Upcoming IP Investor Dialogues: Victoria PLC, Watsco, AWS & Cloud Ecosystem, Spirax Sarco, and POOL - Register your interest at help@inpractise.com 

Anonymous

UK Online and Offline Grocery Models

Former Head of Sales Organisation, Lidl Ireland and current Managing Director, McKee's Foods

Lidl in Ireland: from 0 to 12% market share in 20 years

Founder of Strategic Food Retail and Partner at The Partnering Group

Ocado & Online Grocery Economics

Former Chief Commercial Officer at Ocado Solutions, Ocado

Ocado Solutions Platform: Powering Global Online Grocery

Disclaimer: This interview is for informational purposes only and should not be relied upon as a basis for investment decisions. In Practise is an independent publisher and all opinions expressed by guests are solely their own opinions and do not reflect the opinion of In Practise.

{audio:0:00:00} Jason, could you share a short introduction to your background, please?

{audio:0:00:04} Up until very recently, I was working at Sainsbury’s, having spent about 30 years there. I started as retailer, went into store and property development and then probably the most significant part of my career was 20 years in ecommerce. I joined the founding team in the late 90s and stayed with them through the opening up of a whole bunch of stores and changing our entire delivery proposition, so we could give customers one-hour deliveries. We built a dark store and then just prior to me moving away from ecommerce, I was running Chop Chop, our on-demand grocery business which, in 2016, was a bit of a novelty. Today it is a phenomenon that is gaining much traction in the press.

{audio:0:01:08} Can we take a step back to the late 90s when you were part of Sainsbury’s founding ecommerce team. What was the strategic thinking about approaching online?

{audio:0:01:19} Back then, the strategic thinking was very much about customer acquisition. It was pre the first dotcom boom/bust period and people realized that online was going to be big but everyone also had wild forecasts as to how big it was going to be. There was a bit of a threat about, actually, we just need to get on and do this pretty quickly and acquire customers.

{audio:0:01:52} In the first few years, it was all about marketing spend, investing in acquiring customers and working out how to retain them. No one wanted to lose loyal customers who were coming into your store on a week-to-week basis, to your competitor, who could offer the same service for less money or had better slot availability.

{audio:0:02:17} Once that mechanism of acquiring and retaining customers had got itself established, it changed. It then became about how you incentivized customers to spend with you. It was more emphasis on how you could make your webstore so much better. How do you change your delivery slot pricing so that you can keep hold of customers? It became tactical rather than strategic.

{audio:0:02:48} How did the thinking around the fulfilment model change?

{audio:0:02:55} Very early on, we had some consultants do some work on our network strategy and they had forecast that Sainsbury’s would have needed nine dark stores or fulfilment centers that ran up the spine of the country. We opened one and it didn’t take off as we expected, understandably so. This was pre-broadband; it took you 10 minutes, with a dial-up connection, to get a web page working.

{audio:0:03:33} We rapidly moved from a centralized fulfilment strategy to realizing it was probably going to be quicker and easier for us to utilize our store estates to create much smaller ecommerce fulfilment operations out of the back of stores.

{audio:0:03:51} How do you compare that original thinking to today, where there obviously is broadband and a lot of people online?

{audio:0:04:06} I think we’ve come absolutely full circle very quickly. That is all to the power of technology. Sainsbury’s developed a really good store operating model and scaled that and got to capacity. Meanwhile, Tesco, who had – and still have today – the dominant market share, quickly moved from a store-based model to a dark store model and built several sites around London. Sainsbury’s were a little bit late to the party but the demand wasn’t there. They shut their first dark store, because it was underutilized. The development program of their second dark store started in 2013 and it was opened in 2016.

{audio:0:05:13} For the full circle part, from the power of technology, we have been able to utilize far better technologies in stores now which have been able, in some cases, to double the amount of capacity that a store can deliver to customers, with great picking efficiency. In the UK, there is world class picking efficiency. I am looking at it in Singapore and over in the US, right now, and they are not even anywhere close to what Tesco, Sainsbury’s and Asda are doing. With that level of efficiency, there hasn’t been the need to maintain dedicated fulfilment centers.

