Naked Wines plc

Good morning, everybody. Welcome to the 2019 Results Presentation.

A bit of a reversal of roles this morning, I'm going to start and then hand over to Rowan rather than the other way around because it's not really the normal results presentation we're going to work through this morning. You will have seen in the announcement that we plan to exit Majestic Retail and Commercial, so we split this presentation into 2.

There's a quick kind of run-through the group reporting, which I'll do; and then we're going to very much focus on Naked in 2 parts: The first being an overview of trading, which I'll take you through. And then I'm going hand it over to Rowan to talk much more about -- in a forward-looking way, about the future.

So I'm going to jump straight into the review of the '19 results. A few highlights and lowlights as we see them.

Highlights are that we got revenue growth out of each of the businesses. Naked drove an acceleration of that.

Naked repeat customer contribution, which is the real measure of the growth of the kind of core of the business, was up GBP 6 million to GBP 40 million. And we invested GBP 5 million more in new customer acquisition year-on-year to GBP 19 million and delivered our payback target of 4x.

What I've referred to as some lowlights. Retail gross profit declined by GBP 2 million driven by a gross margin decline.

The fixed cost growth in Naked was ahead of the guidance we gave you in April as we invested earlier in some additional roles. And then some something worth walking through, the cash flow numbers year-on-year look very different but there are some very good investment reasons and phasing reasons for that, which I'll touch on.

So looking at the group revenue lines. Group revenue was up 5.8%.

That is an acceleration from 4.8% in the first half and quicker than we delivered in FY '18 as well. So momentum's moving positively.

And every business unit's seeing growth, and 3 of 4 have positive momentum, too. So Naked has accelerated as a result of the investments.

Lay & Wheeler's had a better year than it looked like at the start as en primeur started off very slowly. And we said 6 months ago that Commercial was possibly turning a corner.

It continued that trajectory in the second half in what remains a very tough market. That translated through to an EBIT reduction of GBP 6 million for the group in the year.

That's the net of a GBP 6 million increase in repeat contribution in Naked less the GBP 5 million uplift in new business spend that we put into the business. The GBP 2 million reduction Retail gross profit, and it's worth it just pausing for a second on some of the drivers of that, but that's a mix of some investment in things like footfall, where we discounted in vouchers to get new customers and existing customers into the stores; driving people into the subscription programs, Concierge in particular, which is at 33,000 people registered on that at the end of the year; and establishing our own brands and very specifically discounting certain brands to start building them within the portfolio and actually seeing some payback on that as we have moved customers into items where we now control the supply chain.

And then the last big driver is the fixed cost increase of GBP 5.5 million across Naked and the plc. We signaled a lot of that at the Capital Markets Day, putting investment into growth teams, optimization teams and the controls and compliance agenda.

And we've got a little bit ahead of that in terms of bringing some of the other roles into the business quicker. So just talking about those fixed costs and to give a little bit of color and a little bit of reassurance.

The fixed cost growth at Naked has gone into the U.S. predominantly.

So we're getting the infrastructure into the market that we continue to see as the biggest value driver for the group. We strengthened the team ahead of some plans around building global winemaker relations, and I'll talk in a moment about some of those that we've made in the last year.

Building kind of more strategic partnerships in the U.S., so kind of large-scale arrangements, be it with airlines, credit card companies, things like that. And that really has brought forward some cost growth that we would have seen next year, such that next year actually across the plc and Naked, the cost growth is expected to be possibly a little bit less than 10%.

And we're guiding kind of 10% to 15% just to give some headroom there in case there are some other opportunities that we see. Should we sell the Majestic business?

There is a net saving, and Rowan will talk through some of the moving parts there in a moment, of GBP 1 million to GBP 1.5 million that will come out of the central cost base. So we just want to kind of leave you with the idea of we know it's gone on quickly but be reassured that it's not going to continue on to that trajectory into fiscal '20.

We're reporting a significant statutory laws today, GBP 8.5 million, and that's driven by the adjusted items that sit below that adjusted EBIT number I just bridged through. The biggest number in there relates to impairments to the store base in Retail.

I'm sure you all know that you have to do an annual test of cash generation and the carrying value, but there are 2 sets of drivers to that impairment charge. About half of it is that we have put a weaker trading outlook in for Retail than we put in a year go into that model.

The second half, which is just worth mentioning, is because we fulfill web sales through the stores in parts, historically, we've included those web sales in the valuation of the stores. Given the shifting channel mix, the increased use of national fulfillment, we took a policy decision to take those sales out of the stores in running this model.

