How is a warehouse for a traditional retailer like Marks and Spencer structured?

In terms how it's actually structured, in either end of the process, you've got the customer at the front, and then, if you go full circle, you've got the people who want to choose what the customer is going to buy or guess what they're going to want to buy.

Everything else in between is a matter of administration, ordering, manufacturing. It's the supply chain. This is really the supply chain point, until it arrives into the nearest territory or the most logical place to actually have a warehouse. If you look at the traditional retailer such as M&S, it's a very supply-chain-led retailer. The warehouses in question, because of the density of stores, they will have multiple warehouses clustered around the country.

You do see this in big retailers anywhere in the world. If there's enough volume, they will have multiple warehouses to try and keep as much stock as possible, close to where it needs to be.

In terms of the actual ownership and structure of those warehouses, it only happens in one of three ways. You normally buy-in to a campus-type facility, usually offered by a 3PL or even a 4PL. The other one is you have a very specific retail model, and you choose to subcontract that entirely out to a 3PL who will build your warehouses for you, and operate them. Or you do it in-house.

There are different hybrids. You can have a bit of an overlap between the three of those as we know, but those are the main high-level types of operation that you can have there.

When the inventory comes in, obviously, Amazon came to pioneer the chaotic inventory storage system in terms of categories that are stored together. How are the traditional retailers storing their inventory?

Depending on how much inventory they've got. They normally stored it by product type. All right? Which sounds like the logical thing to do. You keep shoes with shoes and food with food.

And you do get into hygiene factors when you get into perishables where you can only store certain things in the same temperature bracket. And then there's products that you can't mix so you're really looking for exceptions. For instance, I can't put bleach next to fresh poultry because that could be bad for humans. So there's those sorts of criteria.

How do those metrics compare to the likes of Amazon?

Quite a lot actually. The thing is, well, Amazon have the same issue. All right? Regardless of whatever happens in your warehouse, there's always products that you can't put next to each other. In fact, I probably won't pick on Amazon, pick on Zappos, they're probably the best example of chaotic storage.

What they do is they have brands of trainers. Most shoe boxes are the same size, so rather than have a section of the warehouse for Nike, another one for Under Armour, Kickers, or Skechers, or something, they just basically take any inventory and put it into any available space. They have a smart system that knows specifically which product in which size is in which space. It's not grouped by brands, they're not even grouped by size. It is literally, you fill the space, and the products will know where it is.

When we say chaotic, it's not necessarily chaotic. It actually makes sense. You just put in the best piece of the jigsaw in the most likely available space.

Traditional retailers are usually legacy retailers. They were around when your store manager was a guy who had a back office or back storeroom, and he would know which products were where. It's all done passively within the brain. Of course, if these retailers were a lot bigger, they ended up building supply chains. They've done this long before the advent of really good supply chain technology.

Realistically, it's a legacy thing. It's just managing growth. It's a lot harder than managing contraction. If you were to start again, you'd do it very differently, right? It's just an expansion of how things were traditionally done.

What is the actual process from their hand picking to actually finally shipping the product?

Again, there's a number of methods to doing this. Typically, what will always happen is there is an order. There's a request. There is a requirement or something. There is a message that says, "I need to take X out and take it there." It's as simple as that. You then come to the pick stage, which is knowing where the stock is. It's retrieval of stock and making sure that you've fulfilled that. It's very much a checking things off as you go along process.

Once you've done that, you build a consignment. For instance, you're putting something in an envelope, like a lipstick that's going to a customer, or you've got a handrail to go to a store, et cetera, et cetera. Then you select your mode of transport, and then you pass that to whichever mode of transport you're taking, and then complete your delivery. There may be further additions in there. You may have some rework process that says that all of these need the prices removed or something else, or something else.

As a typical process, that's what happens. Message to order, pick to message, check  there's no discrepancies or take action for discrepancies, prepare for mode of transport, transport, complete.

For a typical delivery then, where does the majority of the cost lie in that? What's the split there in costs?

It really depends. Most of it can lie in the actual cost of moving the vehicle. Moving things across the floor is one thing, packaging, small labor in the warehouse is okay. Cost of the product is significant. The actual big thing is the cost of those trucks, to man those trucks, insure those trucks, pay taxes on the trucks, maintain the trucks, and fuel - very importantly - fuel the trucks. And then mitigate against real-world conditions. That's where most of the cost goes.