{audio:0:06:03} I saw a particular Sainsbury’s, only last week, that delivered its last order out of its dark store, as it has moved all of its volume back into stores. I think that is the right future model for fulfilment because the other thing that technology is doing is shortening lead times for customers. Customers don’t need to make such advanced planning decisions about what they are going to eat in their household this week; they can do a lot more, little and often. This plays back into the on-demand hands. I think the future of fulfilment is going to be more very much hyper local, rather than a centralized model that has got a long lead time from order placement to shipment.

{audio:0:06:03} The centralized method, like Ocado, arguably has better selection, better availability and better customer service, potentially. Are you valuing the convenience aspect so much that you think fulfilment should be hyper local?

{audio:0:07:11} Yes, I think so. If you think differently about online, it doesn’t necessarily have to be a complete replication of a supermarket. It is becoming a channel of its own nature. You can start to curate a range which is more finely tuned to what your customers really want. You don’t have to carry 30,000 SKUs because you are trying to cater for so many different preferences and choices and you’re trying to encourage customers to buy more. A broader range creates impulse. Online, that impulse disappears. On your checkout page, you might get a pop up to say you may have forgotten a few things, but the customer journey is less impulsive.Therefore, 30,000 SKUs is not necessarily that important anymore.

{audio:0:08:21} Ocado, as a standalone model, works really well when you want to put a brand down in a market where there isn’t already a physical presence, or you are a retailer that is last to market. I think they do an incredible job for Marks & Spencer and Morrisons, because they are the latecomers to the ecomm party and have needed to scale operations and I think Ocado is a brilliant solution to help those retailers get up and running.

{audio:0:09:00} Why do you think the big players in the market – Sainsbury’s or Tesco – don’t use Ocado?

{audio:0:09:06} Because we have a store estate. Sainsbury’s have got 600 supermarkets, up and down the country; 200 of those are running ecomm out of them, so you’ve still got headroom for another 400 supermarkets to grow their own ecomm business.

{audio:0:09:28} I think one of the other things that will be interesting to watch in the future will be any kind of future legislation around 3.5 ton vans driving around the country. They start out full and then they are driving back to base empty. The carbon footprint of an online grocer, in a centralized model, probably isn’t as green as utilizing your existing store estate to ship orders direct to customers. The last mile genuinely is a last mile; it’s not a last 20 miles.

{audio:0:10:04} There was an interesting statistic in Sainsbury’s’ and Tesco’s recent earnings, where they said that around 30% of the online sales, during Covid, were actually cannibalizing the store sales, which is obviously critical in deleveraging the store estate for the grocers.

{audio:0:10:25} Totally. That’s understandable as well, because during that period of time, people were queuing to get into a supermarket. I think the big question now is, how much of those behaviors that consumers have adjusted to stick in the long run. Because the pandemic has affected our lives for well over six months, those habits are pretty well formed. I used to be an online shopper and now I’m not. I top up, little and often, on an every other day basis because l plan less.

{audio:0:11:14} Just on that point, you think that the customer behavior is going to shift materially from the traditional weekly, £120 basket sizes, to these often and little purchases?

{audio:0:11:29} I think so. Probably not for all life stages, but if I think about my own family’s behavior changes, my 80-year-old parents, pre-pandemic, were the £120 a week, go into a shop and fill up a trolley, because they’d got their list. During the pandemic, they didn’t want to go out; they shielded. They’ve become online converts and they shop every other week, with either Sainsbury’s or Waitrose. My daughters, who are the other end of the generational spectrum, don’t go to a supermarket at all. They are either shopping from my local on-demand grocer, or they are deal hunting on one of the aggregators. Whoever can deliver them a meal cheapest, that is where they are going. They don’t really see supermarkets as the place to go, to fill up a trolley with groceries and come home again.

{audio:0:12:37} Over time, as those generations become household owners, they probably won’t ever go to a supermarket and fill up a trolley.

{audio:0:12:51} Don’t you think those households, such as your parents, who were purchasing £120, full trolley in store, are they not just replicating that online and doing a big shop once a week or fortnightly?

