[btilly]

It is very hard to get downstream incentives to match the incentives on making the thing. Here are some made up numbers. Suppose that a car costs $25k new. Then manufacture might cost $15k, amortized R&D might be $5k, and profit $5k. After 3 years the car might be worth $15k and the projected lifetime value of being able to try to resell that used, fix, etc might be $5k. Using these made up numbers, profits and lifetime value are about balanced. Unfortunately for the car owner, resource allocation decisions are entirely based on marginal costs. R&D is not a marginal cost - it is a sunk one. Thus selling a new car adds 2x as much to the bottom line as the lifetime value of maintaining it longer. Guess what gets discarded?

Remember, this is the industry which inspired the term "built-in obsolescence". It is also the industry which found it profitable to buy regional transit companies, and run them to the ground to encourage people to buy new cars instead. There is a certain level of distrust backed by history here.

[milfot]

Yeah, all good points. I don't think this should stop tesla trying, or rather, it would be a shame if they were prevented from trying.

i believe it is a matter of restructuring both risks and incentives to make the downstream match the upstream, as you say.. for example, with the saas approach, for say 5k a year, your 5k profit over 5 years per user (add 5k for R&D and 5k for a facelift) turns into 20k profit over 10 years.. and the user gets a guaranteed service - the user and manuf. interests align..

The bigger issue I think, in terms of turning around the current approach, is actually a cultural one in which a car is a personal 'fetish' object rather than the cumulative efforts of a supply chain bundled into a personal transport service.