And that's generated about half that impairment charge as a result of that. Quickly turning to cash flow in the balance sheet.

Net debt ended the year at GBP 15.5 million, about GBP 7 million ahead year-on-year. That's really as a result of the investment choices we made.

So the free cash flow out of the business was about GBP 2 million, but that is after GBP 5 million of increased investment into Naked new business. It is after a stock build because of the timing of Easter and there's mitigation against Brexit disruption of GBP 8 million.

It is after a build of about GBP 6 million of stock in Naked ahead of future sales growth because of the investment up-weight we're making next year. And it is after putting GBP 2 million into a new EPOS system across the retail estate.

And after putting about GBP 3 million into shelving at the retail estate that unlocks productivity savings in store. So there's definitely a view we could have taken to invest much less aggressively in the business and pay down all of the debt, but from a capital allocation perspective, we think the right choice has been to invest the money.

And the balance sheet still supports that growth at 0.8x adjusted EBITDA-net debt. We also announced we would be suspending the dividend contingent on a sale of Majestic.

We will then pay special to the same level as last year's final. That is just to be consistent with the guidance that we gave in March around waiting till we get clarity on the transformation outcome before we put a firm policy in place.

So that's really the group in a nutshell. Growth accelerated, fixed cost ahead but tempering in fiscal '20.

Retail growth, still getting growth out, but margin under pressure. And the balance sheet's in a position where it supports the continued growth investment plans we have for Naked.

That's all I'm planning to say about the group position. Now I want to move forward towards the focus on Naked.

And we'll do that in 3 parts: A very quick word on sale process for Majestic in order to focus on Naked, we have to get that cleared first; talk through the trading that we've delivered in Naked during fiscal '19, that's firmly on track; and then hand over to Rowan to talk about the future. So Majestic.

Goal is to focus on Naked, the goal is to exit Majestic. As of today, we believe the sale of Majestic Retail and Commercial business is the best way to deliver that.

We've been through a period over the last 3 months of testing the market, multiple rounds of offers. And we're currently at a fairly advanced stage of that process with multiple parties.

We aim to reach a conclusion over the summer months. To the extent to which we can't reach a conclusion over summer, we will pause that process in order to protect the business Christmas trading period from the kind of disruption that type of process naturally causes.

And then restart the process in fiscal '20. That's about all I can say on that, as you'll understand.

So moving on to Naked. What are the headlines about the Naked's trading in fiscal '19?

Repeat contribution's up, EBIT's down because we've invested more in growth and we've put the overheads in. Payback's on target at 4x.

The KPIs that support that performance is stable or improving. And the rate of new customer investment continues to accelerate.

So that's pretty much what we said we'd do in the Capital Markets Day in April 2018. 18% increase in repeat customer contribution comes as a result of the investments we've put in, in prior years and the steady KPIs we've seen this year.

GBP 5 million more investment I've talked to, net EBIT down by GBP 1.9 million. Where is that coming from?

The U.S. is driving that growth.

It's both the biggest market now; it's overtaken the U.K. in the course of the last year.

And it's driving the growth in both contribution, which overall for the group, has grown 2.5x since it was acquired by Majestic in April 2015, and also at the sales line. So the U.S.

grew sales by 21% this year, 8% increase in the U.K., 16% in Australia. So we're in the nice position where our biggest business is also our fastest-growth business, which gives us kind of good structural tailwind in terms of the future growth trajectory.

In terms of all that, how that all translates in the EBIT line. We show this chart every year.

I make no apologies every time that we do it. But a reduction in EBIT of GBP 1.9 million is the net of GBP 4 million increase because we've got more customers.

That's some of the return coming from the investments we made in fiscal '18 being seen in fiscal '19. Increased contribution per customer.

So if we look per customer per month, we are getting more contribution out of each one, and that is a function of the quality of customers we're now recruiting. And we've spoken about a lot moving towards higher-quality recruitment.

That is where you see that start to come through on that chart. GBP 5 million more new customer investment, that's going to drive growth into fiscal '20 and beyond.

And then the GBP 2.9 million increase in fixed cost that I addressed in the earlier section. The KPIs that we run the business on are looking solid.

So we've talked about the increase in new customer investment. We said in April 2018 that 4x payback is what we will be targeting.

