I had these in a spreadsheet a couple of years ago when I was studying retailers. I think I got the data from Yahoo Finance like this: http://finance.yahoo.com/

I had these in a spreadsheet a couple of years ago when I was studying retailers. I think I got the data from Yahoo Finance like this: http://finance.yahoo.com/q/is?s=AMZN+Income+Statement&annual http://finance.yahoo.com/q/is?s=AMZN+Income+Statement&an... which breaks out Cost of Revenue and SG&A. I was quoting AMZN for 2010, which had been pretty stable -- in 2011 it looks like that it kicked up spending quite a bit. You made very good points w.r.t. Sam's Club for Walmart or Kindle for Amazon. One needs to adjust the data in order to be precise. The effect though is probably second order and shouldn't change the picture that much. At least it seems consistent with my own anecdotal experience. It is also that possible Amazon's expense is exaggerated by accelerated depreciation. In the expansion phase if your facility is not fully utilized you get dinged again on appreciation. WalMart owns a lot of real estate and has capitalized lease as well I think so some of the operational cost may show up as interest expenses. To get a true picture one needs to account for all those as well. The beauty of Amazon's model is that it uses little to no capital (all operations can be funded by negative working capital and other liabilities incidental to operations), so it can expand unconstrained by capital needs, as long as it can find new customers.