{audio:0:13:08} They are telling me that they are spending less now than they did before and I think that goes back to that impulse point. When you are properly meal planning and placing an order online, you are also seeing the shopping bill tally up alongside. In my experience, what we were seeing in Sainsbury’s was that customers were pretty much spending their budget, whatever that budget was. Whether that was £120 or £60, there wasn’t a great deal of variance, week to week. They were clearly managing to a budget and filling their basket accordingly.

{audio:0:13:55} It’s an interesting point because you forget the subtle differences. Just the price being there and seeing it tot up can stop you adding extra stuff to the basket.

{audio:0:14:07} I think Waitrose have probably got a strong lead here. The consumer’s participation of their smart shop devices is probably having the same issue. They are able to tot up their basket as they are going along and, pre-conversion to smart shop, they were probably spending more than now, when they are using a device in store. It will certainly be in my lifetime when, in supermarkets, everyone will be using a device to check themselves out. It will seem very alien to put a load of stuff into a basket, to unload it onto a belt, to then put it into some bags and walk away.

{audio:0:15:00} Going back to the point on Ocado, why could the situation be different for Kroger, for example?

{audio:0:15:11} There are a couple of points there. There isn’t the same maturity of the ecomm marketplace in the US as there is in the UK. The hierarchy and structure of supermarketing across the US is very different from the UK. It is dominated by only a couple of the nationals, but there is a proliferation of the regionals and the smaller retailers, in the States, that take a good share of the total food market.

{audio:0:15:54} What is really smart about the Kroger Ocado move is that Kroger would have had a choice; put down a supermarket site in Florida or put down an automated facility that enables them to distribute their groceries to a far larger catchment area. Because they don’t have a physical presence in Florida, they’ve gone for the option of building a fulfilment center, to distribute from there.

{audio:0:16:30} That makes sense because they haven’t got a presence, as you said, but that is only one of the 20 odd stores that they actually committed to. What about all the others, where they have a presence? Do you think it makes sense to roll out a CFC, if a grocer has 20 or 30 stores around an area?

{audio:0:16:47} I would question the strategy if you’ve got that level of supermarket coverage as to why you would build a dedicated fulfilment center. In my experience of working with a couple of US retailers, they are light years away from UK levels of efficiency. UK grocers may claim they are delivering small levels of profit on their ecomm business, but it’s wafer thin. The US retailers are operating at two-thirds less efficiency than a UK retailer and I think that has driven them to think more about micro fulfilment centers or CFCs, to get up that efficiency curve.

{audio:0:17:51} We’ve been lucky, in a way, in the UK, that we’ve had 20 years of ecomm development which has enabled us to really fine tune our processes, our operating models and our technology, to get to world class pick speeds. That is having a huge impact on the likelihood of getting some sort of profitability on ecomm.

{audio:0:18:19} On the differences between the US and the UK, do you think we are going to see the US transition to a similar model to the UK, or do you think it’s going to be fundamentally different?

{audio:0:18:32} Fundamentally different. Something like 70% of orders that are picked in stores, are actually collected by customers. The culture in the US, of driving to places to get stuff, is deeply ingrained in their society. Whether it’s more by accident than design, their ecomm businesses will deliver profit because of curbside collection. They don’t have that additional cost – with the exception of Kroger and their CFCs – of delivery cost to consumer. I don’t ever see delivery to consumer being as big as it is in the UK and across Europe. I think they are a creature of habit; they will jump in their cars or trucks, drive to a Walmart and have their groceries put in the boot for them.

{audio:0:19:53} It’s also just a bigger country and there is space to drive and park. It makes it a lot easier than here.

{audio:0:20:02} Yes. As I say, it’s perhaps more by accident than design but I think it’s a strong move to not push the delivery. I think it would work well in the major cities but that is just the East and the West Coast. For everywhere else in between, the culture will be, I can order my groceries online so that I don’t have to walk around a store – the stores are enormous, the trolleys are enormous, everything is massive – let someone else do that hard work for me and go and get it myself.

{audio:0:20:41} You still need to pick it, so do you think they are going to pick in the store or in micro fulfilment at the back of the store?

