Ocado Group plc

Good morning, everybody. Exciting week this week, Cricket World Cup, Wimbledon and Ocado’s first half year results.

You can decide yourself which is the more exciting, which is more important, but Tim will take you through the Ocado bit in a second. Now at previous meetings, and I wouldn’t keep you long, I’ve spoken about the rate of pace of change in Ocado that we’ve had over the last few years.

Not only have we continued to grow our best-in-class U.K. retail business by adding better services and more convenience and better value, but we have also continued to refine and develop our proprietary technology and then package it up in a way that is attractive to increasing numbers of the world’s most innovative and forward-looking retailers.

And in the last 6 months, this rate of pace of change has accelerated. New initiatives which you’ll have read about or heard about and you’ll hear about this morning such as Ocado Zoom and the introduction of automated picking in our CFCs later this year are two good examples of the initiatives that we continue to bring online which we believe will offer even better returns to our partners on the Ocado Smart Platform.

Now, as you know it’s not all been easy because I think it was only February, and I was reminding myself, only February since we had the devastating fire CFC3 in Andover. And thankfully, very thankfully, that did not result in any injuries but it did affect the livelihoods of many and did make a lot of difficulties for our business.

But it’s amazing, how resilient we have been and we are all very grateful for the work not only that the first responders did but also for all our colleagues who worked tirelessly to mitigate the impact of this fire on both our customers and the local community. You’ll hear a bit more from that I’m sure from Tim in a second.

But we did – we have had and reported this morning a 9.7% increase in sales for the first and I think as a testimony for the resilience of this business and the people who work in it. Now, we know sometimes that the analysts and investors sometimes find it hard to keep up with everything that’s going on and Duncan I’m sure will throw some light on how you interpret the results this morning.

And indeed, David Shriver reminded me the other day that one day last month we made three announcements in three hours. I think we announced that Morrisons’ holiday from CFC 4 at Erith and we ordered the second or -- had the ordering of the second CFC from the Sobeys in Canada.

And we also had our acquisition of the minority stake in Karakuri, which as you know is an exciting robotics company which has the potential to revolutionize the preparation of ready-to-eat meals. So it’s a lot of excitement going on.

Now, I can’t promise, we can’t promise that we’ll have the same degree of news flow between now and the end of the year, but what we can promise is, that the board and the management team at Ocado are absolutely focused and we are working hard to remain focused on delivering good outcomes and continuing good outcome for our customers. So on that note, I’m going to hand you over to Tim and then onto Duncan, who will take you through the first half results.

So thank you.

Thanks, Stuart. Good morning.

So 2019 has really seen a shift continue -- acceleration in the shift to the center of gravity of the Ocado Group. We’ve pivoted from being the pure play online grocer in the U.K.

with initially a small separate solutions business to really becoming a technology led global software and robotics platform business providing a unique end-to-end solution for online grocery. The joint venture with Marks and Spencers Ocado.com is now one of eight global partners all amongst the most innovative and forward looking grocers in the world whose online businesses will be enabled by our Ocado Smart Platform.

We truly believe we’ve never been in a better position to create value. I know this is a slightly confusing set of results particularly because of the accounting treatment for the impact of the Andover fire.

And so as usual, I’m going to hand over to Duncan to walk you through and do the complex part of the meeting, and you’ll come back to me later.

Good morning. Thanks Tim.

Yes, we’ve had a busy first half and a complicated set of results to talk through this morning. I’ll try and help you understand when I get through our underlying performance.

The key headlines are that the business is operating as expected, with strong and resilient performance for Ocado Retail after Andover fire, and a great deal of progress in Ocado Solutions. Both segments are impacted by a number of factors which I will explain.

Now, just a couple of points on the presentation of the numbers. First, on this slide we’re showing the statutory results, including the impact of exceptionals.

Second, you also see that the numbers, these numbers are adjusted for the sale of Fabled, our beauty site. As a result of our sale post the end of the half at Marie Claire Beauty, which traded as fabled.com, we’ve classed this as an asset held-for-sale and adjusted both years accordingly.

2018 numbers are also adjusted for the adoption of IFRS-15 in the second half of last year. I’ll focus most of this presentation on the adjusted numbers.

We’ve continued to grow revenue strongly with our retail business up 9.7% despite the impact of the fire at Andover, still well ahead of both the total on the online grocery market. Revenue growth was stronger in solutions.

This primarily reflects the revenues from our arrangement with Morrisons’ as the fees that we earned from other solutions clients are not recognized until the relevant performance obligation is met, typically on opening of the CFC. Our retail EBITDA declined by 4.3% to £44.1 million.

Solutions EBITDA was a loss of £16.2 million as the growth in reported revenues is offset by significant investments in Brazil’s to deliver for our clients. As you’ll see later, we saw strong growth in cash fees, and in the half, we had approximately £45 million of cash fees from international clients not recognized in revenue.

Our other segment includes the accounting impacts of MHE JV Co, the cost of the board and of share based incentive schemes. As to be expected, given the strong growth in the share price, these share incentive costs have risen year-on-year impacting the second.

Net interest costs were up, slightly to £6.7 million given slightly lower cash balances. Depreciation costs, pre-exceptional increased to £55 million with the biggest increases from annualizing the commencement of depreciation of Erith CFC and for our new front end software platform.

In the statutory numbers, there was a £100 million write-off of building assets for Andover. In addition, that was right-off of stock and further costs offset by nearly £12 million of insurance income recognized to date.

Note that we have received £43 million of cash payments by the end of the period. As a result, we reported a statutory loss before tax of and a loss of £43 million on an adjusted basis.

So here’s the split of both retail and solutions EBITDA showing admin costs for each segment. I’ll cover retail in more detail later.

Service Solutions, revenue growth grew 20.6% primarily due to the growth of our arrangements with Morrisons as there was no material revenue from our international solutions partners. Solutions operating contribution grew slower at nearly 10%.

This reflects the mix of the positive contribution from our relationship with Morrisons where our proportion of the cost the CapEx yet the fees were all reported in EBITDA and a negative contribution from international solutions as we do not report any material revenue until the CFC is out. Admin costs for solutions grew significantly to £28.7 million including with an admin costs other non-capitalized costs of developing the platform.

The costs of hosting and support which are currently minimal and the central commercial teams involved in business development and support. A more important measure for solutions at the moment is cash fees.

Revenue grew over 20% in the first half, but cash fees grew more strongly up over 36%. More importantly, for our long term success, cash fees from international partners almost doubled to over £46 million.

The revenue recognized was immaterial, as under IFRS 15 fees are not recognized until operations commence. The cash fees come from signing new clients for and from commitments to the CFCs.

The more material fees will come once the CFCs become operational. Now before it goes to the retail performance, I want to explain the impact of the fire in Andover.

First, it’s important to say that our business is being resilient and despite losing approximately 10% of our capacity at the time, our estimated sales loss in the half was 2% which equates to about 3% since the fire. We’ve been able to mitigate this loss by spreading demand for our customers more evenly across the week.

This is fine in the short term, but it does run the risk of not being out to satisfy our customer’s demand, which may lead to them getting their groceries elsewhere. For this reason, we did the deal with Morrisons to temporarily get their capacity back.

