Interview Transcript

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.

Allan, can you take us back to 2001? What originally interested you about joining Amazon?

Going back to 2001, there was a step before that, which was in 2000. I had been working at Apple Computer since 1995, and we were going to be transforming the whole supply chain, a country-by-country approach to get to a global approach at Apple. Around 1997 to 1998, we worked on the first Apple online store. I think it was launched in the US in late 1997. How quickly has technology changed? Those were the days when you were plugging the mobile phone in, and you were going through dial-up, and you were intrigued by email dropping in one by one, but it was the start of ecommerce. We worked on the first online store, and it was really about the convenience for customers that attracted me to online, which was at Apple.

A guy I'd worked with at Apple had joined Amazon in about 1999. I met Jeff Wilke and Jeff Bezos in 2000. I didn't take the role for family reasons at the time. Still, I got an insight into a disruptive approach to improving customer experience and improving convenience. I was convinced that there was an incredible opportunity there from the early days, so by good fortune, the role came back around in early 2001, and I joined Amazon in April of 2001. I just really thought it was fantastic; really inspiring to listen to the vision of Jeff Wilke and Jeff Bezos and a guy called Diego Piacentini at that time.

What was so inspiring about the way they articulated their vision?

Many things come to mind. They were driven by a customer obsession that was truly different from anything I experienced at Apple. There was a pace, energy, and an authentic customer obsession. How do we truly make a brilliant customer experience for customers? There were no hints of bureaucracy in the company. It was very fast-paced, and it was striving to become the world's most customer-obsessed company. It felt authentic to be part of that.

Was it driven by Bezos himself, or did that also come from below?

I think it came from the team, to be honest with you. Many of these companies are defined by their chief executive, but I would tell you there was a consistency across the executive teams. All the executives in position at that time were singularly focused on a number of objectives, and customer experience was number one.

Every company these days says they're customer-obsessed. What examples do you have that stand out that these guys are really serious about this?

A couple come to mind when I think back to those days. There's no question at all; it was in the DNA of the people. We reviewed metrics every week with the company's executive on the customer experience; how many were satisfied, and how many were dissatisfied. We had this thing called free replacements. When you go back to those days when a customer experienced a defect, we replaced the product for free. That was unique at that time, as I recall.

We built some decision logic into the software of the company and the supply chain where, if we had a product dispatching to a customer that was more costly to Amazon but would meet the customer experience commitment, the logic would favor that route at the cost of Amazon. We truly believed that the long-term loyalty that a great customer experience would drive was much better than a short-term financial gain. Many things were built into the software in favor of customers; free replacements were an obvious example where we went the extra mile to ensure that the customer experience was fantastic.

How would you describe the differences in Amazon's philosophy to build a fulfillment network versus the existing incumbents of the time?

Going back to the early days, the UK and Germany opened in 1998, and France followed. When I joined in 2001, there were only these three companies, and there was this direct link between the websites of France and the distribution center or the fulfillment center in the country, so really, there wasn't a network in Europe at that point in time. That's when we started the journey, and there was this one-to-one relationship between the fulfillment centers and the country.

Our first expansion in the UK was in 2004, when we opened up in Scotland. I am a native Scotlander, I remember it very well. We saw, by the growth of the business, that we were going to have to scale fast, and the second expansion was in about 2006 when we opened our building in Leipzig in Germany. That's when we started to drive the standardization. But the network began to evolve over those years, and we did things like, every single fulfillment center in Amazon could dispatch product to anywhere in the world that had a postal code. The notion of the network and having a kind of global topology, a global set of buildings from which we could dispatch our products to anywhere, that vision was created very, very early in the journey. Over the time I was there, we were constantly and relentlessly working to establish and improve that network. That network provided us with flexibility.

In the UK, to throw out one example, we had same-day delivery. We had a guaranteed next-day delivery, and we had a super saver delivery, which was a three to five-day service at the time, giving the customers great choice. Depending on what the customer chose, our systems would be able to work out the optimum route for that product, based on the customer's service. Today, Amazon is still working on improving and building on that network. It started with no network and evolved. When I left Amazon, we had 22 buildings across the UK, Germany, France, Italy, and Spain dispatching products worldwide.

If I ordered now in London, it's not necessarily the fulfillment center closest to me that's going to ship my product?

