Completely agree. I have never understood how you can sell something you don't have eg more seats than available.
Simple. In a world of refundable and changeable airfares it makes some sense. Is it a great thing to do? No, but it happens on all airlines bar those that explicitly do not do it (very rare) and some have argued that if one does not overbook it can severely impact on profitability.
Let's have a hypothetical example of a flight from say MEL to SYD and your plane has 100 seats.
In theory you have 100 seats to sell and so you work to sell them. If you sell them all that's great your plane is full and hopefully revenue management have done their job and you've managed to max out more or less the yield from that flight.
BUT....
Let's say 20 of your customers are on flexible tickets (say business flyers) and they can, and do, chop and change - meetings finish early or late and plans change at the last minute sometimes.
Do you send your plane out with only 80 seats filled? And what of the other flights that these flexible fare holders might change to?
In a busy market with many flights like MEL to SYD where someone can show up early and (usually) get on the next one you can be left with some flights with a lot of empty seats.
Now how about you know that on a Friday in June over the past 15 years history tells you that that flight at that time has 10 of those 20 flex tickets change.
So you have the prospect that you could go out with anything from 10-20 seats empty.
All overbooking really says is "history says X people will not show up for this flight for whatever reason - change, cancel, etc"
now, what say your overbooking % for that flight is set to 3%, so you oversell 3 seats (the actual oversell numbers are often under 10% as I understand it) so, you are guessing of the average 10 who don't show for that flight at least 3 won't.. which seems on the surface fairly reasonable.
8 or 9 times out of 10 this all happens transparently and nobody worries about it it all works out and the airline sells 3 extra fares and everyone's happy.
But what if this particular Friday they get 18 of those 20 flex tickets showing up (let's assume all the nonref fares are "locked in" and fly for the purpose of this example)... suddenly you have 103 booked for 100 seats, and 101 hold confirmed bookings at departure time. Uh oh. that's one person too many.. and thus you go into an overbooking situation.
The obvious way to stop overbooking is make all fares nonrefundable and changeable. Thus you book on flight X you fly on flight X .. or if you want to change, you basically pay for another seat. That's not going to wash for the legacy carriers who want to keep their business customers happy and allow them the ability to change on short notice and so on.
Now QF usually seem to have very little overbooking on domestic and/or their load factors are such they can absorb this. US carrriers have been far more agreessive with overbooking over the past few years (how many times have we all been at a gate for an AA or UA flight asking for volunteers? either for standby pax or an oversale situation.. it can seem like just about all of them). This has bitten them in the backside more than once.
Finally I will comment that the infamous Dr Dao issue on UA3411 was *not* a result of overbooking. Indeed The UA3411 situation was more akin with this QF situation (both last minute operational requirements) and NOT overbooking. While on the surface yes the end result was not enough seats for all the pax booked who showed up, in both instances all ticketed and confirmed customers were on board those planes before things happened that changed it; neither situation were results of more passengers with confirmed res showing up for those flights than there were seats for. It may seem subtle but these are very DIFFERENT situations.
I'd also note that reduced load 737 flight to PER was probably quite comfy for those that were able to stay on.. who can claim that for a transcon 737?
