First VDB on CO
 

I am astonished to see that
(1) their VDB offer starts at $250 USD
(2) VDB offers are negotiated with passengers individually, and even when you explicitly ask, they don't match the offer they agreed with another passenger. I mean, there are a whole plane of passengers to load, and the agents are taking the time to fight for $50-$100 zoo dollars. Do they get a bonus for giving the passenger less voucher for VDB?

Is this yet another characteristics of Continental, beside the landing skills of their pilots?

I don't know about the landing skills you're talking about, but for VDB compensation, they definitely require their GAs to play auctioneer and try to pay out as little as possible. It's just one of the CO characteristics of focusing on minutiae.

In some cases, they've started lower than $250.

What's interesting is that CO's on-time performance is pretty bad (they were last or second to last for each of the last 3 months, excluding Express carriers). You would think they'd try to focus on boarding efficiency, but instead they're focused on saving a bit of funny money here or there.

They also have one of the lowest denied boarding rates in the US industry so the number of flights impacted by this is dramatically lowered as a result. On-time performance is skewed by having a hub in EWR more than anything else. It would be hard to argue that GAs negotiating with pax on VDB payouts has a material impact on on-time performance.

Which is countered by them consistently having one of the highest involuntarily denied boarding (IDB) rates in the industry.

It's possible that their inability to solicit enough volunteers results in them being forced not to oversell by as much. That in turn leaves real money on the table.

While you look at it from an impact perspective, I see it as a lost opportunity for the airline. I think playing the auction game turns people off the process, so fewer people participate.


Many airlines have hubs in congested airports. UA has ORD and SFO; DL and AA have JFK; DL has ATL; CO has EWR; DL (formerly US) has a focus at LGA; etc. While EWR makes an impact to CO's stats, the other airlines are also burdened as well by their congested hubs.

On my last CO VDB they most certainly took a delay because of the oversale. I talked to them beforehand, we agreed on an amount, but protection was the issue (maybe it was too difficult to do in SHARES, I don't know). I wanted a flight they didn't want to do. So I passed and got on the plane.

Once fully boarded, they come on the plane begging for vols. I ask this GA about my protection. They radio someone and they say it's okay. Off I come, and they push late.

I was quite specific about my language - they have one of the lowest denied boarding rates (VDB and IDB combined) in the industry. IDBs are so rare they barely impact the numbers - there was some recent data posted on this.

I got VDB in Omaha on 12/27. I could tell that the flight was badly oversold; e-Checkin offered $200 for a later flight. I declined. At the gate, they were offering $400, which I gladly took.

I was offered $75 for BTR-IAH. :eek:

Needless to say, I declined.

Yes, I posted it. And the numbers were statistically insignificant. Based on true statistical norms, not "CO has x0% more", which is based on an amazingly small sample.

Another thought - OP - nice cheap shot concerning the pilots and landing skills.

There you are playing your numbers games again.

The vast majority of people fly as ticketed. The number of VDBs and IDBs are both very small. That's what we're talking about. Just because the discussion brings up an aspect that doesn't make CO look good doesn't make the number trivial.

There's a higher chance of getting IDBed on CO than there is on other carriers, despite CO's lower DB numbers.

In fact, when presented with an oversold flight, you're more likely on CO to get an IDB than on the other carriers because of this.

When you look at all the DBs of the major carriers and see what percent are VDB vs. IDB, it looks like CO may have a problem getting enough VDBs.

Now that doesn't say why, but it could be the routes it flies, the clientele it has, or it could be that it not generous enough with compensation and/or makes the process cumbersome

If you're referring to that other thread, a statistician posted that they were indeed statistically significant.

We're not talking about large numbers, for sure, but they were statistically significant.

I'd ask the statistician the p-value, but I'll drop it. No need to resurrect that thread.

:D No, it is a numbers game. You can't just cry "you're playing a numbers game" that as a defence whenever you don't like that a point you are making is massively diminished by how rarely it occurs. I'm sure those IDBs are very unpleasant to every single person they happen to. And the number of times they occur is tiny, for all airlines - not just CO.

Yeah, that was me.

P-value for "PMCO IDBed PAX at a higher rate than PMCO in 2Q 2011" is about 2.8 x 1E-66. (Sorry -- I can't figure out how to get superscripts in FT's version of BB code. Anybody have a workaround for that?)

The P-value (for a given conclusion) is basically the probability of having gotten a false positive. More precisely, if I claim "X" with a P-value of p, then I'm asserting that I have data that (1) support my claim of "X", and (2) would occur with probability less than p if "X" was in fact false.

So, the simple interpretation is: if PMCO's overbooking/bumping process is actually less IDB-prone than PMUA's, then in fact we could have gotten the data posted by aacharya in that old thread by random chance. But the probability of getting those data -- or any equally extreme or more extreme data -- is less than 1 in 100,000,000,000,000,000,000,000,000,000,000,000,00 0,000,000,000,000,000,000,000,000,000, if in fact PMCO's overbooking/bumping process is less IDB-prone.

