Power - per kW cost would not have fallen materially, but efficiency per instance would have improved
Land - flat fixed cost would not have fallen, but sunk cost on existing facilities no marginal cost only relevant to geographic expansion. Also, AWS would own so would be working from increasingly small amortized base
Maintenance - might improve marginally with scale and experience as improve leverage on physical infrastructure
Staffing - would expect headcount per instance to fall in the period and automation to simplify some roles, offset by rising wage
R&D - You are 100% correct this would most definitely be increasing in the period and illuminates an interesting nuance. I am probably perverse in thinking of the underlying 'core' service (compute, storage, etc) margins as funding R&D to compete by creating the suite of services on top
Fascinating topic - yes AWS pricing falls slower than Moore's Law, this reflects the degree of market power shared by few large public cloud providers given the scale needed to compete at the top level. The profit maximizing strategy for each is enjoy rising margins while costs fall and reset prices to a lower level as a pack once costs have fallen so low that it is most profitable to lower the price to serve a larger market (google 'oligopoly' or 'kinked demand' for further clarity).
There was actually a slide addressing this very question used in a Urs Hölzle keynote at Google Cloud Live 3/25/14. It was titled 'but prices are not falling fast enough' and showed 2006-2014 cloud prices falling 6-8% vs. 20-30% improvement in hardware pricing. I included a screenshot in my deep-ish dive I'm developing on the economics of cloud market pricing http://www.stackalpha.com/blog/2015/2/25/cloud-price-wars-th...
Agreed and especially like your points around expert markets about the executives' radar.
Just a to build a bit on the pricing point: Outside of perfect competition, price is not determined by the market. It is determined by executive or committee or some agent of a company subject to the market, specifically 1) observed and expected competitors' prices and 2) the company's estimate of the demand curve composed of all possible customers & workloads.
Great points, but I would (and have - news.ycombinator.com/item?id=911252) argue that prices are not cheap, at least relative to costs. They are spending to develop some great services over top, but the perception that they are charging anywhere close to their marginal costs is purely successful PR.
Also, subtle point but I would add that any firing would be for moves perceived as customer-hostile. Just looking into the EC2 Reserved Instances and the supposed secondary marketplace I see some less-than-consumer-friendly practices that seem to be just fine so long as nobody takes note.
On Reserved Instances: this makes RI purchase decision sound binary based whether or not 100% usage. Even accepting the gross oversimplification of binary based on usage, the discount is enough that the headline breakeven utilization can start at ~60% for 1 year and ~40% for 3 year RIs and then drift up from there as you start realizing the assumptions hidden in the discount, e.g. 0% time value of money, no future price cuts, etc.
Also very small point, but new 24 month RIs do not exist. Only place to find an odd 2 year would be on the RI Marketplace, that is if AWS were actually motivated to have it function...
Power - per kW cost would not have fallen materially, but efficiency per instance would have improved
Land - flat fixed cost would not have fallen, but sunk cost on existing facilities no marginal cost only relevant to geographic expansion. Also, AWS would own so would be working from increasingly small amortized base
Maintenance - might improve marginally with scale and experience as improve leverage on physical infrastructure
Staffing - would expect headcount per instance to fall in the period and automation to simplify some roles, offset by rising wage
R&D - You are 100% correct this would most definitely be increasing in the period and illuminates an interesting nuance. I am probably perverse in thinking of the underlying 'core' service (compute, storage, etc) margins as funding R&D to compete by creating the suite of services on top