, there or thereabout.
Okay. It's a 380-ish.
Yeah.
400. Okay. Thank you.
Thank you. Yes, James Lockyer from Peel Hunt. Mostly follow-ups, I think, actually, from other people's questions. Just on customer acquisition, how much of the marketing spend that you do sort of per half, is for new customer acquisition versus existing customers? How much of your new customers are free, let's say? That's, I guess, the first question, or I can run through all three of them.
Most of the spend is on customer acquisition as opposed to retention or reactivation. We look at kind of retention spend, reactivation spend, and acquisition spend, and acquisition is the majority of it. What you do find, and we've always found, is that if you turn off the acquisition spend, you still acquire customers. The problem is that if you say that we're going to put a GBP 20 offer out on a, on a first order above GBP 80, for example, if you don't have one at all, you still acquire customers.
You put it on Google, you know, it's easy to find, if you see what I mean, and then some of the ones that you would have got for free, you end up spending some money on. If you don't do that, you don't acquire the other ones. Does that make sense? Yeah, there's Whoever was asking the question before about could we go faster, it's like both ways. You could, you could massively reduce the spend and still grow at maybe 40% of what you grow at, or something like that. That's one of the reasons why you have to increase the spend quite dramatically to get a doubling, because obviously, you don't get any free customers, in the second part.
I guess maybe a clarity, though. Of the 959,000, I think, the inactives in the period, how many of those are regular weekly shoppers who don't need to be prompted to come back versus?
Sure.
Who need to come back?
We look at kind of active actives, if you see what I mean. That number is, I think that's, is that 12 or 16 weeks? If it's a 16-week, I can't remember if it's 12 or 16 weeks, but if you're seeing as something like 150,000 of those, are recent as 12 weeks. Something like 120,000 to 150,000 of that number, off the top of my head, right? You know, are people that have been acquired in the last 12 weeks, and therefore, some proportion of those will become active actives but aren't already.
That's why when Stephen was saying in the Q3 last year, we had some TV advertising that we turned from brand advertising to direct response, which massively increased the headline customer acquisition numbers and threw them into that pot. As I was saying before, they weren't the right customers, so they had higher churn rates. Which means that the underlying active actives can be growing while that looks like it's relatively static, 'cause you're, now, you're not putting in as many poor quality ones into the top as you were when you look at the year-on-year comparisons.
... Thank you. The second is a follow-up on the Kroger improvements. I think the CEO was talking at some point in an interview where he thought that the online model could be as profitable or more profitable than the in-store model. I guess that's part of the projects you've been doing that might have informed his comments on that. At what point do you believe that might happen? If that, at that point, could you see prices online be more attractive than offline, which could then drive more people and then the virtuous circle and just keeps going up?
Look, one day that needs to be possible. I don't see it in the next, you know, 6-12 months in the U.S. market, for example, 'cause we've got to see growth, and there's a lot for them to learn, and there's changes they need to make on their own software front and stuff to drive the type of efficiencies that we would look for in other markets. It is theoretically still the goal that those warehouses in the U.S. will generate higher EBITDA and higher PBT returns than their stores can, and therefore one would logically assume that at some point, someone will put that, maybe reinvest that back into pricing, and we will hit that tipping point. We've got, you know, we've got a lot of growth to go before that.
you know, in terms of 10%, 20%, 30% online penetration before that, than in a retailer that post-merger will be over, you know, well over $200 billion of sales. There's a lot of opportunity first before we see a tipping point for the whole U.S., you know, the $1 trillion of U.S. grocery market to go online.
Thank you. Just one final piece.
This will probably have to be the last one or penultimate question.
Just on the split of the revenue in the period for technology, what % is at maturity, so true recurring, sort of not growing, say, sites, new sites that have expanded from last year and new customers from this year?
Oh, gosh. Well, I'll give you the number that I do know. Around 90% of the revenue is recurring revenue of the revenue mix. 10%, the balance, is the release of the previously capitalized fees that we receive upfront. I would say, of the recurring revenue, I mean, the vast majority is going to be sort of existing sites that were there in place at the start of the year, naturally.
That's great. Thank you.
Okay. I think that probably is the time where we should wrap it up.
Thank you very much. Nice to see you all.
Thanks, everyone.
Thank you.