Which leads me to conclude that, indeed, PMCO is probably more IDB-prone. :) In other words, the data are statistically significant.

That absolutely does not mean that it's operationally significant. UA IBDed about 1 in 10,000 pax, while CO IDBed about 1.7 in 10,000 pax. For context, if ua1flyer was at risk of IDB (I'm sure he's not), over his whole career we'd expect him to have around 1 IDB at that rate. So it's not a big risk for any given pax.

It's also probably not a major issue for the airline's bottom line, but some of the related VDB issues mentioned above (profit from overselling at high fares vs. loss from liberal voucher policies) may be.

That's my point. It's a small number for all airlines. The relevance is how small.

Just because it's a small number doesn't mean it's insignificant or should not matter.

Remember the airlines are moving millions of people each month. Even a small percentage has an impact. That's why you look at rates.

When an airline consistently has one of the highest IDB rates in the industry, that is a very likely indicator that something could be improved. And improvement should be a goal for any business.

Good assessment. Thanks for the p-value calculations.

You've created another argument for argument's sake. You posted above that they should be focusing on trying to improve their on-time performance instead of trying to reduce the payouts. None of what you've posted here changes the fact that these events happen so infrequently that they just aren't going to impact that number. And that's being generous - I highly doubt that every departure where there is an IDB or VDB situation is delayed because of the process in place.

I've personally a VDB situation on CO where 3 volunteers were identified at the start of the process, payout agreed, and 1 of them was asked to board at the very end due to a no-show. Flight left on time (actually early) and the vouchers were processed after the aircraft pushed back.

That's my personal experience - but I'm not drawing a conclusion from that which says that this is the norm. You are taking your one personal experience and extrapolating that to being a reason behind CO's on-time performance. EWR is a much more significant reason IMO.

At the risk of belaboring the obvious, 1700 (# of IDBs for CO in one quarter of 2011) is a large number. But 1 in 6000 (fraction of pax IDBed in that quarter) is a small percentage.

My point is that the only way to have a reasonable discussion about this is to be careful with language. If we want to have a rhetoric-laden free-for-all, then we can be sloppy with language. :)

While I tend to agree, it should be noted that exceptionalism is not necessarily a sign of ineffiency. WN had the lowest fuel costs in the industry for years, and that was obviously a good thing. The real question is whether a moderately higher-than-normal IDB rate is indicative of a good business model or a bad one. Arguments have been made for why higher IDB could contribute to CO's bottom line (the cost of 1700 IDB compensations may well be dwarfed by the savings in paying $250 per VDB instead of UA's flat $400).

It's a reasonable surmise that new UA management will crunch the numbers and set the policy to optimize their bottom line -- with the benefit of data from both pre-merger systems. Which suggests that whatever UA does in 2013 will indicate the more efficient system (in their humble opinion).

Not really. You don't seem willing to entertain another opinion on the matter without attempting to slam that person.


Correct. When an airline like CO is plagued with last, or near-last on-time performance on a regular basis, I think their focus should be on fixing that rather than $50 of funny money here or there.


I did not say this was a cause of their delays. If anything it's a possibility, but I said it was a matter of priorities for the airline, not a cause.


And neither am I. I just gave an anecdote of my experience, much like you did. Whether their VDB lowballing makes a material impact to their on-time performance, we don't know.

As for EWR, CO outperforms most other airlines at EWR, and other airlines have their own hub challenges, so that seems like a red herring IMO.


Good point. Though since this is FlyerTalk, not AirlineExecutiveTalk, I'm looking at this from the flyer's or customer's perspective, and in general, I think most customers would agree that IDBs are a bad thing. Smisek and CO can do all the number crunching and set their policies as they see fit.

New Policy???
 

This is new.
Previously, all CO volunteers received the same (highest) comp.
However, you had to be at the gate to claim the comp, if it was raised.

I was once sent from a departure gate to a CO Customer Service Center to have them process the VDB voucher/re-route.
While there I heard the GA raise the VDB comp.
I noted it to the service rep and my voucher amount was increased.

I always confirm that if the VDB offer increases I will receive the higher amount.

IBD's themselves may be a bad thing, but if CO's load management leads to more people taking the flight they booked (lower overall denied boarding %), then their IBD strategy may actually be customer friendly.

It depends how you define customer friendly.

If you figure that a VDB is a volunteer and is willfully taking the alternate flight for the compensation, and that is not a customer negative, then the IDB number is the only number to focus on.

Now if you're saying that a VDB experience is inherently a customer negative, then the total DB number is the one to focus on.

I would guess that most VDBs are satisfied with their outcome, otherwise they wouldn't volunteer in the first place.