We hit that in H1, we hit that for the full year, that is the number we're currently operating at. And the sales retention has increased from the H1 number, 78% to 81% for the full year, which means H2 was 83%.

That was number I think people saw and thought that it had got materially worse and that was a structural issue. I'll show you in a second how that looks monthly versus on a long-term trajectory.

So back to new customer quality. We keep mentioning this and we show different ways of seeing it.

But this chart is kind of showing us how the fiscal '19 customers we recruited are delivering on average 6% more contribution per angel that we recruited than the ones in fiscal '18. And what you see in this chart is you've got the cohorts along the bottom.

So a cohort with -- at the end of the year, with 1 month of tenure, 2, 3, 4, 5 and so on. And every single one of those 12 monthly cohorts we recruited during fiscal '19 has performed ahead of the equivalent cohort in fiscal '18.

What are some of the underlying KPIs that are helping that? Well, on the left-hand side is the sales retention chart I just referred to.

So you've got the actual monthly and then our rolling 12-month sales retention. And you can see that from month-to-month, we see sales retention numbers bounce around all over the place.

Sometimes it can be greater than 100%, sometimes it could be down in the low 70s. That tends to be a function of things like seasonality and promotional phasing.

But when you turn that into a long-term trend in a rolling 12, it looks remarkably stable. And there are 2 things that impact that longer-term trend.

One is the mix between new and old customers. New customers tend to have lower retention than older customers.

So as we increase the amount of new customers we recruited, you actually get a little bit of downward pressure on that metric. The second thing is the mix of countries.

The U.S. and Australia both exhibit somewhat lower retention than the U.K.

That's a function in part of them having younger customer bases who have not matured to the same point. But as the U.S.

grows in particular as a proportion of that, you see a slight downward pressure on that number despite underlying improvement. For those with very sharp eyes, you will observe it's probably dropped about 1% to 2% across the course of that 4-year period, and that mix effect is what's driving that.

On the right-hand side, the second thing that's driving improved customer economics is we continue to improve the contribution margin on repeat orders. And you can see kind of back in fiscal '14, actually we were operating a number of businesses at what I would consider to be kind of sub-optimal scale.

As we have grown those businesses, you start to see the benefits of scale coming through the fulfillment cost line and also the purchasing line for the wine. And that has improved the contribution margin over time.

And then getting to the final level below those KPIs into our operating KPIs. What we want is people to like the wine they buy.

We measure that with the buy it again rating which is stable. We want to have the wines that they like available to them in all of our warehouses, which we measure with availability.

That has increased from 90% to 91%. And we want them to love the service we give them, which we measure with a 5-star service rating, which has stayed stable at our target of 90%.

So the business is still delivering the proposition that we set out to deliver to our customers. Couple of ways that we help do that.

One is through investment in improving service levels. This is where our 4 U.S.

distribution centers are. We opened a fourth one this year, which is the one in the middle, in Missouri.

That now gives us the ability to get an order through to our U.S. customers, 98% of them within 48 hours.

Quite often, when you sit in the U.K., that doesn't sound very impressive. But for a U.S.

service level, that's actually best-in-class. It also supports margin improvement when we truck wine all the way to Missouri in 1 bigger truck, it costs an awful lot less than putting all of those cases on FedEx to get there.

So part of that driver of that margin improvement is as you hit certain scale thresholds, you can unlock another level of fulfillment savings here. And the second area, we continue to evolve the portfolio.

And as I said, we put some investment into dedicating someone to building this. This is the winemaker pipeline.

A number of names on here you might not have heard of, but David Seijas at the bottom left, he was the sommelier at El Bulli, the very famous restaurant. He's now making wine for us.

And at the bottom right, as an example, Penelope Gadd-Coster, is I think known colloquially as the queen of sparkling American wine. But she is a very, very well-renowned sparkling operator in the U.S.

She is now making her own wine for us. Everybody else on this page has their own story which is there, but I won't walk you through them all.

So hopefully you can see that we are building the foundation KPIs that delivered the growth metrics around retention and payback that then deliver the growth in business. That's what we've done this year, now I'm going to hand over to Rowan to talk about how the future shapes from here.

Thank you. So that was the past.

Now let's talk about the future, and hopefully get the future going the right way. So 3 things to say about it.

The first is we've got a very solid core business. If we just take the customers we've already paid for and recruited, we've got a business that would make GBP 16 million on standstill basis right now.

But important thing is that growth is embedded in that. So the level of spend required to maintain a business making GBP 16 million, we're spending more than twice as much as that.