And as of now, we’re using 100% of the capacity in Erith. Second, we anticipate that all the losses will be covered.

That includes assets that will be replaced, not at their original costs, but at their current cost. Importantly, we expect to be fully indemnified for business disruption including, estimated loss sales and cost of inefficiencies, and for the steps we take to minimize these, such as the fees that we’ve lost by taking back capacity in Erith for Morrisons.

Finally, a word on reporting of the insurance proceeds. We will recognize the building assets as the spend is incurred, effectively recreating them without the cash costs of building them.

For business disruption, we will recognize income once our agreement with the insurer in the amount, on the amounts is formalized. This may take some time as the business disruption may spread over a number of years.

Now having covered the impact of Andover, we are summarizing the impact on consensus. The most recent EBITDA consensus published on our Website on the fourth of June was £45 million.

Given the estimated lost sales to date, the expected impact on sales and marketing in the balance of the year, there will be a £7 million hit to EBITDA which is covered by exceptional insurance income recognized EBITDA. We will also lose the fees as a result of our agreement with Morrisons Erith and now incur all the fixed cost of that facility until the end of the holiday.

This is a further £8 million EBITDA hit again covered by insurance. Finally, we expect an additional a £10 million cost for share schemes at today’s share price as a result of the increase since the year end.

So onto retail performance. Retail revenue was up nearly 10% on the year with total order growth slightly ahead at 10.4% growth up to 318,000 orders per week.

In the half, Erith was serving both Ocado and Morrisons, but we benefited from an over 50% increase in capacity throughout the period. Mature CFC efficiency, meaning for Hatfield and Dordon, improved to 167 UPH.

In Erith, UPH also improved through the period and continues to in the second half. There were some strong improvements in delivery efficiency as measured by drop per van per week, up by 2.7% to 194, but with more to go, to reach our new target of 200.

Note that this measure was adjusted for the temporary costs increase from the fire in Andover, as these costs are taking to the exceptionals as they’ll be covered by insurance. Wastage cost remained at industry-leading levels, and the actual product that is wasted and end up in landfill remained at 0.02%.

So on to retail operating contribution. Gross margin was up 50 basis points, continuing the trend from the second half of last year.

The market remains competitive, and we do not expect any material further gains in the second half. We continue to find ways for our suppliers to increase their sales with us, and this helped us grow our supply income by 40 basis points.

Trunking and delivery costs were lower due to the improved efficiency I mentioned earlier, offsetting the wage cost pressures. The biggest cost movement was for CFC costs.

Operational efficiency was better, up in mature CFCs. It was also improving in Andover pre fire, beating Hatfield’s UPH the week before, and it also improved in Erith.

However, fixed costs from Erith and from our second general merchandise distribution center, both operating in a fraction of endgame capacity, drove most of the 120 basis point increase. The balance was from engineering costs where per-order engineering cost in the new generation CFCs are higher currently than in the previous generation.

Our expectations are unchanged, and we expect these costs to fall as our new generation bots and further enhancements to software and hardware take effect. Other operating costs were in line, and marketing costs were up as despite no proportional increase in acquisition costs, we ran an off-line trial and we had both canceled acquisition costs and increased retention costs as a result of the fire.

So overall, operating contribution was down around 50 basis points. Retail admin costs were up 30 basis points with the majority of this increase in technology as we begin the project to move Ocado.com onto our new platform.

This is a multiyear program but the benefits of operating on one platform are significant. Overall, EBITDA was therefore down 80 basis points.

We expect an improved outcome in second half without the annualization of fixed cost increases and as our business continues to scale. Onto CapEx.

We spent £112 million in the period with the first significant spend on international CFCs driving all the growth versus last year. We continue to invest more in developing our capabilities with increases in technology and also in fulfillment development.

On that, it’s perhaps worth noting we’re about to use our robotic pick station on real customer orders for the first time in Erith. Our guidance of £350 million remains unchanged as the new deals signed this year and the extra commitment from Sobeys for a second CFC will not have a material impact on CapEx in the short term.

Our cash position is strong. We ended the half with £360 million in cash, and with our undrawn RCF, we have £460 million of headroom.

Of course, once the M&S transaction completes in early August, this will grow to £1 billion, giving us sufficient funds to support all the growth in the U.K., improve our platform and meet the current needs of all of our Solutions clients. So to finish, I want to remind you of our outlook for 2019.

Assuming normal market conditions, we’d expect our retail revenues to grow between 10% and 15% for the balance of the year with the upper end dependent on delivering sustained, fast ramp-ups in Erith. In the second half, we’ll not have as material increase in fixed costs as the first, and given our ongoing efficiency improvements, we’d expect underlying EBITDA to grow.

Note, of course, that post the formation of the joint venture with M&S, the commercial terms include a payment of fees for our platform, meaning that retail EBITDA will decline with an equal and opposite increase for Solutions. For Solutions, revenue growth will be below retail due to the Morrisons holiday in Erith.

Of course, we’re still involved in a number of conversations, and so there is the prospect of further deals to come, which will result in more cash fees, some further costs, but as no revenue will be reported, may impact on profits in the short term. CapEx guidance is unchanged at £350 million.

So in summary, 2019 has been a busy year but there’s still much to do. And now hopefully, Tim will give you a sense of our ambition for more.

Thanks, Duncan. So I’m going to talk us through five different points here.

The first one is how we’re scaling the business to deliver the outstanding execution that our clients rightly deserve; our joint venture with M&S and what it will mean in terms of better customer service for Ocado.com customers; how we have a virtuous cycle here, as the OSP club grows, the network effects are magnified; how we’re innovating so our partners can grow faster in a new area, in immediacy; and also how we’re innovating in other verticals as well to benefit both our customers and also our business. So in terms of scaling the business, so we are organizing ourselves to be capable of growing not in one or two warehouses a year but in 10, 20, 30 warehouses a year.

So we keep challenging ourselves to see how much faster we could grow so that we can free up Luke and his team to sign more deals. We’re organizing ourselves into a mission focused, decision making organization.

We have hundreds of people working on these deals at the moment. We have a 20-strong transformation team who are helping us to transform and change our business to be this mission-focused organization, very focused on the delivery of these sheds on time for our clients with the associated software to run them and run a fantastic customer business.

And we’ve been continuing to change, as we mentioned last time, our technology area with new reporting structures and a lot of new processes and many new things in our organization to help drive the throughput of the 1,400-strong team we have in technology to deliver on the promises we’ve made to our customers. We’re investing and we’re investing heavily, investing in the right tools and the right resources as well.

So we’re growing the people, the teams. We’re growing the tools.

We’re changing the processes so that we can undertake multiple programs at the same time and in multiple countries. We’ve had to create new abilities in the business such as having multilingual teams on call 24/7 for our customers.

So it’s not just that we’re going to get calls from people who haven’t historically run warehouses but we’re going to need to answer them in French, in Spanish. We’re going to need to understand what’s happening in the United States and Australia, but that should be easier for us.

And then in the execution terms, it’s not about adding complexity, it’s about adding velocity. In fact, the more that we do, the more simple this is getting.

So the more experience that we have in building these projects, the more experience that we’ll have in commissioning new warehouses. And any new challenges that we face – we’re less likely in each building to see another new challenge because we’re likely to have seen that challenge before.

