Cars used to have a useful lifespan of decades. Compare a ’38 Ford to a ’58 and there are some marked improvements, but the older model would have been quite at home on the roads of 1958. That longevity applies less and less to the cars of today.
The very rapid rise of self-driving cars is probably the best example, especially as we head into a transitional period between “fully manual” to “semi-autonomous” to “fully autonomous”. Cars are a big investment, so it isn’t very encouraging to buy a new one knowing that it will be technologically obsolete long before the end of its mechanical life – like pointlessly replacing a smartphone every year, but at 100x greater cost.
Ironically, this may be the biggest nail in the coffin of personal car ownership. If a car can drive itself, and they are upgraded too quickly for amortization to really work, why not sign up for a “car-as-a-service” program, a driverless Uber? That offloads purchasing to a company that will probably churn through cars on an annual basis anyway, and negates any need for finding a car with desired equipment, driving dynamics, etc. I’m guessing it would also reduce the total amount of cars on the road, since there would be no downtime (e.g. cars sitting empty in parking lots for hours).
Car manufacturers would then be reduced to supplying these fleets rather than the more glamorous sales to private consumers. Their advertising would change from wind-in-your-hair imagery to B2B comparison charts, and may end up looking like the market for passenger airplanes. There would still be a niche market for private ownership – probably the luxury and/or sports models – but it would definitely be far smaller due to the relative cost compared to an on-demand service.