Why would AA even want to sell a $200 A->B->C ticket if someone is willing to pay $800 for the same A->B seat? On other forums, people mention the whole "price, not cost" thing, but I'm talking about price.
I'm thinking that limitations in their pricing algo make the seats non-fungible. Maybe they just reserve some % of seats for connecting flights and calculate A->B pricing independently from A->C, figuring it's good enough as long as customers don't exploit it. And if the seats were fungible, the price fluctuations for A->C might be wacky, like staying at $800 until A->B is sold out.
> Carriers argue this practice prevents them from selling seats on the abandoned leg of the journey, leading to lost revenue.
Maybe they could do this with... pricing. Like charge LESS for one leg, instead of more for both legs.
And would they complain if the customer didn't show up for the entire flight?
This seems more like "protect my crazy business model"
But they've already sold the seat on the abandoned leg of the journey, haven't they? What difference would it make if someone sits in that seat or not?
The difference is before all that. They don't want to sell that cheaper ticket in the first place to someone who would otherwise pay for a much more expensive ticket to the first destination.
I thought about it in another comment, and it doesn't seem that crazy. Properly factoring the opportunity cost of the first leg into the connecting flight price would be complicated and might give worse results.
How dare you bypass our price discrimination!