You have an estimated cost because you know what your fixed cost asset is, but you have a lot of variable costs in there around time, recovery, consequence. In the actual physical movement across the world or across the country, that's where many things can go wrong and you end up potentially spending more as a consequence.

In terms of the structure, if you take M&S for example, they have one large warehouse, which is obviously used to manage in their stock, and they will ship stock from there, and then around the country?

No. They have hubs and nodes. That warehouse is the only one in the country that actually held the entire inventory of M&S stock. Then you have several sub-warehouses, which would feed, so you'd have one which was specialized in, I don't know, menswear, and then it would have limited ranges of everything else.

That did two things: number one, it put quite a lot of stock - almost like a little cross-stock staging post - close to where you needed it. It also allowed you to keep a supply of your inventory at regional sites. If you were to lose that site, let's say the big site in the middle, you still have inventory for everything you need, albeit not in one place. It's scattered around the country.

Similarly, if you lost the stock on the smaller sites, you lost a site that held menswear and was crossed off all the local stores, yes, you've lost capacity, and yes, you've lost your resilience in menswear. But you've still got menswear in the middle, and you've maybe got some more around the country. In addition to that, you probably liberate some store stock. Yes, you've got capacity issue, but you are resilient.

That was typically how the networks were built. Having everything in one place is not normally advised, unlike, being able to liberate stock from smaller locations. Then this is where we get to omnichannel, because if you think everything is a warehouse and all stock is waste. It's a really important line, that.

Because you want to sell all your stock, because if you don't sell it, you pay to store it, you pay to manufacture it, and you've got to do something with it, and all that costs lots of money. All stock is potentially waste. It's about how you can get rid of that at the best possible price.

The second thing is that everything is a warehouse, even the trucks that you're using to move things to the planes or the boats or whatever. You know where that stock is at that precise moment in time. The idea is liberating that piece of stock from wherever it is in your entire supply chain, and giving it to the customer that needs it.

That could be as simple as taking something from a store and handing it to a customer when they walk in to using something from a truck that's on its way to a regional cross-stock from the center and saying, "We now need to pick that up, and then put it on another cross-stock vehicle that's going in the opposite direction," and then take that to a nearby parcel depot where a driver is going to take it someone's place of work the following morning. All stocks should be available to all people to take on.

As we shift to omnichannel, what is the split then between using third-party carriers and owning in the last mile for the big UK retailers? For the fashion retailers and M&S.

Very limited in terms of owning. No structural ownerships on vehicles and assets. Typically, very, very few retailers will buy the trucks and even less will own them. Some may purchase trucks, and actually pay a 3PL to operate it for them because that's not what they do. I would say that that's actually waning at the minute. Actual asset ownership is rapidly in decline. This is a market which likes to lease, rent and so on. Typically, most of it is outsourced. I haven't seen the industry specific numbers here, but I think if I heard that 5% of all retailers owned and operated their own trucks, I'd be surprised.

Can you just talk about the challenge? What is the return rate in the UK at different categories, online?

Good point. In terms of categories, it depends. In the UK, we're actually blessed with a moderately okay rate of return. Depending on the product, you could be talking as low as maybe 1% or 2%. You can expect that on high-high tech, like nice big TVs, your iPhones, gaming consoles, things like that, less than 1%. Fashion is much higher. Fashion can be as high as 25% to 30%. As a whole, it's about 10% to 15%.

This has been the issue, obviously, in apparel retailers. Let's just run through that process. Let's say I purchased a product and I want to return it. What are the costs for a retailer then to pick that up and get that back in the warehouse?

The costs are broken down, someone has to pay for it. Whether the customer pays directly for it or not, I think the ultimate thing is here, the customer is always paying for it somehow. Usually a provision for return is baked into the retail price of the outbound product. If it is considered to be free of charge.

Because in all those costs of doing business, the only point that you capture revenue is at point of sale. If you are offering a free return, you're hoping that they're going to follow the lowest cost route to get it back to you as quickly as possible. Because, don't forget, this is product that's out of your inventory and out of the supply chain. Therefore, it's not wanted but it's not immediately available to another customer.

