A Glimpse of Amazon
Visiting the Mountain
I have been to the mountain!
In February I attended the quarterly meeting of the economics group of which I am a member. During that meeting, I gave a short presentation on the trucking market. In attendance was the Chief Economist from Amazon who was impressed by what I offered and asked if I would care to visit Amazon. They like to invite various research experts to hear what they have to say and to explore ways to mine such expertise. Of course, I accepted and, accordingly, I visited their Seattle offices last week. While some of what I saw must remain, understandably, confidential, I can report these observations of a remarkable company – a company unlike any I have visited in fifty years of working life.
Here’s a new word for you.
At least it was for me. I learned that Amazon divides transport into three buckets: “First Mile” “Middle Mile” and “Final Mile.” This taxonomy teaches us something about the evolving transportation world. Historically the bulk of interest has been on the line-haul portion of the move, where goods move in high volume, long distances. Amazon introduces us to a world that explicitly includes origin consolidating activity and the essential delivery of goods to a specific consumption point. That viewpoint forces one to see the entire move and to recognize the special circumstances of consolidation and distribution activity at either end of the move.
Amazon is a big gorilla!
The day was an eye opener in several ways. First, that company is very big and moving very fast in many regards. They are approaching 600,000 employees worldwide, having added 250,000 people in the last year. The flip side of those numbers is the company has bunches of new people who are understandably ignorant of basic facts, including transport. During my presentation, the more experienced people stopped me several times to explain to their new colleagues what some of my terms mean, like ELDs and backhauls. One person I met had six months of experience at Amazon. He told me that a third of Amazon people have been hired after him. There is also a significant bureaucratic load, tied to a desire for control.
Amazon is convinced that brains equal profits.
With that horde of people, there is a heavy core of analytical experts. Part of that is the nature of their work and growth plans. Part is simply a powerful bias towards continuous improvement based on smart people analyzing and improving process. If only 5% of their employment is so inclined, that equals 30,000 people, a phenomenal investment – and, I think, advantage.
Amazon employs more than 100 economists, using them principally as project managers, given their analytical skills. That practice appears to focus talent on tangible improvement. The experts come from a variety of backgrounds: one I met was a chemical engineer. I met another person who was a micro-economics professor from the University of Washington. He was on a two-year loan from his university job.
Importantly, the general manager of their line haul freight operations, “middle mile,” talked at length of the analytical projects he was executing, including several focused specifically on freight matching activity. They have, for instance, improved their load acceptance by allowing their carriers to choose appointment times, from a menu of options. The previous arbitrary assignment process left their carriers unhappy about time-definite appointments, given the specifics of their drivers’ trip circuits. I have never seen such an investment in analytical work from any carrier, broker or shipper in my forty-two years in the industry.
Amazon moves a lot of electrons.
Underlying that analytical bias must be profound data handling capability. Between their in-house products and their consignment products, Amazon must physically inventory 350 million SKUs. For each SKU they create a specific demand forecast in order to determine stocking levels. Obviously, the majority of those forecasts are machine driven. Still, each one must be created somehow and monitored and linked to ordering processes. Talk about big data! In comparison, Walmart manages 4-5 million SKUs.
“Do it my way!”
I detected a strong bias towards control. In transport they prefer small fleets to large ones; they prefer power-only carriers while they provide and manage trailer supply. Importantly, capacity management is their driving goal, ahead of cost minimization. That figures, given their explosive growth and their high-service retail strategy. The need for control is also related to their focus on process rather than results. As Walmart is a distribution company that happens to have stores, Amazon is a supply chain management company that happens to have a website. My IT friends suggest they are a cloud data management company that happens to sell goods.
Amazon is very young.
In addition to the aforementioned bureaucracy, I detected an understandable naiveté in some of their transport initiatives. Without describing the specific initiatives, the lack of experience that so young an organization necessarily has may cause them to make inappropriate assumptions about market structure that will have real costs. They assume that Amazon solutions are always superior, even where eighty years of market maturity has already developed excellent solutions. This is a troubling viewpoint for a company already with a strong control bias. Why, for instance, does a supply chain management company insist on investing in and managing trailers when the market is loaded with people who know how to do that? Surely, they can obtain the necessary capacity levels at low costs if they learn to properly communicate with the potential suppliers.
Bentonville, AR had better be worried.
The Walmart comparisons are instructive, I think. Sam W revolutionized basic retailing by building a distribution system that can fill a large store with a comprehensive variety of cheap goods. He made the big downtown department store obsolete by providing an outlet both more convenient and way cheaper. The only drawback to Sam’s model is the limited selection of any good, the tawdry appearance of the stores and the long lines at checkout. Those were the results of the driving cost focus.