So one of the ones, for example, that we’re dealing with at a moment is seismic risk. Very -- what’s the right word?

Very relevant this week given the earthquakes that we’ve seen in California. Obviously, there are multiple territories around the world that, that’s relevant in.

What we’re working here on is trying to bring the lead times down. So one of the advantages here is that as we do more of these projects and we’re able to do them faster, able to execute them faster, we can build the lead time for our customers down and absolutely making sure that we learn from each project and that everything we learn is fungible onto the next project to increase our capability and our velocity.

So we aim to be an organization that could deliver dozens of warehouses successfully in a single year. The key, in our sense in the joint venture with M&S was to create a business that was better for ocado.com’s customers.

So to leverage the strengths of Ocado Retail, which was already the leading online grocer, with the best service and an experienced team with M&S who have 12 million food shoppers. So just think about that relative to our 300-and-something thousand orders a week or our 700,000-odd active customers on a 12-week rolling basis.

About 12 million food shoppers a week into M&S stores, so a significant ability to drive traffic to Ocado.com; M&S being the highest ranked food retailer for quality in the U.K., bringing those thousands of products into our existing 50,000-large range and the highest innovation and being partners, meaning that they will direct that innovation towards our customer’s needs. And then bringing together Ocado Solutions with its globally validated proposition, IP owned and operated, the evolving offer including Ocado Zoom.

So the idea of a joint venture is to bring all that together to create an improved offer that’s tech-enabled, that’s asset-light and that’s mission-focused. So CFC5, we’ve spoken about today in Purfleet and will go live in 2021.

We’ve announced seven more CFCs over the next 12 years. Together with M&S, we intend to grow this business faster than we would have done as a stand-alone business.

So the club is growing, bringing Coles onboard in this half, Sobeys bringing onboard a second warehouse. And it’s very much about more partners and more facilities enabling an increased level of investment, an increased level of investment enhancing the quality of the platform that does both -- creates faster growth for the partners that are on the platform, a virtuous cycle, therefore a more increased investment; but also making it more attractive to more partners, also leading to more investment.

So it really is a virtuous cycle as the scale increases the rationale and resources for investing for further innovation. And we see the scope for significant further innovation, for significant increases in the warehouse efficiencies and delivery efficiencies, for significant increases in the proposition possibilities for customers in immediacy and beyond.

We’re innovating so that our partners can grow faster. And so one of the interesting ones, and we’re doing so many we can’t obviously talk about them all today, but in this last half is the launch of Ocado Zoom.

Ocado Zoom has got happy customers, and so there’s a couple of quotes here from customers. It’s almost the fourth emergency service.

We quite like that one. We like, just tried @Ocado Zoom.

Holy Batman that was fast and accurate thoroughly impressed. What’s interesting is that we’re not seeing customers cannibalizing their existing primary shop from Ocado.

In fact, the anecdotal customer feedback that we’re getting is they’re either cannibalizing a mission -- a shopping mission that didn’t exist. We didn’t think we were eating at home tonight.

We are eating at home tonight. What have we got?

Oh, we’ve only got some pasta and some sauce in the cupboard. But actually, we can Ocado Zoom it, and the next thing is we’ve got some fresh salmon steaks and salads and broccoli and all kinds of fresh food delivered to our door 29 minutes later.

So it’s really cannibalizing kind of inactivity or cannibalizing things like Deliveroo. And so it’s really, really interesting.

Two-thirds of the basket -- so what we call a small shop, not an emergency top-up, and a small shop being between £20 and £55, the average basket has turned out to be, I think over 30% larger than our prelaunch expectations. So very, very interesting.

It is, we believe, the best immediacy service out there for the customer. So we’ve launched with more than 10,000 SKUs.

We expect that to grow over time to closer to 15,000 SKUs. We are 95% on time within our one hour from order to delivery.

Our average is around half an hour, and we still have 0.2%, incredibly low, substitutions. So it’s a phenomenal service.

But most importantly, because most of the immediacy services we see have been launched around the world has not been sustainable, services have not had economics that will work and we believe that Zoom will have a sustainable economics of scale. It will have slightly higher gross margins due to the product mix and some lower promotional activity.

It will have increased delivery income as a percentage of sales. So with similar delivery charges, obviously, in smaller basket, it has increased delivery income.

And at the moment, delivery fees are between £1.99 and £2.99 per customer, offset against the slightly higher cost to fulfill and only slightly higher because of a network effect of using the large CFCs to fulfill the smaller Zoom CFCs, which are already -- the large ones being already enormously efficient in single pick economics, still with low waste due to the frequency of fulfillment and the data that we’re able to collect and therefore manage our food carefully and a slight increase in the last mile cost due to the point-to-point delivery but also the smaller radius that those sites serve. So overall, Zoom is both a massive positive for Ocado.com in the future but also will enable our partners to meet their customer needs and shape the future shopping habits profitably, which is I think the keyword that’s been missing from immediacy offers to date.

We’re also innovating, as you know, with a series of investments that we’ve made over this half so that our partners can grow faster in new verticals as well. In May, we announced a £4.5 million investment in Karakuri.

It’s a minority stake in a robotic meal prep business, one that we knew for a period of time as our CTO had been serving as an adviser there for nearly a year before we made investment. We will be putting Karakuri’s first live production machine into our office soon, and we believe that it has very interesting possibilities within our larger warehouses next to our smaller warehouses, our Zoom warehouses and in our clients’ businesses.

So we think it’s a very interesting business. In June, we made two announcements, both in Jones Food, where we’ve become a minority -- sorry, majority shareholder and also the joint venture that we’ve announced, Infinite Acres, with 80 acres in Priva, both in the vertical farming space, two slightly different investments and in fact, a parallel to our online grocery business, Jones being the farmer like Ocado.com is the retailer and Infinite Acres being the supplier like the Ocado Solutions business and OSP is to the global industry.

We think obviously that technology is going to transform this space. We think we have a lot to bring in terms of some of the technology that we have already developed.

In fact, back in 2015 was when we first filed patents in the vertical farming space. So we’ve actually been working on it for a while, just not talking about it.

And obviously, huge relevance not just to society but also obviously to our clients. The most obvious places to build vertical farms are next to the most -- the largest distribution facilities for groceries to customers, which is what we’re building worldwide for our clients.

We see these as transformational opportunities with significant potential future returns. We are absolutely keen not to distract our core business, not to distract the focus on delivering for our customers.

And just to give you a clue there, on the left the little circle is to illustrate we have less than 10 people internally in the Ocado group that are focused on these investments and working with these partners. But clearly, we can leverage our technology expertise and some of our existing IP in other verticals to create significant, long-term future value.

The core business is one that has a long runway for growth. Our current partners already are a very strong base.

It’s a pretty interesting number that £195 billion is our existing customers’ annualized bricks-and-mortar run rate sales. And so as we see channels shift in that market, you can clearly see the potential for a sizable online business powered by the Ocado Smart Platform.

If 10% went online, it’d be a £20 billion business. But if 30% went online, it’d be a £59 billion business.

And a £59 billion business would be something like £3 billion of annualized fee opportunity. So that would be substantial just from our existing customers, but the global opportunity in grocery is obviously huge, £7.6 trillion, the global grocery market; £2.8 trillion in the key markets that we’ve identified.

