This story of "profitable lines" has hung around forever. Every retail business has profitable lines and loss leaders. Many non retail business have them too.
The problem with this line of reasoning is it assumes you can cut off the loss leaders and just run the profitable lines. Generally speaking, you can't. The customers come to Amazon because everything is there. The second they think everything isn't there, they come less frequently, and those "amazingly profitable" categories become not so amazing or profitable. Rotisserie chicken is a loss leader for many supermarkets, and the supermarket makes a lot of money on those snacks near the register. They can't decide to be in the snacks only business tomorrow. Neither can Amazon.
The story about the ambition and the growth and all that is nice - that story is true for Facebook, Google, Amazon Web Services, Microsoft and so on. They all got profitable within a decade and continued with high growth, high profitability operations.
This isn't an accurate view of Amazon's current position in physical ecommerce. They always planned for each unprofitable line to become profitable as it matured.
Amazon's doesn't have many loss leaders in physical ecommerce any more. They used them to build those businesses 10-15 years ago, but they don't need them anymore. The only loss leaders as a category they maintain are digital. And even that is pretty limited to Amazon Video/Studios at this point.
It's a false division - physical v digital. Prime customers get free delivery (physical) + video.
AWS is a genuinely different business, the rest of it is all bundled together. So, they have digital loss leaders. And they may not have many physical loss leaders, but the ongoing argument that they have all these lines which are profitable just doesn't mean anything when customers don't want an online shop for X specific thing. They want an "everything store".
Prime video only exists because of the incremental prime sign-ups, prime renewals, and prime spending money on physical retail. It's all they care about.
The problem with this line of reasoning is it assumes you can cut off the loss leaders and just run the profitable lines. Generally speaking, you can't. The customers come to Amazon because everything is there. The second they think everything isn't there, they come less frequently, and those "amazingly profitable" categories become not so amazing or profitable. Rotisserie chicken is a loss leader for many supermarkets, and the supermarket makes a lot of money on those snacks near the register. They can't decide to be in the snacks only business tomorrow. Neither can Amazon.
The story about the ambition and the growth and all that is nice - that story is true for Facebook, Google, Amazon Web Services, Microsoft and so on. They all got profitable within a decade and continued with high growth, high profitability operations.