That's exactly right. We positioned our inventory around these different fulfillment centers. When the order drops, a series of decision-making in the software logic looks at, first of all, to your earlier point, what's the customer request and the customer commit. And therefore, what's the optimum route for the customer and for Amazon to satisfy the customer. It may well be that the inventory could dispatch, in this example, from well north of London and arrive on time in London. So no, it didn’t position all inventory in London for London customers. It looked at optimizing that network. And of course, to your question, that was us starting to build a network of fulfillment centers across the globe.

How did you think about optimizing inventory placement?

I think you look at your inventory profile, and you look at categorizing that inventory into sales velocity. You look at fast-moving, medium-moving, and slow-moving product. You also look at the dimensional data of the product; what's heavy and bulky product and what's small product. You build that inventory positioning based on where your customer demand is. Customer demand typically follows population density.

Then you look at the storage solution. An example for a very slow-moving large product would be a trampoline, gazebo, or hot tub. That storage solution may be different for a large, slow-moving product than for a small, fast-moving one. You looked at your inventory profile, you looked at where your customers were, you looked at the sales velocity of the inventory, and you made some decisions based on a number of factors.

What’s the storage solution for large and bulky items at Amazon?

One of the things we used to talk about, as I moved on in my career into grocery and then on to Wayfair, into heavier and bulkier items was that, an item's an item. It has a certain set of characteristics. In Tesco, we dealt with fresh, frozen, and ambient products. We were dealing with eggs, or the example I just mentioned, we may be dealing with a gazebo. You have to understand the characteristics of your inventory profile in terms of sales velocity and dimensional data. Is it fragile, hazardous, bulky, or is it awkward? Look at establishing our storage solution for that range of products and what's our throughput, which is the pick, pack, sort, dispatch, and delivery processes.

You build a process path for the product type that any company is looking at. One of the things that I learned over a period of time is it's just a product, and whether it's coming in from an FBA vendor, a CastleGate vendor, and whether it's an egg, a bottle of water, a head of lettuce, all the way through to a gazebo, it's a product. And then, you optimize the process path and the technology for that product.

How did FBA change how you thought about optimizing the network, given that you now had first-party and third-party inventory?

From a supply chain perspective, it's just another product. Now, look at the evolution of the supply chain. In the early days of FBA, when we didn't have the standards in place, I remember some suppliers would arrive unannounced and deliver a large caravan full of books or music or DVDs that would be kind of dumped on the receiving dock. But of course, going back to the point I made just now, we had to establish process, establish standards; what type of labeling, what type of packaging.

I loved that evolution because you had to innovate. It was a fantastic way of increasing selection on the website; you connected suppliers with customers which was great for everybody; great for the supplier, great for the customer. From a supply chain perspective, it was just another inbound lane, and we had to put in place the right processes and the correct standards to manage that variation. That's something that evolved, so whether it’s FBA or CastleGate or touchless inventory, you develop a process and a standard for that inventory. To me, it didn't matter whether it was FBA or not FBA. It was an inbound lane.

Did you store both FBA and third party in the same warehouse, with the same balance?

Absolutely, we did. Again, back to that set of philosophies or principles on the inventory profiling, we captured the dimensional data, we captured the sales velocity, and we had software built that used the data and understood where every single item was in the fulfillment center at any given point in time. We could well have an FBA item sitting next to a purchase item, and that would be in that location, not because it was FBA or not FBA; it was in that location because of the dimensional data of that item. The software accuracy and the discipline around the process drove accuracy into that inventory positioning. We were constantly measuring to ensure that our accuracy was incredibly high on the putaway process or the stow process, therefore driving a much greater accuracy on the pick. Leaning away from the FBA or non-FBA section, it was more about the position of the inventory based on the dimensional data.

What was the biggest challenge in scaling FBA?

I think it was the lack of standards and processes in the early days. We put a huge emphasis on a very robust sales and operations planning process so that we knew what the total inbound volume would be and what the categories were of that inbound volume. And we knew what the outbound volume would be, so it was another approach where we broke it down to what's the process, what's the technology, and what did the people process take for each process path.

In the early days, there were no standards. When I left, we had made great progress on the maturity of those processes and standards. I think it was the diversity of the product because, again, it was brilliant for customers because we had a much wider and broader depth of selection. It was the diversity of the categories and making sure we had the right processes to deal with that.

How can brick and mortar retailers learn from Amazon in how they've organized their supply chain and fulfillment network to compete today?