If a quarter of that went online, you’d see a £700 trillion online market. So you can see there that if -- or if we took 25% of those key markets, that would be £700 billion of an opportunity.

If 10% of it went online, you’d see £70 billion online from our customers or up to £350 billion online from potential customers. So you can see the fee opportunity could range up to £15 billion to £20 billion in the existing model.

That’s before we further innovate towards lights out and human less deliveries, all things that we are working on in our 10x and other innovation departments. So in conclusion, as we’ve seen over the last couple of results presentations since we signed the major deal with Casino in the back of 2017, the center of gravity for our group has continued to shift and really has accelerated the shift over the last six months.

By partnering with Marks & Spencer, Ocado.com will now offer U.K. customers an even better experience than they’ve had for the last 17 years.

We can then focus on -- in the group on creating sustainable value for our solutions partners. For now, outstanding execution for those partners remains our priority.

But at the same time, we’re continuing to innovate, continuing to innovate at our part -- making sure our partners remain ahead of the game and our stakeholders can benefit from the future value creation. It’s an innovation house that we’ve created, and that’s absolutely at the heart of what we do.

Duncan and I are both ready to take Q&A. Difficult questions to Duncan.

You might bring a chair.

I’ll bring a chair.

Sorry. Sorry, go on.

It’s Bruno [ph] from Bernstein. My first question is you sort of said that Sobeys signed the second CFC ahead of plan, so there was a plan for more.

You know, I watched your interview with the Kroger CEO. You talked about the plan has more CFCs beyond the 20.

So clearly, they said as part of the contract, of your growth commitment, there is more to come. Would it be fair if I aggregate across all your deals that you’ve signed so far the kind of annual run rate for more CFCs beyond the ones you’ve told us already with probably about at least 5 to 7 for Kroger, adding up to possibly, with the other partners, closer to 10 CFCs per year?

As part of your committed growth plans that you haven’t disclosed, would that ballpark be the right level of additional CFCs from the existing partners?

I mean, look, I guess, what we’d say is that we haven’t -- we deliberately haven’t mentioned everything that we hope to do with those partners and they haven’t wanted to. There are not firm commitments for sites.

But if they don’t -- as we said before, that we’ve signed conditional exclusivity with a number of these clients, and obviously, it’s conditional on growth rates. So that’s the -- expectation is what’s required to maintain exclusivity.

Sobeys, as you know, ordered the second one prior to the requirement for them to do so to maintain exclusivity for Canada. We did say in terms of Kroger that it was 20 in the first 3 years, so that’s obviously about 7 a year.

And then if you add in some of the other clients, then your 10-plus a year doesn’t seem like an unsuitable number. Obviously, going forwards, we would hope to see an acceleration in our capability and in our customer demand or client demand for warehouses, not de-acceleration.

And I guess that links back to the dozens per year you just said before about -- capable of delivering. My second one is on Zoom.

We’ve tried it out. It is indeed quite an amazing proposition.

Help us with time lines; about how quickly could this become a London proposition here in the U.K. And how quickly could you think about signing up partners for that?

Are we talking end of year for the U.K.? Or is it a lot further in the future?

Look, I think there’s a couple of points to make. The first one is that the physical operation that we’re running today and all the software running today is not end -- well, it’s never endgame.

But it’s kind of a bit of a bitter test. It’s not yet in its -- kind of as mature as our existing facilities and software.

And so there’s ongoing work there that will really mature over the next probably one to two years in terms of what it’s going to look like and the supply chain software behind it and stuff like that. In terms of erecting one, it’s much, much faster than building a large shed because just physically much smaller.

And so it’s possible, if you can find a suitable site, to put one of those things up in a matter of months rather than over a couple of years. And so it’s something that we are talking to some clients about, and it’s something that we’re looking at second -- locations for a second facility here in London.

It would be super nice if I was in catchment so I could try the service because you’re a step ahead of me and I’m a little bit jealous.

[Indiscernible]. The last one is a bit more technical.

Sorry, Duncan. When I look at those cash fees you receive, should I presume those are like annual cash payments, therefore I don’t have to double them up to get to the full year?

So when you sign up a partner in H1, that’s a one-off lump sum and so doubling up wouldn’t be a good idea for those cash fees you’ve received in the first half linked to signing up partners? Or do they pay in quarters and half years, those people, for the preparation phase?

Yes. Unfortunately, it’s not that simple.

We typically get a signing-on fee for a new client. We typically get a commitment fee for a CFC, and then we’ll get ongoing annual fees post the CFC going live.

Now what you’ve got in the first half is some signing-on fees, Coles as an example. You’ve got some new commitment fees starting, and those commitment fees can have slightly different payment terms or will negotiate slightly different for each client.

So whether they’re paid in three equal sums over two years, in quarterly sums over two years, in two sums over two years can be slightly different. So unfortunately it’s not easy.

I can’t say predict twice in the second half.

But so assuming the current customer base knowing your deals, can you guide us how big the cash proceeds would be for the full year given that wouldn’t be new information?

Well, I can guide that you’re going to get a decent amount in the second half. Simplistically, today, if we want to count or something, just assume twice.

Thank you.

Anyone?

Hiding behind Nick here. Andrew Gwynn from Exane.

Let’s go through three questions quickly. I mean Bruno had about 10.

But on the gross margin guidance, you obviously saw pretty strong progression in the first half, but I think you said second half, it should be relatively stable. Could you just elaborate what the moving parts were in the first half and why it’s not coming through in the second?

Second, sort of bigger picture question. On CFC4, obviously a sort of test bed for how quickly you can ramp up new CFCs for partners.

So could you just share a little bit more as to maybe the sort of utilization of end capacity, how progression has been there? And then you sort of hinted a little bit what we’ll see basket size for Ocado Zoom, the average basket size.

I mean I appreciate gross margins better, but obviously the gross profit per basket presumably materially lower than it would be in sort of regular proposition. Thank you.

So let me just start backwards. And I might ask you to repeat your questions.

The Zoom basket, we’re not putting a number on it. We don’t want to for competitive reasons.

But I think what we have stated is it’s -- well, I’ve said it before. It’s about 30% larger than we were anticipating.

Most of the baskets are that £25 to £55, not £15 to £20, and the cash margin is slightly higher than the grocery business. Over time, we expect it to be higher than that.

Yes obviously, the total cash margin of the basket is lower, but so is every other aspect of serving that order. Although it’s delivered in immediacy on a courier, that’s cheaper than putting it in a van because obviously the distance is significantly smaller.

You’re only delivering to a single -- a mid-single-digit kilometer map radius maximum from the delivery site. So the facility is smaller.

And overall, as I indicated before, we believe that the increased margin and increased delivery charge as a percentage of the sale will match off quite closely to the increased cost of picking the increased cost of delivery as a percentage of sales and the increased waste. So we think that ultimately, it should generate a similar margin to a large facility with a slightly different service, different proposition.

So the previous question...

[Indiscernible]

So, CFC4 has continued to improve its productivity. It’s continues to improve its scale.

I think we stated that it reached volumes of around – it’s around…

Just under 50. So…

Thousand orders a week. So it’s continuing to scale.