{audio:0:20:47} Because the technology is now ready, I think micro fulfilment will be the UK’s equivalent of in store picking. Once they get 500 orders a day, I think they will move, pretty quickly, into micro fulfilment and they will automate the processes. Then I think I can see that technology being proven in the States and then moving over into the UK as we start to think about downsizing some of our enormous supermarkets and repurposing that space to deliver a higher rate of efficiency.

{audio:0:21:29} Can we walk through the in store picking process, historically, in the UK?

{audio:0:21:47} Back in the day, it would have been a single order pick bill. Your order would have been received and, right at the beginning, a picker would have been given a pick list on paper and they would have assembled your order and it would have been dispatched on a van to you.

{audio:0:22:10} These days, it’s far more efficient. Your order will be one of 500 orders that are going to be leaving that store that day. All of those items – say 20,000 – end up in the pick system’s algorithm and the pick system is going to work out the least amount of effort to assemble those 20,000 items that are required to be picked that day. Your order could be divided up into six component parts, picked by six different pickers across the sales floor. Once those picks are complete, those boxes arrive at a staging area, in the back of the store. Systemically, the jigsaw puzzle is put back together and it is loaded onto a van. The van driver turns up and doesn’t really need to do anything other than validate that they have got the right number of tote boxes on the delivery vehicle, and off they go and start doing the delivery.

{audio:0:23:16} The technology has replaced all of the decision making and thinking that is required in that process.

{audio:0:23:27} Is that a third-party technology, like Dematic or Takeoff, that assemble that part at the end or is it proprietary?

{audio:0:23:35} In most cases, it is third-party technology but not the Dematics or the Takeoffs because they are automation and technology providers. In the UK, it is Systems Technology that is being used. The core of the product is a warehouse management system; an old-fashioned logistics warehouse management system, that has been adapted to cope with the demands of ecomm grocery.

{audio:0:24:13} Do you need to carve out a section at the back of the store to do this in store picking?

{audio:0:24:19} Yes; typically, 1,000 square feet of space.

{audio:0:24:29} Yes; for an MFC, you are typically talking about 10,000 square feet of space, but you are holding inventory in that space. The staging area in a supermarket is just for holding on to completed orders, for a short period of time. It’s like a departure gate at an airport; for a moment in time, it’s full of people and then when the plane leaves, it’s empty until the next plane needs to leave and then it builds up with people. The space turns over fairly frequently, throughout the day.

{audio:0:25:06} How do you link the inventory and availability to the existing availability online?

{audio:0:25:25} They are real live challenges today. To your point about Ocado, the great thing about them is that it is 100% known inventory. As soon as something is out of stock or expected to be out of stock, it can be greyed out on the webstore. But a supermarket can’t do that. They haven’t quite mastered the AI ability to be able to forecast when something is about to go out of stock, or low stock, and take it off the website.

{audio:0:26:03} The way that retailers have been able to mitigate for that is to, in effect, offer 100% availability but then shift when the picking is done, from during the day to before the store opens. That’s one of the other advantages/disadvantages over the UK model is that customers today, place their orders today, for tomorrow. There is a small percentage of them who have got the ability to be able to order this morning, for a delivery this afternoon. 90% of the volume is going to be known today, for picking tomorrow. That enables a store to get themselves really ready to be able to pick that volume.

{audio:0:26:55} So availability is still compromised, but it’s not as compromised as it used to be, because you don’t have an offline customer fighting for the last banana with an online customer, because the online customer has already had their order picked before the store opened.

{audio:0:27:14} That could only be an issue if we see a huge transition to customers wanting same-day delivery?

{audio:0:27:21} Yes, and that’s what we see in the US. Their proposition is very different. Again, they are bang on the money for the future of ecomm because they tend to have more of a four-hour lead time. It is not a day one for day two proposition but it’s, if you place your order now, in four hours we will get it to you. I think, generally, consumer trends will move towards this more immediate fulfilment, rather than planned purchase.

{audio:0:27:58} They will struggle from an availability point of view, but the customers would understand, if I’m placing an order at lunchtime, for a delivery at 4PM, my availability isn’t going to be as good as it could be if I ordered it first thing in the morning, for a lunchtime delivery. It’s small percentage terms at the moment. It will only really become a problem for them when 20% of their business is shipped to customers rather than the 5% or 6% that it is today.

