White Papers, Analysis, Presentations

Topical materials and analysis


Amazon Value Chain Analysis

Amazon is the juggernaut retail corporation of our day and there is much interest in how they do things. Just about every retailer in the world has had to come to grips with how to address Amazon's onslaught. Walmart, the world's largest retailer, is bound and determined to not have Amazon do to them what Walmart did to Sears, Kmart, and others. Retailers lament that Amazon runs a high-growth, zero-profit business model, and is rewarded for doing so by the investment community, while these same retailers are held to a different set of rules by the same investment community. Amazon's answer to this is that they just focus on customers, providing ever more innovative ways to serve customers; everything else just flows from that focus. The investment community says that Amazon could turn into an incredible profit engine at any time. But is this true, and if so, how will they do it?

Using only information available in Amazon's 10K reports, and making some basic assumptions and calculations, Worldlocity has assembled a profile of Amazon's business, in the form of a value chain chart, which is shown below. For this analysis, we separated Amazon's business into two segments: 1) their retail products business; and 2) their services business. Their retail products business includes all physical product sales for which they own the inventory, along with various digital products sales, which are downloadable. Their services business includes retail third-party product sales for which they get commissions and fees, and also includes subscription services, Amazon Web Services (AWS), and certain advertising and co-branding products. 

One caveat: this analysis relies on assumptions and cost allocations, for which there is no publicly available information. Having said that, while the numbers lack precision, they are most likely directionally correct. This analysis shows that Amazon, on an operating basis, loses about 2 cents for every dollar of products it sells using its traditional selling model in which it owns the product inventory, while it makes roughly 15 cents for every dollar it sells through its services business (which includes AWS, which has operating margins of 25%). The net result of the two is that Amazon overall makes about 3 cents in operating profit for every 1 dollar it sells overall. (If readers of this analysis disagree or have better data to improve it, please contact Worldlocity). Thus, it's the business innovations that resulted from the original business model that are the profit engines now, and possibly into the future. It's difficult to see how the original business model, given its cost structure, can make money comparable to that of a traditional retailer (without some combination of price increases, cost reductions, and service reductions).

This of course poses significant challenges for traditional retailers, who do not have a marketplace operation or an AWS to subsidize their product-based ecommerce. Thus, they will have to subsidize it out of their brick-and-mortar operations, diluting the margins there. Walmart is rapidly trying to ramp up its third-party marketplace operation, not just to match the choice of Amazon, but presumably to also provide fatter margins. 

Finally, Amazon's supply chain is a cash-generation machine, with a 2016 cash-to-cash cycle of approximately minus 35; this equates to about $13B in cash on a 2016 end-of-year revenue basis (a more precise calculation would time-phase it over the year of 2016). This money float, courtesy of its supply chain, is one of the critical flywheels of the Amazon success story. In a sense, it is not unlike the float on insurance premiums that Warren Buffet has historically used as an investment vehicle. It's a powerful statement for supply chain management.


The net-net is that Amazon manages its business around cash flow, rather than bottom-line profit and profit margins. Its cash flow from operations in 2016 was $16.4B, and its free cash flow was $9.7B. 

Download the full analysis

Retailers and Manufacturers - Can't We Just Get Along?


Over the course of the past thirty years, retailers and manufacturers have spent an enormous amount of time trying to figure out how to work better together. In supply chain management, we like to call this collaboration. An entire sub-segment of the supply chain management software market is devoted to visibility and collaboration. We've had CFAR and CPFR and all sorts of variations, along with various software companies trying to sell snake oil and pixie dust to manufacturers and retailers. None of these technologies and techniques have been able to scale for the collaboration problem between retailers and manufacturers (for a good discussion on why this is, see this old, but good article describing the challenges). Collaboration within enterprises is hard; collaboration between enterprises is doubly hard. 

While all of this was going on, Amazon figured out how to drive $200B+ of gross product volume through its supply chain by connecting customers with suppliers. Why? It's certainly not because they were better at collaboration. Collaboration is important but it's only important if you have a maniacal focus on what Clay Christensen calls the unmet needs of the customer. In Christensen terms, Amazon figured out the latent needs of the consumer. Everyone likes to focus on the Internet and all the related technology aspects, but it's actually much simpler than that. For most people, going to the store is a pain. How do I solve that pain point and do it better than anyone else? Procter & Gamble and Walmart can have the best, most seamless relationship in the world, and the consumer might still find the buying process a major pain. 