So if we don't increase the spending, the inherent rate of growth in the business remains strong. We don't need to increase spend to increase growth, but we are planning to.

. So let's take each of those in turn.

So core business is solid, the best way to measure this is on a standstill EBIT basis where we're only taking the proportion of new customer investment required to maintain sales retention. When you add that back, you can see that it's growing pretty consistently over the years.

In the future, we are -- this is going to have to carry all of the plc cost as opposed to having that split with Majestic, and I'll come back to that. That growth in steady state is driven by continuing growth in repeat customer contribution.

This is the engine that fuels the rest of the business. And we point more and more towards contribution as the key KPI rather than sales because this number tells no lies.

Sales can grow, sales mix can change, sales can move around all over the place. There are very different levels of contribution in different markets, but the engine that drives the business is contribution from the existing customers.

So that's grown to around GBP 40 million this year. And that in turn has been driven by a growth in new customer investment.

And that's grown from around GBP 8 million at that time Majestic acquired Naked to GBP 19 million in the last -- in the prior year, and the current run rate is at around GBP 26 million. Now the important thing I want to point out is that GBP 19 million that was spent in the prior year, almost all the benefit of that will come through in the next year.

So there tends to be about a 1-year lag between investing the money and that appearing in the customer contribution. Okay.

So what are we going to -- what is our approach to capital allocation? The first thing to say is, this hasn't changed.

We've consistently said we want to run a healthy balance sheet, we want to prioritize growth and we want to return surplus capital to shareholders. So what we're trying to signal in here is that once -- assuming the Majestic deal is completed, there is no intention to go and do acquisitions or to fundamentally change the business.

We intend to stick to the business as it is. If we talk about the growth side, our investment in growth is driven by our ability to find places to put the money.

And we are not going to get ourselves into the position where either we are constraining that growth to hit an EBIT target or we're desperately looking for growth opportunities because we committed to a growth spend. The nature of the growth is it's going to not be a straight line.

And the reason for that is we try lots of different things all the time. We have hundreds of tests running.

If they work, we back them and back them further. If they don't work, we don't.

So looking forward, the rate of growth, we're looking to accelerate it but we're only going to accelerate it where we can find and prove opportunities. We're not going to force that to try to hit some growth number and we're not going to constrain it to hit an EBIT number.

Okay? So what does that mean?

Over the next year, we think that the investment to new customer acquisition's going to grow from GBP 19 million to around GBP 26 million. The reason for that is there is just underlying organic growth in our existing channels of around 15%.

Digital marketing has been growing faster than that. And we are testing a number of other channels.

And its early days for them, but we will let you know how that goes. Overall, we expect around GBP 26 million in 2020.

And clearly, we think that will grow in the future but we'll tell you more as we're able to prove out those opportunities. The second thing is we want to be investing more money into our intelligence: AI, data science, IT.

And 3 things we wanted to do with that. The first is automation.

So in a rush from -- to get from a startup business to where we are today, there are a number of processes that are still done manually that would be better done by machine and free people up to do more interesting things. We want to be able to put the energy into doing that.

Optimization, reducing friction. We're still at the stage, even though we're 10 years in, where we're getting significant improvements by optimizing the funnel, for example.

And we haven't been able to do that as much as we want to have over the last 4 years because most of the IT energy has gone into Majestic. We now want to get that energy back into Naked and really optimize the funnel.

And finally, innovation. We want to find opportunities to improve our economics.

And you should look for that not as increased payback, our target is 4x. If we can improve the economics, we will try to find more opportunities to invest at 4x.

Okay? So we want to go smarter and we want to go faster.

So what does that look like on a pro forma basis? From a P&L point of view, EBIT next year on a pro forma basis, we expect to be about the same kind of number.

Largely, that's the 2 factors there are -- is this -- no. The 2 factors there are the repeat contribution, we expect to grow and we're pretty confident about that because we've already acquired those customers; but we're also increasing the amount of money we're investing in new customer acquisition.

Below the line however, we will be carrying a significant chunk of plc costs. Last year, that was around GBP 10 million.

We expect we will lose about GBP 4 million that and we will keep about GBP 6 million. And we're making a GBP 3.5 million decision to increase those to acquire new capability.

Just to be clear, some of those are the same people, right? So we've gotten one IT team that's been serving both businesses.

We've got one digital team that's been serving both businesses. We want to keep all of that resource and now put that into Naked.