I think brick and mortar can learn from Amazon, and I think Amazon can learn from brick and mortar. It’s not that one is fundamentally better than the other, but they are different. In Amazon and most ecommerce companies, as an example, there's talk of forward-deployed inventory and brick and mortar stores have been doing this for hundreds of years. It's called the store; it's called the shop. Usually, these stores and shops are very close to population densities.

I think there's an opportunity to take the best practices from other ecommerce and brick and mortar to drive a better customer experience, whether it's coming from ecommerce or brick-and-mortar companies. When I reflect on what brick and mortar was – this is a bit generalist and may not be entirely fair – many companies hadn't invested in the technology and in the data to truly optimize the customer experience. I would say that was the opportunity there for the brick and mortar to think long-term, to make that investment in technology to optimize the supply chain through brilliant processes, brilliant technology, and building great teams. But I think each of these customer experiences can learn from each other.

Let's say you're a similar business to Shopify, and you have a huge customer base, merchants, and transactions. How would you approach building the fulfillment network today against Amazon?

The supply chain comes down to several principles that you might want to be thinking about. In the end, the product moves from a manufacturer through a set of steps to the customer. You look at your customer demand and, personally, I always start with the customer and work backward. Where is your supplier base, and basically, how do you simplify that connection? How do you bring accuracy to every step of the process, so you know where that is. I would start with my understanding of where my customer demand was and my supply base, mapping that out, looking at the growth opportunities, and looking at the customer proposition on the speed requirements. Are you offering same-day service? Are you offering a timed service? What reliability are you offering? Then you build your network based on that, looking at, again, sales velocity and the product and inventory profiles that any company has.

What could be the most difficult part of building out the network for Shopify?

When you're scaling a network, you’re looking at all things storage, all things throughput, and all things people. When I talk about throughput, I talk about inbound, the processing within the fulfillment center; the pick, the pack, the sort, the dispatch. They're also looking, very much, at any middle-mile activity you may need and any last-mile activity you need. You look at every element of the process, understand each element's capacity, understand exactly how many items or how many customer orders you can process through that.

I would add the people and recruitment elements to the scaling question. Where can you attract and retain the best talent possible for your network? It's all things people. It's the receivers, the people that do the putaway, the people that do the pick, the pack, the sort, the dispatch, the final mile delivery, the customer contact, the returns process. You again break it down to those process paths, but in my experience, it wasn't the automation; it wasn't the process definition or the standardization. It was the recruitment challenges and ensuring that we always remained ahead of the curve in terms of getting people on board, getting people trained, and retaining the talent. And sometimes, we skip over this point, but I think scaling the people has to be a high priority for any leaders in supply chain today.

How would you compare the supply chain philosophy at Wayfair with Amazon?

Of course, there are some nuances. I would say, fundamentally, regarding moving products from manufacturers and suppliers to customers, the approach is to build. I'll go back to the consistent approach areas at Wayfair and Amazon. When I go back to my Apple days or anywhere, it is really to build very solid foundations of processes and standardized processes; real operational excellence in the definition of those processes.

The second big area is the definition of the technology, so the process step and the technology step should be absolutely hand in hand. We should look at every step of that process. The next step is your customer, and to start that philosophy of the customer experience right at mile zero, in fact, at the supplier. For example, the customer of putaway is pick, the customer of pick is pack. Really think about how you orientate your processes with that customer obsession.

Is the process materially different on large versus small items between Amazon and Wayfair?

Again, I'm thinking through the approach to it. For that bulky item – let's use the gazebo or the hot tub – it has a certain set of requirements around storage, throughput, and labor. Then you optimize for that product type, so the philosophy and approach to these items are the same. The physical deployment of the storage solution for an egg or a gazebo is, of course, going to be different. You don't want to put a gazebo at height in your building, and you also want to put a fast-moving product closer to the sortation place, these types of things. When you ask about the supply chain philosophy, the philosophy or the principles are very similar. You adapt or adopt a different physical deployment based on the product type and the sales velocity, but the philosophies were very similar.

How do you compare FBA and CastleGate given one if for large gazebos and the other for small books?

