Amazon's Value Chain - Quick Update

Here's a quick update on the analysis published in 2018

In February, 2018, I published an analysis of Amazon’s supply chain using public information available in Amazon’s financial reports. At that time, the latest 10K that was available was for 2016, so the analysis I did was based on their 2016 full-year financials. This year, after they publish their 10K for 2019, I will update the report, along with further refining the assumptions and methodology I used in the 2018 report.

Last week, I started thinking about updating the analysis, so I downloaded Amazon’s 2018 annual report. On the first page, in Jeff Bezos’ letter to shareholders, I discovered something very interesting.

Let me provide some background first.

To construct the 2018 report (based on 2016 financials) I had to devise a methodology based on certain assumptions. One of the foundational things I had to do was to determine the revenue split between Amazon’s first-party sales and sales attributable to their third-party sellers. This information was not disclosed by Amazon but various analysts had speculated on what that that number was. In first-party sales, Amazon purchases and owns the inventory that flows through their distribution centers. This is basically their original business model. In third-party sales, the third-party seller owns the inventory and uses Amazon as its storefront marketplace, along with some also using fulfillment by Amazon (FBA).

Once I could figure out the revenue split between first- and third-party sales, I could then attribute costs to each of those value chains and then figure out profitability. Therefore, determining revenue split was a foundation for the analysis. Since this was based on certain assumptions and a simple methodology, I expected the resultant numbers to be directionally correct at best. The split numbers I arrived at for 2016 revenue flow (value of all merchandise sold) through Amazon’s site were: 46% first-party and 54% third-party. This meant that gross revenue flow through Amazon’s site for 2016 was $204B. I had seen others speculate on this number, but they were all significantly different from my number. This gave me some level of pause, but I decided to stick with the numbers since the methodology, while simple, was at least understandable and repeatable.

Let’s go back to the Bezos letter to shareholders, published in Q1, 2019. In this letter, he starts with sharing the historical revenue split for third-party sales. Here is what those numbers look like:

I immediately zeroed in on the split number for 2016. With some level of astonishment, the 2016 split number is 54%, which is exactly the split I derived using the methodology in the report. While there are other assumptions in the report, this gives me confidence that the numbers and conclusions are directionally correct (and maybe even more than directionally correct). That said, I continue to add the caveat “directionally correct,” since there are other assumptions on top of the foundational split assumption. When I update the report later this year after Amazon publishes its 2019 10K, I will stress test those assumptions, and modify them, if necessary.

Finally, I have to give Amazon kudos for the level of operational transparency in their financial reporting. I read a lot of 10Ks and I find theirs to be excellent. While their report does not provide explicit financials for each of their value chains, it does provide a lot of operational data that is very useful.

Below is one slide from the analysis. Click here or on the slide to download the full (free) report.