But there will be some new heads and there are obviously some heads leaving with Majestic. On a standstill EBIT basis, at the bottom line, we expect EBIT, if we're doing it in an apples-for-apples basis, to grow to around GBP 21 million next year on a standstill basis.

Net of the cost, that will come back to GBP 11.6 million. If we weren't making the investments in capability, that would be something closer to GBP 14 million.

So that's the pro forma. Looking at it slightly longer term.

Again, this is the chart we showed you before, growing to GBP 14 million -- sorry, ah, there we go. So as you can see, we've grown invest -- contribution from existing customers up to GBP 40 million, right?

Every year, you can see the new customers have added around GBP 10 million or GBP 11 million of additional contribution per year. That in turn has been driven by this level of spend.

So you can pretty quickly start seeing the correlation between spend GBP 10 million get GBP 12 million; spend GBP 14 million, get GBP 11 million; spend -- sorry, spend GBP 14 million, get GBP 12 million; spend GBP 14 million, get GBP 11 million. We spent last year, GBP 19 million, so we expect this to go up by about GBP 13 million next year.

And next year, we plan to spend GBP 26 million, and you should expect to see exactly the same thing happen. So if you -- I'm sure you can all extrapolate this out for yourselves, but very simply, if we just keep doing what we're doing, the GBP 40 million goes GBP to 80 million over around 5 years.

So we're now saying that we have growth embedded in the business. But point was, we are spending at the rate of GBP 26 million already.

We don't need to find new investment opportunities to hit GBP 26 million. We're getting 4x payback.

If we keep investing GBP 26 million at 4x payback, this number doubles over about 5 years. All right.

So the other part of the picture is the -- I've heavily flagged the U.S. is our biggest and most important market, it's where we expect the most growth in the future.

Therefore, we are reorienting the Board around the U.S. So first of those is the appointment of John Walden as Chairman-elect.

I'm sure many of you know John, he was CEO of Argos. And aside from being an American, which should help, he also is someone who brings great insight on the customer side and will help rebalance the Board in that direction.

Katrina Cliffe has been at American Express for many years and has deep understanding of the subscription business. And Nick Devlin, who's sitting in the front over here, is the person who took our American business and took a potential business and made that potential real.

And so we're bringing him on to the Board as COO in order to make sure that we put the energy and emphasis on the American business that it deserves. Greg Hodder will be stepping down as Chairman.

Greg is going to stay on the Board for 6 months. We've been asked that question, were there any strategic differences or anything else?

Absolutely not. Greg is very much a U.K.-focus guy.

Our future is U.S., we want to rebalance the Board in favor of the U.S. So the future is Naked, where we want it to be a simple business: 1 brand, 1 business model, all of our focus and attention into driving that business with the resources to be able to drive it properly.

With a very clear focus, which is profitable growth by going faster and going smarter. And on track for growth, which we've got a solid core business underlying this.

We have growth embedded in our current rate of spend and we can see a way to increase spend in the future. Thank you very much.

Ben Hunt from Investec. Looking at the business, assuming that you sell Majestic, it looks like you'll be, in FY '20, loss-making to a tune of about GBP 3 million.

If you want to continue to grow at the rate that you want, GBP 4 million or GBP 5 million investment a year, taking into account the working capital that's required for that and the lower CapEx, when would you envisage yourselves becoming free cash flow positive again? And how much cash basically are you going to be required over the next few years to fund growth, as it were, at the rate that you want to grow at?

Any color on that, how we should think about it would be [ highly appreciated ].

From a principal perspective, I think I will take you back to this chart, where you can pretty quickly see that the relationship between what we spend and what we get back from it is -- well, you can see what the relationship is. It's kind of, you spend that amount of money and you get 70%, 80% coming through for several years afterwards net of attrition.

The working capital is largely funded by customers. The only working capital that isn't funded by customers is the capital for new customers.

So obviously existing customers are self-financing; new customers, we finance. So the answer is, if you spend GBP 1 a year every year over about 3 years, you become self-financing.

So then it just becomes a question of how quickly we can accelerate that.

Just with regards to the -- Wayne from Liberum. So if -- you mentioned that if you sell Majestic, you'll be paying a special of around 5p, the final dividend from last year.

But if you get your target of GBP 100 million, clearly if, can you just explain...

Just to clarify that we've never stated a target.

No. Sure.