I might even sound a little repetitive here, but from my supply chain perspective, I think of both as an end-to-end supply chain. In the Amazon example, FBA was a mechanism to connect suppliers to customers and expand selection. Similarly, in Wayfair, the CastleGate service was a fantastic service for suppliers, it was fantastic for customers, and it was fantastic for Wayfair. In my supply chain perspective, it's a product range, and therefore, how do I optimize my processes and technology for those products range. I don't think about the different commercial elements. I'm thinking about the products and optimizing the end-to-end supply chain.

Let's take the first mile in Asia. You have some suppliers, whether Chinese general merchandise sellers or furniture manufacturers. Are the individual steps in the process, the drayage onto the ocean freighter, very similar for large and small items?

It’s a process thing. For example, in Asia before the pandemic – going back to the principles we discussed earlier in the discussion – I was working with manufacturers to understand how we take waste out of the process. When I talk about simplification, in simple terms, every time we hand over a product, it's an opportunity to introduce a defect or a cost. How do we reduce the number of handovers, going right back to the supplier? How do we get the suppliers' demand signal to be cleaner? How do you start that forecasting process?

Using an example, going right to the MRP level of a supplier and connecting it. How do you eliminate waste? And that goes back to one of the principles of simplification; touchpoint reduction, defect reduction. You map out for the bulky product or the book. How do you optimize and reduce the number of steps in that process for that product type? The approach, philosophy, and principles are very similar. How do I optimize for my gazebo, and how do I optimize for my book? But think about it starting at the source. People often talk just now about the last mile. First mile is equally important. Go back to the source of the supply.

If we just go to the source then, how did you find the Amazon approach to large and bulky items different from Wayfair, given that Wayfair is, I guess, a specialist in home?

When I was there, what we called the heavier, bulky items were a smaller piece of the puzzle. We were focused, like a laser, in optimizing the smalls process. It was a conscious decision to go after that volume, so I wouldn’t be able to give you a completely detailed answer on how the supplier connection was available, but we did put in place programs; we talked about that in sales velocities. For example, you have to build flexibility for things like seasonal peaks, so let's talk about a program we had called Vendor Flex.

Vendor Flex was a program that looked at how we optimized that supply chain. How do we reduce those touchpoints for things such as large-screen televisions, as an example? We looked for some suppliers at that seasonal peak, where we deployed a team of Amazon people to the source of that large-screen television and took the order in there, in that vendor, slapped on a shipping label, and shipped it direct. That was skipping the notion of shipping it to the Amazon fulfillment center, receiving it at the fulfillment center and dispatching it from there. We cut out four, five, six, seven different nodes in that chain. That’s an example of working with the supplier to help the supplier, to help the customer, which those types of initiatives did do, but the philosophy and approach were the same. How do I simplify? How do I reduce touchpoints?

In the first mile, what do you think of the potential advantages that Wayfair could take in that specific category, given they're specializing over the likes of Amazon in large and bulky?

I think the Wayfair model is all things home. By definition, it ended up the opposite of what I just talked about with Amazon, where we focused on the small parcel. It was just the volume that was there then. At Wayfair, with the furniture category, with our beds, sofas, tables, chairs, etc., by definition, we had to focus on how to optimize these process steps for bulky products. That’s what we did, but again we would look towards the material handling equipment that would reduce lifts; safety is first and foremost and a priority at both Amazon and Wayfair. How do we move that heavy, bulky item in the most efficient way?

We ended up deploying many initiatives that would ease the movement of that product through the supply chain, right to the customer. We worked on the way we would package it, the way we would consolidate the freight in Asia, the way we would book that freight and load the container, the way we would receive it into a fulfillment center and bring in product.

How is that different? Take what you just said, receiving it from a supplier, filling the container. Would it be materially different at Wayfair than it would be for Amazon?

Honestly, I'm unfamiliar with how Amazon does it now.

But they wouldn't split up their large and small?

When I was there, we had dedicated buildings for bulkies; we would receive that bulky product to those dedicated buildings. We also had some mixed buildings when I was there. In Wayfair, we have the flip. Most of the buildings are bulky buildings, and we're receiving those bulkies, so you optimize a bulky receive process, a bulky storage process, but we also had mixed buildings. We had several buildings positioned across the network that could dispatch light bulbs and gazebos from the same building. We would set up the receive process for small, medium, and large, and the receive process for bulky items.

What’s the most difficult part of the first mile for large and bulky?