{audio:0:28:31} How do you see that evolving? Do you think we’re going to get to that stage where the UK grocers will have to start adopting MFCs?

{audio:0:28:42} Yes, definitely. The holy grail at the moment, for UK retailers, is all about efficiency; getting to profitability by being efficient. You are going to have to compromise some of that efficiency when you move towards a more in-day fulfilment model. The way to square that circle is to think about what steps of the process you can truly automate, with humans to begin with, then robotics later. That’s not far away; I would say we’re two or three years away from robotic picking arms, in MFCs, that will start to assemble orders. Then you can offer a proposition which is more attractive to the customer, without having to take the cost of picking inefficiencies as a consequence.

{audio:0:29:37} The long-term model here is, if we assume that convenience wins in the long run, people typically want their orders quicker, as opposed to waiting longer, the UK grocers will have to move to an automated MFC model, rather than purely in store?

{audio:0:29:53} Agreed. While we are at 20% today, in the Far East, they are running their ecomm businesses at around 26% of sales. The balance of space in stores is going to have to change. Range rationalization will create space in stores. Proposition changes in stores will create space. The best way to repurpose that space is to create automated fulfilment for your ecomm business. It will be at that point that you can see the model changing from batch ordering day one for day two, to a more fluid, on-demand way of picking and dispatching orders.

{audio:0:30:51} Why haven’t the likes of Sainsbury’s adopted this already?

{audio:0:31:02} It’s a good question. I don’t know whether we just haven’t really thought this through yet. There is a strategic angle here around how big could ecomm be, how much are we prepared to compromise our P&L to run an ecomm business and who wants to go first on flipping the model and saying, you know what, day one for day two was yesterday’s ecomm business; the future is all about in-day, it’s all about on-demand, it’s all about speed of fulfilment. To be able to make that happen requires two investments – one in space and one in technology – to be able to get that infrastructure in place to be able to satisfy on-demand.

{audio:0:31:59} I think Tesco are there. They’ve got their first micro fulfilment center and plans for more. I look at Tesco and think, they are a year or two ahead of most. The smart money is on watching what they do with micro fulfilment and being a fast follower.

{audio:0:32:20} Sainsbury’s and the like, they could be right? If online gets to 30% or 40%, which is still a huge percentage but, actually, 30% total of sales are happy to be next day – maybe I don’t need the food immediately – then maybe it can actually be only a few MFCs or a few dark stores, instead of the majority?

{audio:0:32:48} True. I often think about this and there are other dynamics that change that retailers tend not to consider that often. As a retailer of old, I know what it’s like; you are focusing on what is going to happen tomorrow, the next week, this quarter. Everything else around it is the status quo. But of course, it’s not the status quo. There are generational shifts; the population and socio-demographics are changing all the time and I think we’re just on the cusp of a big change.

{audio:0:33:40} Back when Piggly Wiggly, in the US, opened the first self-serve supermarket, they were probably considered to be the radicals, doing something really insane. Suddenly, it was a generational shift that adopted the self-serve supermarket as a concept. I think we’re on the precipice of that next generational shift where ecomm could go exponential in food and, at the same time as going exponential, lead times will shorten. If I can order something on Amazon Prime this morning and I can get it tonight, why can’t I get my groceries from two miles down the road? Why can’t I order this morning and get it tonight? I think the whole ecomm time shift is going to hit grocery retailers, at some point.

{audio:0:34:38} Obviously, all UPHs are calculated differently, but I’ve read that Ocado has 180 UPH, Oda is at 220. What can you actually get in store, consistently, on a fully allocated based or apples to apples with Ocado?

{audio:0:35:02} There are two types of measure of UPH. There is what retailers tend to use which is what we would call ‘on the gun’. As soon as I have logged onto a device and I’ve started my pick walk and when I’ve logged off after I’ve finished by pick shift, that is my on the gun units per hour. That’s a really credible number because it is comparing apples with apples, very closely, but it also helps the retailer to plan their labor very well. Best in class in the UK is, I would say, Tesco and Sainsbury’s, neck and neck, probably knocking out 170 units per hour.