About one-fourth of it is endgame. But I think more importantly than that is the absolute amount and the absolute amount relative to our – most of the sheds that we sold, which are materially smaller than it is.

So, there isn’t a client facility that we’d expect to ramp to bigger than Erith in its first year, so that we’re building at the moment. So, the fact that we can run them at that size, at that scale is very important.

And its productivity has continued to improve, hasn’t yet hit the levels that Andover was at. I think we mentioned the week before the Andover fire, it had exceeded the productivity of Hatfield for the first time and was – we still haven’t even implemented any individual tracking of productivity.

So whereas we would gain a significant amount from just being able to tally to the individual people, what they’re doing. We haven’t implemented that in favor of some of the internationalization work that we were doing and security what we were doing and stuff like that.

And so there’s a lot of scope to continue to drive up. We still expect Erith to be the most productive warehouse that we have with a significant increase in productivity over the levels that we’ve achieved live today in Dordon.

I can do the gross part. On gross margin, we saw a decent uplift in gross margin in the second half of last year.

This is a continuation of that and we’re not necessarily predicting, because we think the market will remain competitive for another, as it was, step change. So second half pretty flat.

I think just a word on supplier income. I don’t know you always hear me say it, but I really mean it.

Don’t expect a growth in supplier income in second half of the year, so – but 4% in the first half, if that’s replicated in the full year its still a big growth in total across the whole year.

Good morning, Nick Coulter from Citi. Three, if I may.

Firstly, on execution, could you just talk about the constructs of CFC project team? How you go about managing the projects?

Just give us a little bit more assurance on that the granularity of hiring an expertise that you’ve developed? And then comments on each of the warehouses on the construction?

And then I had a question on engineering costs within CFC for just to understand the evolution and how that’s progressing. I think you touched on it briefly, but didn’t really give any quantification?

And then lastly, on the robotic arm rollout, how you envisage that would proceed from here? Thank you.

Sure. So the first one is about how we build.

That we have an internal design team that can take a building that our client or we are looking at and very quickly translate that into an optimum design, which we can then put through our internal simulation teams to tell us whether it’s at -- whether it works, as well as we expect it to work, the viability, the scale et cetera, and any challenges that simulations show. One of the main concepts of our own solution is its modularity.

And therefore the kind of more simplicity of building because it’s the same equipment in every site that we’re getting better at teaching teams how to erect grids faster. The – we’re creating rigs.

But – some of the difficulties, for example, you put a pick station and you need to make sure the pick station is aligned with the top of the grid. If it’s slightly misaligned by few millimeters that you get some number of errors per thousand, and so as before that was being done manually, we’re creating kind of rigs that enable us to do that on site with less experience and more accuracy.

So at -- the more we’re doing, the more capable we are of actually building more and faster. But the whole key to those buildings is that they are much simpler to hand over almost to a third party and say, this is what has to be built.

These are the manuals that go along with how you erect something, how you line all up and stuff like that. And what we’ve been doing and I think we’ve mentioned it before is we’re working with outside parties that are very big in the global building construction world and creating teams that used to be solely Ocado U.K.

based teams with some Ocado people that are moving around the world and we’ve got the department busy with people’s work, visas and stuff like that, so that we can locate them in Australia, locate them in Canada et cetera and work together with the building contractors. So that if we used to have 12 people from Ocado, we might now have two in the area and 12 people from a third party, which margin increases our cost, but in the scheme of the whole building is not substantial and that enables us to build multiple buildings.

We’ve created a single team in Hatfield that will watch all of these warehouses. So whereas historically we’d have an individual team that will kind of watch them to finesse them and fine-tune them and first of all these are massively more digital in nature than the analog nature of the traditional infrastructure that we used in Hatfield and Dordon.

And so actually it’s all predominantly done in software. But just in case something isn’t looking right that all going to be monitored from a single central place.

So, you don’t need to be on-site to monitor because all that data is just streaming to us. And so these are some of the ways that we’re able to expand our capabilities.

Obviously, we’re growing our internal teams and training people all the time and we’re trying to go as fast as we can in that.

And on manufacture of robots, how you’ve been…?

So that’s actually an -- the actual manufacture setting is enormously easy to scale. You just need -- notice that you want to do it and that you’ve -- the correct lead time for the individual components.

The more challenging thing is getting to the robot that we wanted to manufacture 50,000 or 100,000 of. If we actually could lock it down today and say we were 100% happy with it, and we’re not 100% happy yet, but if we were 100% happy, you can manufacture them very quickly.

There are lots of people -- think about how many iPhones get manufactured and how much harder it would be to assemble something and fit everything in your iPhone that’s about that thick compared to building one of our robots. It’s the design of it that’s trickier.

And obviously, the design of it is benefited by our ability to have an increased investment, which we’re able to do because of the amount of robots we’re planning on building.

And I guess that’s a good segue to engineering costs at CFC4.

So our engineering costs are coming down. Our engineering capabilities are going up.

I expect them to come down still materially further. We have been putting through some fairly chunky changes on some of the robots as we’ve had king of -- as we develop design enhancements that we know will significantly increase their reliability.

Slightly weird when you do, of course, because what happens is you say, I know that this component causes me this much engineering costs and it shouldn’t. So I find a replacement to it that I believe will reduce that by whatever is 95%.

But then I actually go through a period of increased engineering costs whilst they swap out the component for the new component cost and put that cost through. So we’ve seen a few of those bumps on our way to significantly decrease engineering costs.

And the signs overall are very positive. As is always the case in a real business and not a spreadsheet, you often take 10 steps forward and you take one step back and then you take 10 steps forward one step back.

But the engineering, all looks great and we’ve built the first few of the prototypes of the new -- the next generation is kind of third main generation of robot, very significant change. That’s the one that we’ll roll out predominantly for our clients and they’re in live test at the moment and we’re building more of those right now and that program is coming along very well.

Then lastly this one on rollout of the robotic picking; how you see that over the course of the next few years?

Yes. So, look, the way works is that we’ve been doing research on it for almost seven years now.

So, it’s an area that we are quite knowledgeable in. What we now need to do is to turn that research into live sales and gain that amazing learning that you have from actually operating them.

The first one was unfortunately being turned on in Andover the week that we have the fire. So we lost it.

We’ve replicated it. In fact, we built more than one in Erith at the moment.

The first -- the key then is to take the 50,000 products that we sell and start to put them into groups and say, well okay; this group of products are kind of light. They have incredibly regular shape.

They have similar characteristics and they can be picked by a simple suction gripper. So let’s just take that.

They might represent five 5%, 10%, 15%, 20%, 25% of the velocity of picks, and let’s create some automated picking cells that will pick that part of each customer’s order. And then, as you know we’ve been involved in some EU funded and collaborative programs with institutes and universities across Europe in SOMA, the soft-hand manipulation.

They were quite successful projects. And so ultimately, they ultimately, that’s how you’ll pick grapes or something like that and carefully place them in a bag.

But that’s not the first one to knock off from a production perspective. So it’s just the constant kind of game of basically increasing our capabilities and increasing the amount of stuff that we’re capable of picking with automation.

The value of it varies. We’re probably in the middle of the pack of our client’s portfolio at the moment.