I wouldn't put it in terms of difficulty. I would say it was a process step that we had to optimize. We looked at the best way of moving that product from A to B; what's the best way of dispatching it? I'll come back to again to the issue of, how do you optimize that supply chain? We worked with manufacturers, we worked with suppliers, we went right back to the manufacturer and its process. Building on what I said earlier, you take the mentality or the approach, at every step of the process that the next step is your customer, and you start to understand the customer pain points from the source.

We would go into the manufacturing sites and look at your next pain point. The next pain point is loading the truck in the yard. Going back to the analogy that the best person to tell you how to improve picking is the person who picks all day long, we would load that truck together with those people. The best person to tell you how to load a truck from a manufacturing source is the person that does it. We would deploy our leaders, our engineers into the manufacturing site and the suppliers to understand the customer pain points at every point in that supply chain. How do we make it better? How do we simplify it, how do we take the pain out, how do we reduce the touchpoints?

What were some of the pain points for those furniture suppliers that you found?

For a lot of people, it would be lifting. Many of the pain points were in the packaging process that we improved. We learned, through that approach, where the damages would be induced and where we had difficulty. When I talk about things like throughput, what’s the productivity associated with this? How do we load, in a safe manner, fast but more effectively; maybe not even faster but more efficiently? How do we optimize that? You look at having the right metrics.

Let me give you an example of this notion of each step of the process. Your customer is the next step in the process. When you’ve started, people might have looked at optimizing a truck fill as a metric, densifying that product into the truck, and congratulating themselves on the silo of that metric. You cannot get another square centimeter or square inch into that truck. However, you need to look at the end-to-end metrics in a balanced way. How efficient is it to unload that truck? How much damage has occurred in transit in that truck due to the packing? This is when you start to look at the end-to-end process and have the right metrics in place to drive the right behavior. The right behavior is always, first of all, as a priority, safety. That is followed by your customer experience metrics and your employee experience metrics. You need to have those end-to-end metrics, those end-to-end processes, and understand your customer pain points. That is the process and approach you take.

Which part of the process do you typically get the most damage for large and bulky items?

That's something we made incredible progress within Wayfair; we had a full team that was passionate about making sure that we reduced damages. Damages can happen anywhere; as I mentioned earlier, the more touchpoints, the more opportunity for damage, and the more opportunity for costs. So we put processes and audits in place to understand where those touchpoints were and where those damages were incurred. It's typically involved in the movement of product in and out of containers, in and out of trucks at the fulfillment centers or the middle-mile stations, or the last-mile stations. Therefore, you look at how you eliminate that step altogether, or if you can't eliminate the step, how do you optimize the step?

In this example, we put in different types of storage equipment that would hold the material in a better way, in the location. Different material handling devices would help move that product much more efficiently and safer. Each step in the process can induce a defect, and it's about understanding that. This is one of the foundations I talked about. It's that culture of defect reduction. A defect is a good thing. You encourage your team to identify that defect and remove it. You praise and reward that behavior because taking the defect out of the system as early as possible is a win for everybody. The ultimate failure of the defect is it gets to the customer. Those are examples of improved storage, material handling, packaging, and metrics around how to load and unload a truck.

Just walking through the processes and philosophies that seem very similar between Amazon and Wayfair, they both focus on defects. What advantages would you say Wayfair, specifically, has over Amazon in shifting large and bulky to reduce the touchpoints and the defects in the first mile and fulfillment?

I wouldn’t want to draw a direct comparison, to be honest with you, and say one is better than the other. I would say that the process focus at Wayfair was the right one, which was to be the best in the world at home delivery for all things home. Therefore that obsession around the customer experience led to an optimization of the Wayfair processes, which improved the way the heavy, bulky product was received, stored, and delivered. I wouldn't be in a position to say how Amazon does it today. Amazon is very good at what they do, but I would emphasize the nature of the inventory profile that Wayfair demanded optimizing each of these steps. Therefore, they became very, very good at it.

Can we walk through the bulky process? Let's say I'm a supplier, I understand the product is shipping via the ISC, the international supply chain, it goes through a consolidation center in Asia and then is sorted, to then be placed in a container.

The first step in the process starts with the manufacturers. The manufacturer manufactures the product. We wanted to get very close to the start of the supply chain. In this example, an Asian one, the product is manufactured and moves into a consolidation center. One of the opportunities is to consolidate that freight from multiple suppliers in Asia, to optimize that next step in the process, load the container, the container goes onto the vessel, and position those vessels into, for example, San Francisco for the west coast or into Savannah for the southeast and into New York for the northeast. We had multiple points in that process that would bring that optimization. Still, it started with a manufacturer; it started with the consolidation of the freight in Asia, in this example.