But if that kind of sale number is in that ballpark or somewhere of that magnitude, capital allocation, how much cash you're going to be obviously be sitting on the balance sheet but then you've already just committed to that kind of a special. Just your thoughts around what the optimal balance sheet then should be looking like.

I'd point you to our capital allocation policy which is, we don't intend to hold on to cash we can't deploy. And if you combine the cash requirement, you guys can probably figure out from this kind of chart.

We've got GBP 15 million of debt on the balance sheet. There, you can pretty quickly figure out how much change will be leftover depending on the selling price.

We don't intend to sit with a pile of cash on the balance sheet. Our policy is we want to return surplus cash to shareholders.

We don't intend to do any acquisitions or major diversifications. We'll be definitive if and when the deal happens.

But until then, there's some pretty good pointers on that.

[ So it could then be ] at digital spend. That's -- [ only been baking off ] GBP 4 million to GBP 5 million of the GBP 26 million.

So can you give us some stats and KPIs of actually how digital acquisition spend is faring versus vouchers and your traditional channels that you are spending in? And why digital probably isn't a bigger percentage?

And just give us a favor of the mix between the channels.

[ Basically ] around all channels [ in ] Existing on exactly the same basis, which is, can we get 4x out of this? And if we can't, we aim to reprice it to get 4x; and if we get more than 4, we aim to deploy more capital until we start getting the marginal returns dipping down.

Digital, no different. So we're not going to disclose exact differences between how one channel's performing against another.

We apply exactly the same metrics to every single channel. Sorry, the second half of that question was?

It's -- you answered it. But I'm just trying to get a sense of why digital for example -- because obviously digital is the untapped opportunity versus you've been using the other channels.

And just to get some sort of sense of success versus failure and what your learnings are and what we could expect in the future.

Right. So I think that your -- the way you first asked that question was, why isn't it a bigger proportion of the spend?

And the answer is because we have to prove out things. It's not a case you turn on digital.

There are lots of opportunities within digital. And our constraint historically has just been, number one, our ability to test and prove all of these out, with the big constraint being effectively access to IT resources to, for example, build a specialist funnel for different channels.

So if you're on Facebook and you see Naked Wines ads on Facebook, you'll see there's several different approaches we use to try and recruit customers. And where you can design a funnel specifically to that approach, you get much more efficiency than if you put everyone through the same funnel.

That's an area we've been constrained over the last 4 years, it's an area we're not going to be constrained in the future. So digital, we expect to grow partly because we can see just statistically looking at other companies, our proportion of digital spend is relatively low.

So from a top-down view, it looks like there's opportunity there. The key to opening that up is having the ability to test multiple options and prove them out.

And we're going to have the resource to do that.

[ Chris Redmond, Paragon and Black ]. Your customer churn rate of about 20% a year, looks to me like that translates to about a 2x payback, not a 4x payback.

And following on from that, your -- it looks as if you're sort of having to buy back your customers every 5 years. So how can you improve your customer retention?

First of all, that's not correct. So for example, if you take 20% churn rate, that's about a 5-year customer life because it means 20% attrition, 1/20 is 5.

If we converted, let's say, 78% of that spend into contribution, then that's about a 4x payback. And if you download our presentation from the last Capital Markets Day, we explain that at some length.

So our payback is 4x. We are not having to -- you can see the pattern of -- from year-to-year.

You can see the length of time that cohorts mature at and how long they pay back for. The split into how much of this is replenishment and how much of this is -- how much of our spend is growth is simply done by taking the GBP 40 million and applying the attrition and working out how much of the new spend we need to replace that.

You can see that's about GBP 11 million, which leaves about GBP 15 million for growth. The majority of our spend is going into growth.

To the point of increasing customer retention, because having to effectively replace your customer base every 5 years is quite an expensive thing to have to do. What can you do to retain customers better than 20% a year?

So first of all, if we weren't increasing investment, retention would grow over time anyway. So if you look at customers that are kind of 2 or 3 years old, retention's closer to 90% where the sort of life span starts looking more like 10 years, right?

If you look at the U.K., which is our most mature market, it's more like 7 years. So the one thing we can do is just wait and customers get better over time.

The bad customers leave, the good customers stay and they get more valuable over time. As far as improving retention is concerned, if you go back to the chart that James showed how the cohorts this year are trading better than the cohorts last year, a chunk of that is simply down to retention.

We are just over time getting better at targeting customers who are likely to stay. The key to that is having the data skills to be able to identify the right kind of people to be targeting.