So, the sum is using modern supply chain management tools to give adequate selection at low cost, assuming minimal amenities. Now, along comes Jeff Bezos who sees that Internet connectivity can flip that tradeoff on its head to emphasize the far more lucrative engine of service rather than low cost. Amazon has one hundred times the products, all offered with no lines, plus an increasing collection of online amenities. His business development model was to start with the specialty products (books, replacement parts) where inventory economics made traditional distribution more expensive and way less convenient, given the need to search hard and often settle for a less appealing substitute. Now, progressively, he has used the process excellently that I reference above to take cost out to steadily increase the number of SKUs available at reasonable cost. Moreover, the market is discovering that the convenience of Internet ordering is sufficient to cover some extra cost, further expanding Amazon’s market reach.
Back to the future.
Someone said recently that Amazon is the modern reincarnation of the 1880s Sears catalog. Supply chains in those days were incapable of stocking anything but the minimal basics anywhere but in big cities. With the advent of cheap printing, reliable mail service and the telegraph, small-town consumers could then gain access to those big city inventories. Finally, the railroads and the Wells Fargo wagon could get those products to the small towns in reasonable times at reasonable costs.
Of course, as digital tools improve, this new Sears catalogue is aiming at performance that will deliver goods in excellent times at lowest cost. Amazon is implementing same day service, as I write. With seventy U.S. fulfillment centers, it can offer next-day service for more than a million items, many arriving the same day. How many of those million SKUs do you have to have sooner? Some milk? Some bread? A part for your deck project? Amazon also knows that we are moving to a place where your Internet ordering process will manage your staples inventories, the bread, the milk, and order all the parts you need for your deck beforehand. The ultimate service level is not needing to source any last-minute supplies.
In summary, the comparison is a powerful one. Sam’s breakthrough was to provide a high percentage of our basic needs at low cost. The emphasis was on basic needs and low cost, with an unpleasant shopping experience as a necessary side effect. My wife and I put up with Walmart stores because they carry all our basic products and the products are cheap. We use a lot of bird seed. Sam’s strategy was essentially bottoms up and had a hard ceiling, the four million SKUs that could be handled in superstores. Jeff’s breakthrough was to see that an Internet-based supply chain could offer everything at reasonable cost and with a very convenient shopping experience. Ever use the one-click feature on Amazon.com?
He is likely, very likely, to catch and pass Walmart for two reasons. First, his model is service-driven and is based on revolutionary intimacy with the customer. He can use what he is learning about us to further stimulate demand and further lower cost. The only Walmart person that knows my wife and me is the checkout clerk nearest the bird seed shelves. Second, the hard floor to his top-down strategy, delivery costs of heavy, bulky items (those bird seed bags) is steadily receding under the pressure of supply chain innovation. We already order some of our birding supplies from Amazon. Soon it will be all of it.
So, what does this all mean for Trucking?
- Join the analytical revolution: Amazon is demonstrating the tactical value of analytical resources. That is to assign them to specific business development projects where analysis is directly tied to tangible outcomes. They are proving that the exploding availability of data, powerful digital tools, and smart people can take supply chain performance to new levels. It troubles me to have found evidence of improvements there that traditional practitioners should have adopted a decade ago. We in the industry must make those same investments or be left behind. Importantly, Amazon has the scale to perform any transportation function in-house should they choose to. That they have a strong inclination to bring those functions in-house means they will make such choices frequently if their current suppliers persist in traditional behaviors. Suppliers to Amazon will be increasingly relegated to low-value, subordinate positions.
- Leverage our transport knowledge: Amazon knows a lot about supply chains, but the inexperience of many of their people leave them ignorant and naïve about truckload transportation. Moreover, their preference for small carriers provides them visibility to the least sophisticated segments of our industry. They present a major opportunity to suppliers of information, analysis and operational services who can combine experience with the analytical language that Amazon speaks.
- Understand and adapt to the retail and supply chain revolution: Amazon is the embodiment of the digital revolution, which will be turning supply chains upside down in many ways. Their example makes certain that we in the truckload industry will have to make profound changes in order to survive as the supply chains we serve change. The difficulty is that, in the short run, more of the same, our traditional approaches, will suffice for a while longer – lots of people are making great money in 2018 with their usual customers – nor are the eventual answers visible yet. But, sometime in the 2020s the new forces will come to bear and rush by the unprepared. It is time for us to give this matter hard thought NOW! That will enable the aware to recognize the winning procedures when they emerge. Whether or not Amazon emerges as the winner of the next phase, they are the best example we have now, the best organization to study. Following them closely should help stimulate our understanding of this process, again, so that we recognize what must change when the answers become visible.