{audio:0:35:53} The other measure is a fully loaded units per hour. That takes into account all of the labor required to fulfil an order; the staging at the beginning, the staging at the end and getting it onto a van. What you could typically do is take 30% off that 170, to get to a fully loaded rate. None of the retailers use fully loaded as a measure because it doesn’t help them with building their labor standards. They are looking at individual tasks and building labor up.

{audio:0:36:29} I would say that the industry measure is on the gun, units per hour, and the best in class are running at 170. When you think about an Ocado or an MFC, those guys, on an equivalent unit per hour pick rate, should be hitting north of 400 units per hour because they are not moving anywhere. They are at a machine, the product comes to them, the customer tote box comes to them, they are picking a product out and moving it over.

{audio:0:37:02} Whilst you can’t compare the two operating models, what you can do is say, look, if you are in store, you should be hitting 170 and if you are in a micro fulfilment center, with goods to person picking, you should be hitting 400 plus.

{audio:0:37:20} That’s where it becomes tricky to actually understand how profitable it is. Again, Tesco and Sainsbury’s say that they don’t look at it like this, in terms of online as a specific standalone business because, as you said, they are using the fixed assets of the store. How should we even look at the profitability of online if they are using that in store model?

{audio:0:37:50} We shouldn’t; that’s the simple answer to that question. You have to look at the business as a whole. You can’t differentiate an online channel from the checkout operation or the cost of a bakery. If you are an omnichannel retailer, you have to look at it in the round and say, as a business, am I delivering a profit? Yes, I am. Am I satisfying my customers? Yes, I am. Am I growing my customer base? Yes, I am. If you are doing all of those things, then you are a healthy business.

{audio:0:38:30} If you are not profitable and you are losing customers and your customers aren’t satisfied, then it doesn’t matter how good your ecomm operation is and whether it’s profitable or not. The fact is, in the round, you’re not a profitable business.

{audio:0:38:46} Perhaps therein lies the challenge for Ocado. You’re not an omnichannel business; you are a single-channel business, with a lot of fixed cost and a lot of assets that you can’t completely deleverage and you might not be delivering – on a supermarket’s playing field – profit.

{audio:0:39:14} As you said, this is an interim solution. If we believe in the consumer behavior change to more immediacy, shorter lead times and the batching and availability issues, then there is going to have to be a transition to MFCs or dark stores eventually, for the grocers.

{audio:0:39:35} For the grocers, it is a transition to automation. I’ve got no doubt in my mind that that is the end state. What I ponder is, what is the end state for a pure play grocery business, whether that is Amazon Fresh or Ocado. We’re already seeing the indications of what they are going to become, not what we consider them to be today. Amazon make more money from their web services business unit than they do by shipping billions of parcels around the world and I think Ocado are doing the same thing by becoming a technology provider, a solutions provider, as opposed to being a grocer. Maybe in a decade’s time, there won’t be CFCs, but there will be Ocado-run MFCs in stores, because they have invested in developing the technology that probably most grocers are desperate to get their hands on; it’s just that they didn’t want to be a part of the big shed in Hatfield.

{audio:0:40:50} But that changes the landscape they compete in because you have another 10 providers that do similar things for MFCs, such as Dematic and Takeoff?

{audio:0:41:04} Yes, but they’re not all equal. My top three or four for advanced fulfilment technology providers are Ocado, Alert Innovation, Fabric and probably Attabotics. They’ve really thought about not just automation in the traditional logistics sense, but they’ve gone one step beyond in their thinking about assembling multiple order lines into a small as contain as possible, that is ready made for shipping or for curbside collect.

{audio:0:42:02} I think Tesco have six or seven dark stores today. Could you explain exactly how that has evolved in the UK?

{audio:0:42:13} I think you’re right. It’s probably best to use that Tesco’s example. The first one that they opened is still open today; it’s in Croydon and it’s 100% manual pick. It is a supermarket that just hasn’t opened its doors to customers. The next generation on from that was still picking as if you were walking up and down the aisles in a supermarket but at the end of that pick, those picked orders were fed into an automation solution that sorted them all out and brought them all together. It was like when you are at an airport and your bag ends up getting to the right gate at the right time; that is the technology that the second generation of dark stores used. It was an automated facility to consume loads of component parts of picked orders and then bring it together and dispatch it to a loading for a van to be shipped at the right time.