For some it’s insignificantly more valuable than it will be for us because of their different labor laws and labor costs. For some less, so but it’s an exciting year.

We just put – we’re just putting a significant increase in the amount of resource on our side that we’re putting into this. We’re pulling teams together that have been working on this in diverse areas of our business into one central team to accelerate our progress in this area.

[Indiscernible]

Rob, we’ll get to you eventually, sorry. Don’t forget your question.

It’s Sreedhar Mahamkali from Macquarie. Three questions really.

Just firstly, picking up on Nick’s question on engineering cost; I guess about year and a half ago to now, it’s about 200 basis points delta. It’s about -- it used to be about low 8s.

Do you actually see it going back to that level in about year and a half from now? Or how should we actually think about the phasing?

Because clearly, it’s not just important for Ocado Retail but also your customers. I mean just give us an idea in terms of we should think about that?

Over the next two to three years, we’d expect the engineering in our new facilities to start to look more like the engineering in our historic facilities, but initially at higher levels, I would say.

The biggest driver of the CFC cost increase was fixed costs of 120 basis points. Assume a 100 of it was fixed costs.

So it’s not an engineering point. UPH is, as we’ve said, lower in Erith today than it is in the combination of Hatfield and Dordon.

So those are two drag effects. So engineering costs are higher, but that’s not what’s driving CFC costs.

It’s a contributor, but it’s not the main driver. And the biggest driver of lowering CFC costs over the next couple of years is just using more of Erith.

That will be the biggest effect.

Got you. Okay.

Secondly, on Jones Food Company; just interesting to hear your thoughts on a three-year view or a five-year view. What -- is it vertical farming itself or vertical farming solutions like you’re doing at Ocado Retail solution?

Is there a play here?

I think we see opportunities in both, and that’s why we’ve made two investments as opposed to one. And the scale of both is early days for us to comment on, I think.

We definitely see opportunity to expand the farming operations themselves, and we definitely see very significant global opportunities in the Solutions side.

Okay. Last one, just in terms of Zoom, you’ve talked about looking for an additional facility in London.

But have you got a long-term view, five years out or something? Corresponding to your 12 CFCs, how many Zoom facilities could you have potentially?

So we haven’t sat with our partners. Our new Board is only getting formed next month to start looking at that, and the data from Zoom is only a few months old.

So I think what we’ll need to do is build two or three of them, gain more data and then work out whether we want 10 or 100. I think that remain -- or several hundred.

And that remains to be seen. There is no doubt, though, that this whole area of immediacy is of interest to our global clients.

And so making sure that we have everything available from store pick through kind of micro sites for immediacy, smaller warehouses, medium-sized warehouses, larger warehouses, dark stores, manual warehouses as a full suite of what we can offer our clients, so that they can deploy the right things in the right geographies at the right times as they grow and roll out their own businesses is a critical part of our future success, and it’s something that we’re very committed to. Can we get to Rob, please?

Poor guy. Because his hand is going to drop off soon from holding in the air.

Okay. Thanks very much.

You normally have an advantage as you’re normally on the home turf.

Yes. I know.

You put me away like this time. I’m sure it’s counts for double.

So, three from me. First one, on numeric simply.

It looks like from this slide on the opportunity for OSP. You are sort of confirming it’s around 5% fee there in terms of percentage of GMV, if I do the math’s on that.

How is that looking now you’re further through -- dropping through in terms of EBITDA percentage of GMV? And are we still looking at around the £45 million CapEx cost for each of the Andover size sites?

So I think you’re asking a few too many detail questions there. If you’ve deduced that to where you think it will, on average, be, then that’s your deduction.

Obviously, as we said before, the fees for a facility are based on its capacity largely in its peak hour because its peak hour defines how much we need to put down. Now obviously, if you don’t use the front end, then the hosting fees will obviously be lower if you only use it for one hour a week.

But a lot of the fees are based on the amount of kit that we put down and therefore will depend on individual -- the trading patterns of the different retailers as well as their average item prices. So in some markets, we’ve seen average item prices that are 30%, 40% lower than we see in the U.K.

We’ve seen other markets where they’re 30% to 40% higher. So that fee will vary by client.

And as I say, it’s not based on sales. It’s based on capacities.

The CapEx per site -- our share of the CapEx per site, we’re not going into the detail, but plus or minus, I think we said that Andover originally would be 12.8% or -- 12.8%, 13% of sales capacity, and plus or minus these are kind of sensible areas of -- or ballpark areas. Obviously, that excludes the enormous amount of R&D and learning that we’ve put into creating these solutions as opposed to creating an incremental £0.5 billion of capacity.

So I’d just add slightly to that two things. Tim mentioned earlier seismic.

So in a seismic CFC, it wouldn’t be at that level. There’ll be higher costs, but that’s clearly part of the commercials that we would discuss with our clients.

So you shouldn’t necessarily think that we suffer that. And secondly, that’s a sort of a U.K.

price. If you’re going to a location where there’s some additional tariffs, it might be slightly higher as well.

But again, those are part known costs that we’re -- when we engage with clients that we know about that, and you should expect a commercial deal that works for us there. But the underlying cost of the kit, yes, I think it’s still a good indication.

Broadly in line?

Yes.

Yes. Second was on CFC4.

So a little bit of maths. If I roll the current run rate out, it looks like you should be finished maybe end of 2021.

Given the demands there, given the clear opportunity in terms of operating efficiencies from that one, can you -- could you roll them out any faster than that? Or is that the right number as well, 2021?

The fabulous and challenging thing about Erith is, it’s certainly the biggest site we’ve committed to for anybody to date and probably the -- possibly the biggest site we will ever build. And so it is the site that should always be at the edge of testing the software and other kind of interactions of the building, of the quantity of robots interacting with each other and stuff like that.

So I’m not sure I would expect it to grow any faster than we’ve outlined before. It doesn’t mean that future sites need to scale to that level, as we showed when we turned Erith on and ramp it in the same number of weeks it had taken us that we’ve been operating Andover for in months.

So we ramp the two of them to the same volume in a matter of weeks post opening Erith. So the beauty of what we’re doing is that we could now ramp a site to the volume of Erith in a very short time.

So that gives us in the future better ability to decide to grow faster from a facilities perspective. But at the moment, to take it from 50 to 100, to 150 to 200, to 200 and something, there will be a lot of challenges to overcome along the way there.

We’re expecting them, but there will be a lot to overcome.

Sure, you will. And final one, just in terms of Zoom.

Just again, numbers. CapEx, have you given -- been able to give sort of rough CapEx-to-sales or CapEx on sales for that?

We haven’t. Look, I -- as you can imagine, when you miniaturize something, there are some things that don’t scale.

There are some things that scale linearly. There are some things that don’t quite scale linearly.

So we would expect to see a higher CapEx-to-sales ratio, a slightly higher operating-to-sales ratio and a slightly higher delivery cost-to-sales ratio, as we illustrated in the chart earlier. Having said that, we expect to see a slightly higher margin and a slightly higher delivery cost.

And we expect it to be predominantly incremental sales. And so overall, we believe that it’s possible to replicate similar levels of profitability and return on capital as we would see in the main business.

And does it -- just one. Does it fall under the existing exclusivity agreements it would sit within?

Predominantly. So we’ve got two.

