I have a small number of friends who are executives who negotiate for big contracts. They've always been able to beat analyst expectations when negotiating in high volumes.
I'e seen some ridiculous cost estimates of the Kindle fire that use the cost when buying chips and other parts in thousand unit quantities. At those quantities, one calls a clerk and has a fixed price. For the Kindle fire, it is an estimated 4 million units in 2011. I'm sure for large minimum order quantities and a six+ month production run, Amazon is paying less. Far less.
The trick is:
1. Buy vendors (e.g., Amazon bought their touchscreen chip vendor, Apple with their CPUs)
2. Bid three or more vendors against each other (Amazon did this with LCDs and CPUs bidding Qualcomm, Nvidia, and the winner TI against each other for a huge order).
3. Do not rush the bidding (no evidence of this on Amazon's part). The vendors need to go to their sub-vendors to find unit price reductions for volume guarantees. For example, in chip testing.
4. Listen to vendor suggestions for alternate design ideas. I'm certain Quandra removed cost from the blackberry playbook and was willing to 'give' the design to Amazon in return for volume guarantees. I doubt this was Amazon's going in position.
Analysis where the Kindle fire costs a little more that $199 to build.
More believable $150 BOM for Kindle fire.
The above analysis breaks down the costs . My only real complaint is Amazon wouldn't pay $19 per processor (CPU). Rumors are Amazon is paying TI far closer to $10. That alone would make the Kindle Fire a profit maker. I've also heard Amazon negotiated at just the right time for the screens and is paying less for them. True or not, Amazon is the first Apple competitor to be able to jump in and guarantee multi-million unit volumes. Amazon probably paid upfront the setup costs, so the more they sell, the more profit. :)
The above doesn't include all the costs of building/selling a Kindle fire. It is probably $30 to $40 short of the full cost. So do not think Amazon is making $49 profit per Kindle Fire. It is probably in the $1 to $10 range in 2011 (better in 2012 as post holidays chip and other costs drop).
My best estimate is the Kindle Fire goes out the door at a tiny profit. I just cannot believe the high component costs quoted for Amazon when they are bidding on multi-million unit quantities. It isn't that Amazon isn't spending and losing money somewhere (e.g., India fullfilment centers). I just do not see the huge 'loss leaders' for the new Kindles some articles are claiming. Not in the volumes Amazon in buying in.