And again, that's one of the areas where we are being able to invest more resource in the future than we've been able to in the past.

Eleonora Dani, Stifel. Let's talk about Naked Wines in the U.S.

Can we talk about the demographic breakdown of the customers there? And what's the plan to engage with the millennials, with younger demographics?

I don't think we have said there is a plan to engage with millennials, to be honest. Just being really to the point about it, we -- the things that define our customers are much more they're likely to be early adopters.

They're the kind of people, they value authenticity. They're far more cultural things than they are age and income things.

And so it is true that customers who are 50 spend more on wine than customers who are 40. But if you look at age distribution, it's actually much flatter than it is for most wine companies who spike very heavily some time after 50.

We've got a significantly younger base anyway, most probably just -- probably just a function of the fact we're online and 100% online and most people aren't. We have no policy of particularly going out and targeting millennials.

We have a very strong policy of targeting customers who love the fact they're getting something special by helping a winemaker and therefore they're going to stick around. And I think every time we've focused our energy on finding more customers like the customers we've already got, we get good payback.

Every time we've tried to define our targeting in terms of socio-demographics, the results are pretty mixed. We haven't been able to make that work.

Best example of that is on Facebook. There are 2 ways you can recruit customers, one is you specify the demographic, the other is you say to them, "Here are a whole bunch of customers we've already got that you've got, too.

We want more people like this." The second way is dramatically more successful than the first.

When we've researched it, just as a matter of interest, the greatest correlation has been between attitudes rather than age.

Just on Majestic wine, the central costs. Clearly, what's going to be forecasting to come out is GBP 1 million to GBP 1.5 million.

Just explain how that breaks out because I would have thought that the central cost for Majestic was a larger number. So just how that number comes about.

So if we end up doing the deal to sell Majestic, part of the deal is likely to be that we're going to be providing a transition period for IT, and that's where the GBP 1.5 million comes from. I think the number you're really looking for is the -- if -- in the absence of anything else, if we sold Majestic, the GBP 9.4 million that we spent on central cost last year, what happens to that?

The answer is that falls to about GBP 6 million. And it doesn't fall to exactly half because there are just a bunch of costs of being a plc that aren't divisible by 2.

So I think when you look at the components of it, the biggest single decision we're making, for example, is we've had an IT team who have spent 70% of their time on Majestic over the last 4 years. We want 100% of that time to be on Naked now.

And we're pretty confident we can get a payback for that. Kate?

Kate Calvert from Investec. What further infrastructure investment do you need in the U.S.

to support growth over the next 2 to 3 years? And if you were to sell Majestic and have more money to accelerate it, how far in advance do you have to sort of engage your winemakers so you've got the supply in place to make sure you've got enough supply?

I'll let Nick answer the question about infrastructure in a second. But I think on your -- the first part of your question.

On the second part of your question, just to repeat it. Our rate of investment in Naked is not cash-constrained at the moment.

So selling Majestic isn't going to mean we're going to go faster or slower or anything else. It is determined by the rate at which we can find and prove new opportunities.

So don't expect us to be turning up in 6 months' time going "Oh, GBP 26 million has gone to GBP 100 million because we've got the money to spend." That's not going to happen.

It's likely to be much more -- the future is much more likely to be like the past, where we've grown at about around 15% organically. Nick, do you want to answer the first part?

Yes, happy to. So in terms of infrastructure, I think you kind of will have seen from the chart we showed around cost investment in the U.S.

business, we've actually put a lot of the investment in that infrastructure we needed kind of ahead the expected kind of further growth in the U.S. business.

So that's had a couple of components. And partly, that's -- and we signaled kind of clearly around investing in the controls and the security to make sure that we're able to operate a business at scale in the U.S.

So we've invested a lot in areas like compliance and building out the kind of finance infrastructure of that U.S. business.

We've also invested in resource around strategic partnership growth. Just one of the things we see as a future growth opportunity for the U.S.

business. So actually to deliver the plan over the coming years, we don't feel like in terms of kind of people, we need a big kind of big step up in terms of infrastructure.

If you're thinking more from a kind of capital side of things, again, actually the way the business operates, we don't own kind of large capital assets, and there's not a big capital investment required to support our growth. We talked a little bit about the warehousing infrastructure we've put in place.

And again, we've got a scalable network with a best-in-class service provision. There's not a big investment required to service our growth there.

Okay. Thank you very much, everyone.