Your cost would be, first of all, whichever returns route you choose. Be it a portion of your store to drop off returns, which you'd hope it'd be because then if the item is okay, you can do a cursory inspection and it can go back into stock immediately at that store. If not, then it will fall back into the supply chain, and travel through the loop, be inspected at a big return center, and then made re-available to the general stock pot.

If you're using an e-commerce return service, then maybe you're using a third party PUDO such as a CollectPlus, or a La Poste, post office, or an in-post locker, or something like that, or you'll offer a pickup service. The favorite is towards drop-off because they do tend to be a lot more affordable. You, as a consumer, will not pay for it, but your retailer will have paid for it. Maybe a few pounds, you will get a tracking number, and you will see that processed through.

The next staging cost comes from the inspection and reconditioning. Fashion items will need to be repacked. They have to personally be inspected, so it's a visual human inspection to make sure that it's not damaged. It will then potentially need to be steamed because don't forget these things crease when in transit, so they have steam tunnels. Seldom do you get tailoring involved. They just have to be made sure that they're ready to go back out.

Then it follows one or two routes. They're either repacked and sent back out to the market, or it's considered as written off. Maybe the customer is charged for it, there's a fall out from that, or it's write-down and it's then sent to either an outlet or a what we call a jobber.Someone will actually buy that stock seemingly cheaper then try and sell it again and make some money on it.

What percentage gets written off? Is it a low percentage?

I think an alarmingly high amount of product is written off. If you look at most websites, they always have an outlet or you do have these big shopping villages like Vista village, and McArthurGlen, and things like that. They're all over the world. I've seen them in Vegas, in Orlando, in Singapore even in Dubai.

Again, depending on the retailer, so I'm not going to say specifically to M&S or anyone, you could be talking as much as 20% of your total stock pot is sent to outlet which is wrote-down.

No, not just returns. It could be anything, unsold product.

Yes, because the whole move towards omnichannel makes all stock available to all customers. It does mean that you do end up with that whole stock that you can't sell, but may have been through the loop a couple of times. If you think in some extreme cases, you're getting free returns now in the UK from companies that sends stuff to Australia.

If you run through that, if I buy an item and return it, the free legs of the delivery cost that's baked into finally shipping that product. If you do ship that to someone else, who's bought the product. If I buy this online, there's the first class delivery to the actual customer, and then if I drop this off at some kind of, locker or midway hub, that then gets collected again. Especially in the apparel space, I think this is pretty common.

Then a third-party delivery operator has to pick it from the depot and ship it back to the warehouse. If you ship it out again to someone new then that's free legs of delivery cost that the retail has to pay.

Is that getting baked into the final price or...

Sure, it is.This is the whole misconception. Everyone thinks things are free, and they're not. Far from it. In fact, when you look on the news, and you read about serial returners, I ordered 10 dresses and I returned nine, think about the extreme cost there.

A, you've actually got the cost of taking it out, and taking it back. B, you've got the cost of taking that product out of the supply chain. Then you have the cost of re-inspecting every item, the cost of repacking every item, and the cost of reselling every item. That item will go back out free of charge. Realistically, if it wasn't baked into the price, then major retailers would be losing a fortune.

There's this drive towards offering free returns so that it will help drive sales, which is absolutely true, and it's wonderfully consumer-centric. What will also drive sales from repeat sales is getting the sizing right if you're in fashion, so that people will then order less because they're not trying sizes, they're just trying styles and designs.

There are things like that, but Will, there's no easy way to say it. At the end of the day, the reason why e-commerce has been so successful is because they're not burdened with huge bricks and mortar cost where you got business rates, and taxes, and staff, and parking, and supply chain, and delivery to those stores.

Stores that are inherently expensive to operate. EComm retailers, a lot of that cost, you can take that out, put it into the price of your products, sell it for a marginally less than it was in the store, have a fatter margin, and give that back in terms of convenience and service, hence the rapid decline of the high street.

Part of it is baked in the price. You can't bake too much in I guess because you haven't got the cost space, especially for bricks and mortal retailers to actually compete with the cheaper online retailers, especially in apparel again.

It all comes down to scalability and leverage though, doesn't it? Can I buy enough stock at an affordable rate? Can I leverage my volume, on every parcel carried to make sure my delivery costs are cheap? Can I direct my consumer to buy products that they're less likely to return? Can I make sure the products I'm manufacturing are desirable in the first place at the right price? It is purely a scale of economics. Simple as that.