One here, one here.

Thank you. This is James Lockyer from Peel Hunt.

Two questions for me, please. So supplier income in the half was 4%.

But given that mobile brings with it a smaller screen and hence potentially more competition for the eyeballs, with Ocado Zoom being, I’m assuming, virtually all mobile, more premium and arguably a focus on right now versus brand preference, can you talk about what you might be able to do there with respect to supplier income? Could it be higher than you are getting with Ocado.com?

So James, I think you’re touching on an important area. And you’re right, the smaller screens of mobile mean that the retailer is more influential in the purchasing decision than they were on a browser, where they were already more relevant than they were in a store-based environment.

I think that yes, I mean, Zoom may have more in terms of slightly more limited range, and people’s desire to shop quickly means that they’re less likely to spend a long time browsing when they actually want the food in 30 minutes at their door, and that has some opportunities. Given the relative scale of it at the moment, it’s not a short-term thing because we’re not going to go to the supplier separately and say, look, we’ve got a site over here doing a few million pounds of sales.

Can we get a larger income source for it? But over time, when that became a £1 billion business, then you can imagine that there may be some greater opportunities there.

Having said that, the direction of travel in the main business, both mobile and on browser, into less catalog searching, into more algorithmically predicted, whether that’s because I’m predicting meals and -- or ingredients or suggested orders or stuff, the more we’re predicting and saying, would you -- this is what we think you want to buy, the more influence we have. Because we don’t say, we think you want to buy butter.

How about you choose between the Lurpak, the Anchor and the President? We just simply put butter in their order, and we choose which one of the three-plus-so label to put in there.

And therefore, there’s more opportunity for us to influence what that shopping is anyway. So there is definitely a long-term trend.

And I’m emphasizing the word long term to emphasize Duncan’s point to say don’t expect that number to grow in the short term. But there is definitely a long-term trend here that retailers that are operating digitally with their clients like this have got more influence over purchasing decisions than bricks-and-mortar retailers have over the last 20 years where the power has dramatically shifted to the brand owners away from the retailers.

That’s coming back.

Second question just on the 255 new software heads that you hired in the half and the announcement of your new London hub; I was wondering if you could talk about that a bit more in terms of what the focus is there? Whether you’re seeing that speed of change increase because of obviously your tech credentials are now much more obvious than they have been in the past?

Look, we’ve done a good job, I must say, in the past in the tech community of pushing our credentials, and we were being successful at hiring outstanding talent in all of our tech offices. It’s always been a focus of ours, and we’ve always looked at quality over quantity.

So there were times when we would like to have added heads at twice the rate we were. But to get the right quality, we went slower.

And ultimately, that tends to drive a quantity later because you become known as the place to try and get a job because it’s difficult to get through. And we’ve seen that quite prominently in some of the markets that we operate in.

The London hub is about just helping us have the ability to increase our headcount. We often see some very talented individuals who really like what we’re doing and would be interested in working on it with us but don’t want to -- are based in London and don’t want to work out in Hatfield for their own lifestyle reasons, for others it’s an ideal place.

And so we decided that we’d like to have the opportunity to hire them as well, and so we’ve opened a hub in London. We want to keep increasing the headcount at the moment.

We may, at some point, put a stop to it. But at the moment, we’ve got an increasing demand for development to hit time lines that we promised to our clients, to migrate the main Ocado.com business off of the -- the majority of the kind of traditional software and onto the same platform as our OSP clients.

And so certainly, for the next couple of years, we’re extremely busy. And with our new organizational structure, we are seeing a close relationship between the heads that we’re able to bring into the organization and the throughputs that we expect -- we think we’re getting out of it.

Thank you.

It’s Simon Bowler from Numis. Two questions from myself.

Firstly, there’s a bit of step-up in kind of marketing across the first half. Just wondering if you can talk about the nature of that and any learnings that came through from that, you mentioned kind of off-line trials within that.

And then secondly, thinking to kind of Ocado ventures, how should we think about those, the pots of money that are sitting there to be spent? Or how do you think about the scale and potential of what you can spend in that area?

And I guess the examples we’ve seen thus far are, to some extent, technologies that can be incorporated into your fulfillment centers. And then is that a good way to think about what you’re ideally looking for when considering those investments?

Or could it be more far reaching than that?

So let me...

So I’ll do the marketing.

Sure.

On the marketing, the marketing is an offline trial. Research is ongoing at the moment.

So the search is not complete. And so nothing to say too much.

And I did mention there was some canceled acquisition costs because immediately post fire, there was no point doing acquisition marketing when you couldn’t satisfy the demand; and some increased retention costs because part of the benefit that we’ve had of recovering demand is spreading the demand across the week, and what that’s doing is encouraging some customers to shop when they ideally wouldn’t shop. So we are moving out the peak days into other days in the week, which cost us a bit more marketing.

So those, I think, are two effects that are not long term. In fact, all of the first half performance is not indicative of the long term.

Sorry. And...

Look, on the ventures front, we have -- over the last few years as we’ve been inventing more and more and we’ve been patenting, we sometimes see that something we invented for the online grocery space has significant application outside of grocery if it’s amended in some way. Hence, we filed patents as diverse as indoor farming, I think, at the same time in something like car parking, for example.

And so we have the ability to create spinouts that are from our IP, but they may be not connected to our clients, although, funny enough, car parking again is something that we always sure of every location, so -- but they may be unconnected to what we do. Some of the stuff that our ventures teams are working on are not connected, but are connected to the IP that we’ve created.

And the other area that we’re interested in is where we haven’t created the IP, but we can see an application of it with -- either within or adjacent or with our clients. So those are our main areas of focus.

We’re not -- we’re unlikely to come to our Board with a completely random "I’ve seen a really good investment over here" kind of thing. There’s going to be some application to what it is that we’re doing, but it could be something that makes a core component for robotic pick or it could be someone has got some software for something or it could be someone that’s just trying to innovate and it’s an early-stage startup.

And -- but in a field that we don’t feel that we have -- that we could create the teams to do it ourselves at the moment and have the time for the focus, but we still think it’s very interesting, which is more of a Karakuri, it’s relatively early stage, but with a good team in an area we think is really important but isn’t one that we can get to at the moment, and, therefore, we’d rather be an investor than not participate in the area. At the moment, as you say, relative to our scale, the investments to date have been relatively small.

I expect they’re likely to continue that way for the moment, but that we -- and Stuart over there has not -- we have looked at things that are quite substantial should we have chosen to attempt to acquire them. But so far, when we’ve looked at them, we’ve said they’re not worth anything like the amount of money that these people want for them, so we’re not interested.

So we haven’t been shy to look at things that are 10 or 100 times the size of some of the investments that we’ve made. We just haven’t found that.

We’re not desperate to do it. It’s just it -- if there was something that we really felt we should buy, we wouldn’t be shy to think about it and to look at it.

But it’s not something we’ve seen to date and we want to do.

Thanks.

Morning. It’s Andrew from HSBC.

A few, if I may. One for Duncan, first off.

Could you just remind us how many customer CFCs you’re incurring startup losses at the moment? And give us an idea on how that should evolve going into 2020?

Second one was on the JV. What sort of -- what changes are you making to the Retail business to sort of prepare it for going into the JV?