How integrated can you get with the production? Again, if I’m a supplier, do you know the units I’m going to churn off my manufacturing line, for example?

Every good company I've worked for puts a huge amount of emphasis on building a fantastic sales and operations planning process. It looks at demand forecasting, what products are hot, what new categories are coming and how to add selection. We also, therefore, equally look at what's the supply planning. It's about driving accuracy into that whole process, looking at the understanding of the availability, and that comes down to connecting suppliers with brilliant data and brilliant visibility into when the products are arriving.

You want to start with that sales and operations planning process. It runs weekly; it's always looking at the current week's performance. It's looking at what that means to the two-week forecast, what that means to the monthly forecast, the quarterly forecast, and the annual forecast. This is an iterative live process that's constantly looking at demand and constantly looking at supply. Then you want to really understand and drive visibility and data through your technology into exactly the status of suppliers, exactly the status of inbound lanes, exactly the status of your inventory profile, and exactly the status of your demands. The dream is a real-time dashboard that connects every supplier with every customer, and you're watching it live from your laptop from wherever you are, and it has the daily and hourly iteration of your demand and your supply.

Interestingly, both Amazon and Wayfair are taking very similar approaches. They're moving upstream into the supply chain to own more of the inventory and understand more of the data at each point. How do you compare that, given that it seems like Amazon takes more inventory risk because they own first-party versus Wayfair, which kind of takes no inventory risk?

The way I would think about it is that the ownership model is a commercial decision in how you want to approach these kinds of things. What's important for me, from a supply chain perspective, is the visibility and the data, and how a company sets up commercially is for others to comment on in a different interview. But from a supply chain perspective, what I needed was data and visibility and accuracy of that data. For example, through a connection to suppliers, everybody's working very, very hard to drive accuracy, right at the start of that process. Accuracy is great for suppliers, it's great for the customers, and it's great for Wayfair in this example.

The way I would look at and answer that questions is, it’s about bringing the tools and the processes necessary to drive that accuracy into the process where everybody benefits. Because in simple terms, if you can't measure it, you can't manage it, and to be able to measure it, you have to have the data. That's how the supply chain's set up, and there's an extremely efficient and effective way of doing it.

What is the biggest improvement you saw from Wayfair in taking ownership of the middle mile versus using a third-party trucking company?

I’m also giving my personal view on these things. I think operations are kind of the crown jewels of ecommerce, and supply chains are the crown jewels of ecommerce, in my opinion. Therefore, I would have an approach of reducing dependency on third parties. Of course, you need third parties, but reduce the dependency on them. Own your intellectual property. Own and control the destiny of that customer experience. I often use the term, you never see anybody washing a rental car. You see people washing their own cars. You own it. You own and build that customer experience in, so having a last-mile delivery for something like furniture and controlling your own destiny in that, is something I believe brings a huge amount of value into the entire process, the middle-mile and the last-mile product.

As you scale the platform and the supply chain, it's important to be thoughtful and prioritize where you want to control your destiny, what data you need, and the IP you want to control. Then you bring that in-house to ensure that you drive a brilliant customer experience; this is a personal view of mine that I would want to continue to drive in future supply chains.

Is the cost difference materially different when you own the middle mile versus using a third party?

That's going to be a function of volume; how much volume you put through and you build the capacity. Again, it goes back to that sales and operations planning process. You want to make sure that you provide enough capacity, never too short but never too much, because there's waste in there. You need to be really robust and drive continued accuracy in that process and then build the right amount of throughput, as I mentioned earlier in the discussion. The throughput, in this example, is the throughput of the fulfillment center, but it's all things throughput. It's throughput of the middle mile, it’s throughput of the last mile, so it’s all connections of how you move the product from A to B and establishing the right capacity in that.

The commercial element of it will be a function of volume. One of the great things about Apple, one of the great things about Amazon, and one of the great and consistent things about Wayfair, is that these companies are thinking long-term. They’re making decisions for the long-term, and if you look at the Amazon example, who would have imagined Prime Air back in 2005. We started the very first early deliveries of Amazon logistics back in the UK, I think around 2012 or 2013. At the point in time, the economics of two or three parcels were, of course, ridiculous, but the company had an incredibly brilliant long-term vision of what they wanted it to look like. That was derived from understanding the capacity required for those packages at scale in 2015, 2020, 2025, 2030.