How is it managed then? Let's assume in terms of a warehouse, if a portion of your inventory is returned, which especially in apparel is 25%, is that all organized and managed in one corner? How do you get that back into the one point of entry quickly?

In a number of ways. Firstly, if you've got a store presence, and like I said earlier, if it comes back into the store, you check it, and then you're paying the staff to have look and make sure that's okay. They can put that back out on the shelf straight away. And that's immediately made available to the public. Also, in addition to that, if it's a super smart store, it could be made available to people online. There's an algorithm that can pick that up.

Typically, in stores, if that's not happening, it goes back into the normal supply chain loops. The vehicles that would have brought product in will be taking product back, unsold stock, plus stock that's been returned. After that, they may as well as sending products back, they may actually have to pick up other things as well like, I don't know, consumables from the store, and desk chairs, stuff like that.

Then when it does get back in - it'd be the same in terms of the ownership of the warehouse, if the store runs it or if they've got a third-party logistics company: the 3PL will then operate the warehouse, do the open check, inspect, et cetera. May also receive it in, do a quick check because it's been in a transit loop, and then just make that available by cross stocking it into one of the other distribution centers.

If you look at M&S, Clipper would receive a lot of the returns for clothing and home. Once they checked to the spec, they know the stock is available. Then the network will then tell them, "You need to take some to Hemel Hempstead because it's going to be exported." Or "You need to now take a load into Castle Donington because you know what? We're going to keep it in the retail supply chain, and make some of it available for e-commerce."

That's what's happening in the background of retailers. Remember all this costs money. It's not just the cost of buying the product and moving it. There is an inherent cost in terms of managing that and having smart systems, and smarter people than average to know where to placeace it.

What is that cost though? If you just take the average online apparel delivery, say it was £50 in the UK?

Youshould budget 15%. 15% of total cost base. 15% of your total cost base will be supply chain logistics.

That would be baked into that Will, to be honest. In terms of specific companies, you're probably talking-- Again, it's interesting [crosstalk]

Do you see that increasing though or not?

Yes, I can. If returns are going to increase, then it will. Then you also balance that off because those locations that have retail stores and start having less retail stores will actually see their cost base drop. That's why you've got all bricks and clicks now moving towards more clicks.

It's entirely possible, as we saw with Woolworths, that they could become online only retailer. Anyone who is on the high street in the minute, which is your Devlins, your House of Frasers, Sports Directs, not just talking UK there.

If you look at the typical cost of your outbound, you'd always just take your returns rate percentages, and add that to the cost of your outbound for your return. In Germany for example, you expect about half of everything you've sent to come back.

Therefore, in Germany, say for your transportation cost, you would plan for every €5, you would have €2.50 to come back, so €7.50. It's a nominal fee around about €1 per single for pick, pack, and dispatch including packaging, and then the same for the returns process.

If I was a betting man, I would have multiple singles of average out. This is how you build a model, I'm not actually picking on any retailer here. Just in terms of the concept. I know that I sell an average of 2.67 singles. That's a pair of jeans, t-shirt, and a pair of socks, 2.67, average of three.

I would then say, right. Whatever my shipping cost is, so if I'm going out at €3 to make a delivery. I will budget €1 per single, but I know my cardboard packaging, et cetera, et cetera is roughly about 50 cents. Therefore, I put 50 cent per single. So €1.50 per single to pick, pack, and transport is my budget.

Then I also know that I'm likely to receive 40% of that back. Out of my €1, I'm going to add 40 cents for my return journey, and I'm going to add another 50 cents for my repack, and reintroductions. Just on one leg of the journey, you would budget your euro, your 50 cents that's your outbound, your 40 cents, and your other 50 cents. You're going to budget  €2.40 on the high returns market to process a single item.

If you're selling a pair of, whatever your average order value is, if you're only talking €10 for t-shirts, you could actually make that work. Because if you're buying good t-shirts for €4, you've got your cost of doing business on top, as well as warehousing and things like that. And IT infrastructure, let's just say that's going to be another 15%. So, you'd say 30%. On every €10, I've got €3 to spend on cost of doing business, supply chain, and returns.