{audio:0:43:26} The generation on from that started to introduce automated picking, where step one was a tote box that was moving around a conveyor belt system and colleagues were stood in an aisle, picking items off the shelf and putting it into a box. The other more commonly seen technology was both the product comes to the picker and the customer’s tote box comes to the picker and the human just moves product from box A to box B. That is the current, most advanced version, of automated picking in a dark store.

{audio:0:44:06} The dark store would just serve orders from a 10 or 20 mile radius of the store?

{audio:0:44:12} Yes. It all depends on order drop density in an area; how many customers, within your catchment area, are shopping with you? You would rarely see a dark store out of a city center or out of the peripheral of a city center because these things need density of volume. Typically, 25,000 to 30,000 orders a week, within the tightest area possible. If you can do that in five miles, brilliant. If you’ve got to cover 20 miles, that’s okay too. But if you’ve got to 50, 100 miles then you are just eroding all of that efficiency you had in a centralized fulfilment model; you are losing all of those gains by traveling so far to dispatch.

{audio:0:45:04} Because the last mile is the largest part of the cost?

{audio:0:45:08} Yes, because you are down to a human and a vehicle, the two most expensive assets that you will have in your operation. Whilst they are driving and delivering to customers, great. But there is no backhaul. Unlike a traditional logistics model where you get the opportunity to bring something back to your depot or warehouse, whether that is product from a supplier or empty containers from a store, you don’t get any of that in online. With the exception of the empty tote boxes, they are coming back empty. That is all sunk cost.

{audio:0:45:52} Why did Tesco even choose to roll out seven dark stores, rather than just use their existing stores? I guess that does still deleverage the stores, as well?

{audio:0:46:02} I think this is question about the strength of Tesco’s store estates, in and around London. One of the advantages that Sainsbury’s have is that they have a very good sized supermarket network in that area. When Tesco were expanding, they just didn’t have that footprint of real estate. If they did, I would be surprised if they had those dark stores open today.

{audio:0:46:41} But they are at a scale now where they haven’t got supermarket estate, therefore they don’t have the capacity and if each of those fulfilment centers is knocking out 20,000 orders a week, that is 140,000 worth of order capacity that they have got to find in the supermarkets and that’s a really tough ask. I think their CFCs are there for the long run.

{audio:0:47:06} How is the store estate of Sainsbury’s different from Tesco, in and around London?

{audio:0:47:11} There are just more of them; and they are bigger. In the days when Sainsbury’s was number one and growing its footprint, its heritage is in London. Their customers were in London. I think London has been their heartland. I think they were very lucky in the late 70s, early 80s, to have put down quite a bit of real estate in London.

{audio:0:47:46} And they are bigger stores?

{audio:0:47:49} Yes. Even some of our compromised stores in London are operating good ecomm businesses now.

{audio:0:47:59} I’m just thinking about the US and, given the size of their stores, it probably makes complete sense that, if you see any online penetration, you can just carve out 5,000 or 10,000 square feet at the back of the store, rather than build a new one?

{audio:0:48:11} Easily. The micro fulfilment centers that I have visited in the US are operating a 10,000 square foot MFC that could easily be a 20,000 square foot MFC and you wouldn’t even notice the difference on the sales floor. These supermarkets are huge and their real estate costs are a fraction of what they are in the UK. The ROI on MFCs looks so much more healthy in the US because their cost base is so much lower.

{audio:0:48:48} What do you think is the major challenge with rolling out MFCs? Let’s say we are looking back in 10 years’ time and MFCs are a fraction of a percentage of penetration, why could that be?

{audio:0:49:24} There is still going to be a place for MFCs with the big guys; the Walmarts, the Krogers, the Albertsons, the Publixs. I think when you get down into the regional grocers, such as the Giant Eagles and Wakeferns, they may start the MFC journey because they can see that, in their neighborhood, the big guys have got them. But equally, they may be able to prove to themselves, over time, that if they can just make what they currently do more efficient, they can either delay the need to invest in an MFC or just completely mitigate the need for an MFC.