And does that affect the financial profile of the Retail business going into that? And the last one really, why Purfleet?

Given the proximity to Erith, your biggest facility, can you give us some idea on the sort of logic behind that and how you’re thinking about CFCs going forward?

So let me answer the last question first. And someone in the press asked me the same question earlier.

They also had their map out. As the crow flies, it’s very close.

But the crow flies and doesn’t go across land. And in between Purfleet and Erith is the Thames.

And crossings across the Thames on the east side of London are poor. And not just poor and slow, but very unpredictable as well, it can sometimes be several hours, an amount of time that you could have driven from Hatfield to Birmingham and quite easily sometimes.

And therefore, Erith has been a facility that we’ve been using to serve south of Erith, and Purfleet is a facility that we’ll use to serve north of Purfleet -- or north of the river and south of the river. And so when we look though at expansion, it’s not the only facility that we need for the next 5 years.

And therefore, you can say, oh, that’s weird. I would have expected them to go further north or something like that.

We’ll need some facilities further north. We’ll need some facilities further south.

We’ll need some facilities west. And therefore, it’s actually -- you then look at where do you need them.

And then it’s more -- it’s a bit tactical around what is the availability of sites that I can turn around as quickly as I want to turn it around given the planning, given the condition of the site, given the expected negotiation with the landlords. There’s a bunch of considerations.

And so sometimes, what might look to you to be, I would have done that one before I did this one, it may actually be that just timing-wise, I can do this one before I do that one. And the difference in the long term is irrelevant, and the difference in the short term might be a bit of extra transport costs in one year versus doing it the other way around, but one’s doable and one’s not actually doable.

So it’s a little bit like that. So I wouldn’t view it as that looks like a weird one to do.

I think you need to look at and think about the next three or four that we’re going to build and how it all fits into a bit of a picture around where our customers are. And I think we’ve said before around 50% of the business today is inside the M25 and around 50% is outside the M25.

One of the advantages of Ocado.com and its joint venture state is that the Marks & Spencer store business is massively more geographically diverse than our supply partner at Waitrose, so they have much more presence in the north up in Scotland, for example, than Waitrose do, so much higher levels of brand recognition there. And therefore, in the future, we’d expect to see more of an evening out of our geographic spread, which may lead to other different decisions in warehouse locations.

The middle question, just remind me again what that was.

Just around any changes you’re making to the Retail business as it goes into the JV in terms of infrastructure.

Right. So basically, we are separating it out as a legal entity.

Obviously, that means that the JV is going to have its own Chief Executive and run its own P&L. But it has over -- it’s about 300 people, so most -- really the buying organization and then some associated of the support -- some finance, some HR, planning and stuff like that that’s gone with it and a limited amount of things like technology.

And it is a buyer of logistics services from group. So we have announced that we created Ocado Logistics that’s serving both the JV and serving Morrisons because the drivers that turn up in the Morrisons livery, much as the drivers that turn up in the Ocado.com livery are from Ocado Logistics that remains a wholly owned subsidiary.

So there’s not much change, to be honest. There is a small amount of -- where there used to be 1 person, now there’ll be 1 doing something -- group being kind of with a relationship person in the JV.

But there’s not much of it. And the first question was on the sort of costs to CFCs prior to open.

There’s a number of costs as -- such as the -- you talked about startup losses. Actually, there’s only a number of CFCs that are in that phase just as it were prior to opening.

So if you think about it, you should expect Casino and Sobeys they are both set to open in the first half next year. So we’ll have people on site starting to install kit.

For other committed CFCs, if you’re in the effectively, call it, the first year, we have some people in our office, as Tim was talking earlier, around designing them. We don’t effectively have people on site, but the biggest cost in Solutions today is the cost of developing the platform.

And about a third of our development cost is not capitalized. So that’s our cost just flowing straight through the P&L.

So developing the capabilities that, that business will use. So that’s the biggest cost.

You’ve got effectively the central and support costs for signing deals and signing new deals. There’s a sort central cost associated with that.

And then you have got some, as it were startup losses because we already have some tech environments live for Casino and Sobeys so that they’re able to do some testing, some integration testing, start building the interfaces on their side as well. So there are some straight startup losses but only for a few CFCs, not for the scale of the program ahead yet.

Yes. Tony Shiret from Whitman Howard.

Just you’re discussing Marks & Spencer. Historically, they’re not sort of very good at big projects.

And I just wondered how you’re -- or whether you anticipate at this stage provisioning against any friction costs when you change over the suppliers. And I think also whether you’re planning to boost marketing to the transition on to Marks & Spencer product, those types.

These, I know, are sort of not this financial year but maybe give some thoughts on that? And much, much longer-term question, it just sort of occurs, looking at the amount of first CFCs you’re going to have with a load of clients over time, that maybe there might be some other things you could do with those clients in terms of buying and stuff like that and having a common Ocado pasta globally or whatever.

Have you thought about that?

So look, I think there are obviously going to be other opportunities with the client base to look at ways that they can all benefit from being in a club together. That’s not just about sharing our technology platform.

There’s a lot of learning as well, which I think is more important than common pasta, to be honest, now -- or operating in these different markets that they can all benefit from. The -- and you know the M&S business well.

I think that it has quite significant seasonal peaks today. And therefore, their supply base is very geared up to an increase in volume that they’re very looking forward to.

And I spent some time myself with, I think, probably their largest supplier last week discussing some of that opportunity. And so we don’t see a significant cost.

Obviously, there will be some costs relating to winding down the inventory of one lot of stock and bringing the other in. Marketing-wise, obviously we need to keep our customer base aware of what we’re doing.

But I think it’s worth remembering that we’re interacting with our customer base on a very frequent basis. We obviously talk to them digitally at very low costs.

So, most of our marketing money is spent in getting new customers to trial our service. One of the advantages of the new joint venture is that we’ll have a partner who wants to direct that traffic to us as opposed to where we are today where Waitrose doesn’t spend any time or money or effort directing customers to us because they want to direct them to a different service.

And so actually, the marketing opportunities to -- are there to reduce cost going forwards. We’re very confident about the working relationship with M&S and achieving any change that needs to happen in the time frames that we have.

And obviously, it’s all -- we call it a joint venture and it is one because it’s a 50-50 owned business and the Board has got 50% representation from both parties. But of course, it’s actually M&S buying half of an existing business.

And so I think when the business is being created, we’re tipping over, I think, 300 people. There will be less than a handful that will be -- would be moving over from M&S into the business.

So I’m -- we’re very happy with the way that’s going to operate.

And so no material provision?

Not expecting one.

No. I mean I think the idea of spending some more marketing costs around the point of transition is clearly an obvious one.

But I think Tim is right. Long-term marketing cost is perhaps more of an opportunity rather than the short term.

And it’s in -- Waitrose interest, it’s in the supplier’s interest for this transition to happen successfully because the supplier certainly doesn’t want a disruption. Waitrose certainly doesn’t want a disruption.

M&S doesn’t want a disruption. So there are clearly, in the week or so around it, may be some issue.

And we’ll -- we can probably say more closer to it but no. Overall, I think everybody will manage this quite sensibly.

Anymore? No?

Thank you very much all for coming this morning.

Thank you.