Therefore, if I'm selling at €10 euros, I've got €7 left to say, "Can I buy this product at €4?" It sounds like it's a 60% markup, and it's not. That's typically how eComm retailers work. For bricks and clicks, it's a much, much, much, more complex process. In fact, it's very difficult to narrow it down to a cost per single.

The only thing you can really do is look at the entire cost of the network, it's fixed asset base, and then divide it by the number of units, orders or sales at the end of the year, and get a measure. This is why there's such cost focus in the industry because if there wasn't, you'd suddenly lose money, and you wouldn't really know why. Very complex business for those ones.

Is that the rough split then between outbound delivery and the return? Because obviously, this one delivery portion is more expensive than the outbound, and then the handling --

Actually, what is it on a per unit basis in terms of split?

Yes. On a per unit basis. Let's give you an example here. Because again, it depends on the complexity of how you pack units, and it depends on how likely you are to get a return. If we're just picking on fashion, it's a relatively low cost because you can put it in polylopes. Packaging is cheap. It's just mainly warehouse, automation, IT, and manual labor.

A euro was the cost of the shipping. If you think your average parcel delivery outbound is €3, typically, your return journey is a little bit more because it costs as much €3.20. Because this is only rough mathematics, then you just budget. If you're saying you've got a 40% return rate, you just budget 40% of your outbound cost. Then adjust accordingly if you see any spikes.

What I'm trying to describe is a really fluid situation, Will, with these guys. Which is why you sometimes see suddenly online retailers can be doing exceptionally well. They've got some killer products, then someone else enters the market and they can't sell that killer product anymore. The units aren't flowing through, so they don't have the scalability. They can't stack it high, sell it cheap. As a result, they start to lose money very, very quickly.

This is the whole dotcom bubble. That's why we have not just dotcoms, but dotcom eComms collapsing.

How much do you pay the carrier per delivery?

That'd be €3. The carrier's charge is never relative to the warehouse. He carries charges relative to the size of the parcels, the service you expect, and the expected distances to travel in, and the average weights. Again, they're trying to do a mathematical sum on how many trailers can we get to collect the stuff? How many parcels per trailer?

What's the densityin location? What does it cost to run vehicles in those locations because Paris is a lot more expensive than running vehicles around La Rochelle for example because of the higher cost of living.

All of that goes into the big pot. So then, it's just literally a "We know we're going to roughly get this kind of volume at this kind of rate spread over these kind of areas. We think as a business, it will cost us X." It is literally X divided by number of parcels per annum equals rate to charge. That's a really high-level example. I'm sure there's many, many data scientists out there that will tell it's a lot more complex than that, but it's not.

So it works out at €3 per average parcel for delivery cost.

That's just a hypothetical. I could be saying, you could be paying €3 a parcel in Europe in general for up to 30 kilos. You could also be paying £7.50 in the UK for small half kilo parcel, but because of that, you've got a lot of insurance baked into it and a very, very high service expectation because you're delivering an iPhone, for example. This is the one thing I keep saying to students.

Same with returns, mate. I keep saying this to all the students, and everyone I've talked to in this industry. Everyone's trying to look to for this magic formula, and the reality of it is it doesn't exist.

There's so many factors in play. There's your customer expectation, the service, your own credit terms, cash comes into it, the kind of products you're delivering, what your customers likely to do with that product when they get it, how long it's out of the supply chain, what your order book is like, when you actually get your money from the customer.

All these, they're very subtle factors because all of that leads into the position of leverage that you've got over your providers. This is the sort of stuff that they're looking at. Where does the customer live. Tricky to say that at a high level. It's the base one I would work from, but the variables can change.

I would work between 50 cents to 70 cents with the cost of pick and pack, and the cost of inspection for a returned item. Think about that. If you've got a €1.70. You've got 70 cents, say, for outbound. You're expecting 40% of it back so, let's just say that's 30 cents returned.

Last Mile Delivery Challenges

Warehouse and logistics network setup for traditional retailers

Inbound to outbound processes for traditional brick and mortar retailers

Structural advantages of online versus brick and mortar retailers

Process of managing returned products throughout the supply chain

Last Mile Delivery Challenges

Former General Logistics Manager, Marks and Spencer

inpractise.com/articles/last-mile-delivery

Why is this interview interesting?

Executive Bio

James Doyle

Former General Logistics Manager, Marks and Spencer

Interview Transcript