Would you want to rely on third parties for all that capacity? Are those third parties investing in the seasonal capacity and the capital needed for the seasonal peaks that the scaling and growth require? That answer is typically no. Therefore, when you take that long-term strategic view, you will position the capacity that comes back to the prioritization point I made. A decision was made, probably in late 2008-2009, to invest in developing an Amazon logistics, which has turned out to be an enormously successful and strategically brilliant move at that time. That's how I would answer the question.

The middle-mile or the final-mile economics in the very short term, they are what they are. Suppose you're scaling a company to $100 billion, $150 billion, $200 billion, for example, in-home products – not to mention Wayfair, but I'll throw that out there – you want to understand what capacity looks like in every process step from supplier to customer and the returns process in that journey. Then prioritize and choose where you're going to start investing now to ensure that you have a brilliant customer experience that is also competitive in five, 10, 15 years from now.

I've kind of morphed into this, but I want to capture one of the learnings I’ve had before we close; companies are investing in technology. Companies are understanding technology and the consequences of that technology on customers and truly taking a long-term view. Of course, managing the daily and hourly tactical stuff is best positioned for success, in my opinion and, less so, companies focused purely on the financials of this quarter, the financials of next quarter. Really make that investment for the long-term.

Like you said, it comes back to providing the best customer experience.

Totally. What's going to be competitive in five years. Another thing that I think ecommerce has brought around is customers are loyal until somebody does it better. There's always a disruptor out there, so companies need to be nimble. Companies need to be continuously innovative, and in fact, the pandemic forced an incredible innovation for the supply chain. How do you onboard 10,000 people remotely to support the demand? We invented ways to do that, and we were extraordinarily successful in that process. Always keep that agility, be nimble, innovate, simplify, always focus on your customers, focus on your employees, and focus on safety and keep these principles at the core.

And it seems like Niraj and Steve do have that customer obsession at the core of Wayfair, and WDN, the last mile, which seems to be somewhat similar to Jeff Bezos 20 years ago, and Wilke and FBA. How do you compare the strategies and leadership of, say, Wayfair in the last few years and Amazon back then?

What I'm going to tell you is that I admire both companies incredibly well for the focus and the energy, and the relentless obsession that they've got in each of the markets that they are serving. Like your question on brick and mortar, there is room for choice in the supply chain and for these customers. I admire the passion, focus, and innovation that each company, including Apple, has brought to it. That's why customers keep coming back, and unless you keep that focus on the customers and employees and keep that innovation, the customers won't come back. In the end, it's a mechanism of survival because if the customers won't come back, there's no operation. I would answer to say I admire both approaches, and there's room for both in this world, and I'm glad there's choice and customers have choice because that's what keeps us hungry, that’s what keeps us innovating, and that’s what’s keeping the supply chain running right now.

As you said, it's the customer obsession. The focus on the customer experience lends itself to a more fully vertically integrated model where you own each part of the steps that can scale and provide that great experience. That’s the moat, in the long run, it seems.

I think that's right. I think, pardon my oversimplification, but you've got platform companies, product companies and customer experience companies; companies that are brilliant at each of these. If you can do more than one of them, if you can have a platform that sells brilliant products and a brilliant customer experience, it’s the old flywheel. Customers are always going to come back.

That's why I would go back to your point on vertical integration. I would start to prioritize which parts of these are going to scale to $100 billion and which parts aren't, and therefore, where do we start prioritizing and investing now to ensure that customers’ experience not only continues but improves. Every competitive environment will only get more competitive as the world continues to innovate on platforms, products, and services. I think it's an incredibly optimistic and fantastic time to be part of the supply chain, and I think the future is enormously bright for the supply chain moving forward. Of course, it's going to be tough, and of course, it's going to be hard, but I very much look forward to it.

Last question, Allan, where would you place Wayfair in those different buckets that you had like platform, customer experience? How would you place Wayfair?

It's a technology platform company that offers fantastic home products to customers all around the world, so it's a company that's firstly, a platform; secondly, has great products on it through the suppliers; and thirdly, is driving a great customer experience. Wayfair is ticking all those boxes.