Having said all of this, even Amazon will have to figure out how to manage categories of goods in the sense of a traditional retailer. It will need to manage the nuances between house brands and manufacturer name brands. In essence, it will have to learn the collaboration game, albeit with its trademark singular focus - on the customer. 

Worldlocity has worked with both manufacturers and retailers over the course of the past thirty years. At a certain level of abstraction, we are struck by the the similarity of their business processes. Making stuff and selling stuff are completely different operations, with different assets and skill sets, but, at a certain level of abstraction, the business processes are surprisingly similar. The below chart shows key business processes for both retailers an manufacturers across the planning-execution spectrum and how they map from manufacturer to retailer. A future report contrasting these business process will be posted here. 

Cash Flow and the Retail Supply Chain


Retailers lament that Amazon is able to run a near-zero-net-profit business model and is rewarded handsomely for it. They are missing the point. Amazon - by courtesy of its supply chain - is a powerful cash flow engine. The engine it has created goes something like this: increased revenue begets increased cash flow, which begets increased revenue. And, it appears that cash flow is a second derivative of revenue - its cash flow is growing at a rate 50% faster than its revenue. This, of course, has not been good news for other retailers; for example, in 2016, Walmart, which is a strong and healthy company, generated 3.5 times more revenue than Amazon, but only 1.6 times more operating cash flow. At the pace of the past five years, Amazon's operating cash flow will exceed that of Walmart by the end of 2018. A simple ratio like the rate of growth of operating cash flow to the rate of growth of revenue can be a strong indicator of the future. (Of course, many retailers are burdened with debt, which can be a complicating factor in this discussion). 

The below chart shows the cash-to-cash (also called the cash conversion cycle) cycles for thirty of the world's largest retailers (the gross margin versus inventory turns chart for these same retailers is found here). 

Blockchain and Collaboration


Blockchain is a powerful force in the business world right now. It is both a technology and a technique for multi-party transaction management that eliminates the need for middlemen, central bodies, and clearing houses (although as the deployment of blockchain grows, it will become apparent that completely eliminating central bodies and clearing houses might be impossible - see the Wall St. Journal article on lost passwords). Will it provide us with the long-sought multi-enterprise visibility we want? Maybe. Will it help us collaborate better across enterprises? Maybe. The truth is no one knows. Until the investments are made and it is tried for such problems, we don't know what we have on our hands. Either way, something will come of it simply because of the amount of money being thrown at it. The entrepreneur greed factor has set in, so everyone is jumping into the game, from currency exchanges to iced tea companies.  

Collaboration is a squishy term that has been thrown around for quite a while. Software companies have been building so-called collaboration tools for a couple of decades now. B2B hubs were supposed to be the ultimate deliverers of multi-enterprise collaboration. Blockchain turns B2B hubs on their ear - the point is the opposite - there is no hub.


Today we collaborate on plans, purchase orders, shipment notices, capacities, promotions, inventory, and all sorts of other things, with varying degrees of efficiency and effectiveness. We can get visibility for product movements across the supply chain, including hand-off points between different enterprises, carriers, and modes. However, none of this is done very easily or without friction. The general objectives of collaboration are simple - two parties have different goals, objectives, and constraints and they must work together to make globally good decisions, or decisions that are aligned to value chain goals, which presumably are good for the end customer at the ultimate downstream decision point. 

Enterprises - and even organizations within enterprises - represent authority domains that are governed by all sorts of rules and regulations. These rules and regulations hinder the ability to create a network effect similar to that which has become common in the consumer world. For example, an individual in the consumer world can sign up for Facebook simply by filling in a form and clicking a terms and conditions button. In the corporate world, the individual represents the corporation and the corporation is governed by internal and external rules, some of which may be common across corporations and some which may be specific to a given corporation. 

In the B2B marketplace world, we had public marketplaces and private marketplaces. Initially, everyone was talking about public marketplaces until corporate reality set in and companies decided they would rather not be on a public marketplace; thus private variations became common. The same discussion is now happening with blockchain. As companies figure out what they can and cannot share, they will have to figure out what is public and what is private and for what and for whom information is public versus private. 

And, with blockchain, there is the collaboration before the collaboration. Before companies can collaborate on business, they must collaborate on technology standards. Or, there could be a standard that emerges by way of its market strength (which is the historical path; think TCP/IP). Either way, this is an exciting technology (and set of techniques), but there is long way to go if and when it is to achieve critical mass as a value driving enterprise tool.