Originally Posted by
channa
Because CO lacks the infrastructure to rebook and process volunteers efficiently. Having 10 VDBs at an outstation with light staffing could tie up that outstation's staff tremendously because the business process and systems are cumbersome.
Though oft repeated here as a factor in just about everything, I highly doubt this comes into the scope of things IM/RM considers when deciding to oversell another seat or not.
Originally Posted by
channa
I understand that, and that's part of it. But the other part that's outside the company's control is the hub conditions and potential for delays/misconnects/other issues.
Even those sorts of uncontrollable irregularities have some patterns that are modeled and taken into account.
Originally Posted by
channa
SFO in winter is a perfect example. The airport is often delayed in certain hours. Does one account for that and assume there will be misconnects and oversell accordingly? Or do they use a perfect world scenario and then oversell only slightly?
Yes, they do.
Originally Posted by
channa
It's this sort of unpredictability that UA tended to be aggressive in. Oversell the plane by 10 or so, there's likely going to be fog/weather/inbound RJ or prop cancels/other delays. When they were right, they win because they sold extra tickets. When they were wrong, well, they're VDBing a bunch of people.
If you're building your systems based on the expectation of failures in the process then I'd suggest you're building it wrong. Account for them and be able to handle them, sure. But basing the underlying financial models on them, not so much.