Agreed. On the other hand... a lot of discussions around here lately have diverged from "This is what I, as a flyer, wish COUA would do" into "Well, it's a for-profit business, what do you expect?" and then into "But then they should do it this way."

Probably some of that discussion is useful and productive, in two senses. First, we all occasionally need to be reminded that UA is a for-profit business, and we will not get ponies on board (no matter how much we want them). Second, FT is a bit of a collective bargaining organization. On occasion, by collectively voting with our purchases (or threatening to do so), we can influence airline policies, as long as the cost-benefit calculation works out in our favor. UA Insider is here because it's good business for UA. And UA keeps me more or less happy because that, too, is good business for UA. So it behooves us to have an accurate perception of how those policy changes that we want (or fear) affect the bottom line... because that determines whether we might get them.

Personally, I'm mostly indifferent to IDB. It's unlikely to happen to any pax more than once in a lifetime (based on published stats), and even less likely to happen to me (as an elite).

I care much more about VDB policy, which is linked to IDB policy. I like PMUA's VDB policy. I think overbooking is inherently customer-friendly (although frequent standby flyers will disagree, because overbooking makes standby harder). I wish every flight was overbooked, and I could choose whether to earn $400 for hanging out in the airport. I want to encourage them to continue it, and one way to do that is to emphasize that it cements my loyalty... because my loyalty is a [tiny] part of the bottom line.

EDIT: I wonder what fraction of IDBs actually end up happy with the outcome? IDB compensation is supposed to be pretty good, especially if it results in serious inconvenience (>4 hour delay, IIRC).

Sorry, what does 'protection' mean in this context? I assume it's not a bodyguard. :)

There have been repeated assertions that the VDB/IDB ratio somehow changes the on-time performance of the carrier. I'd love an explanation of how that happens. If they don't get the VDBs they still close the door and push the plane once it is boarded. They're not sitting around in the terminal playing games rather than closing the flight out if people don't respond to the solicitations.

It might be that more people are IDB'd because the offers are so low that no one will take them (I don't think that's the case but I'll accept it for this discussion). That doesn't change that NN people aren't getting on the plane anyways and the door has to close.

And if it not related then why bring it up in the discussion? Talk about derailing a cogent and reasonable discussion with off-topic rhetoric...

channa wanted a guarantee that the p.o.-ed CO agent wouldn't beat the cr*p out of him (sorry, gender assumption) once they figured out his identity. :D:D:D

Seriously, "protected" on a later flight means you have a confirmed seat, rather than just being on the standby list. A reasonable thing for an volunteer (especially an elite) to insist upon.

I'm not sure who made the assertion that CO's poor IDB rates impact their poor on-time performance. It's a correlation, no doubt, but I don't think anyone said that they're causal.

I did say earlier that I think their priorities are messed up if they're worried about nickel-and-diming VDBs while having one of the industry's poorest on-time records.

:D

The CO agent was very nice -- most are (I won't get pounced on for making that generalization because it doesn't make CO look bad). ;)

Anyhow, she just didn't want to do what I wanted. She offered other solutions which were later (and I wasn't interested in), so we didn't come to an agreement, and we parted ways. The conversation was pleasant the whole time.

I burned some Delta miles @ Christmas, and at kiosk check-in @ IAH, a message popped up they were looking for volunteers. I accepted. Then they asked what amount of $ I will willing to take for the bump. The message very clearly stated that volunteers would be taken in order of the lowest amount one was willing to accept.

That's kind of hard to bid due to the factors of not knowing when the next available flight would be (would it be 2 hours, would it be overnight etc), could I keep my FC seat, etc.

For all info - I put in a bid of $250 for the IAH-ATL leg and it apparently was not accepted as another guy's name was called and I heard the GA talking to him about his alternate flight.

Hopefully UACO doesn't start going that route for VDB.

I've had three CO VDB opportunities, all of which I have declined. In March and September, I was on SEA-EWR. They started with an offer of $400. In May, IAH-LAS, they offered $250.

While UA's VDB process was generally good to me last year (3x $400 and 1x $600 [I declined the hotel and cab voucher and it was the last p.s. JFK-LAX flight of the night oversold by order 20 pax]), I wish the GAs could be a tad more generous on occasion. Last Monday, I declined $400 for the last IAD-LGA flight as I was going to have to put myself on Amtrak to get in by the early morning and pay for my own cab (I stupidly forgot to ask about the US Shuttle out of DCA in the morning). Even had I declined the hotel, the agent wasn't willing/able to offer a cab voucher into the city. When I was still at the gate, they had exactly the right number of volunteers, so my guess is that someone had to get IDB'd as a result.

Well - the number said it. CO sucks in VDB (mostly because of the offer I guess).

I think that one of the reasons why SMI/J wants to keep PMUA's IM.

(For real - based on the SMI/J we know, he would have killed that $100 off from the standard offer by now.)

Beside - CO really has issues when dealing with DBs.

Incorrect. IDB's are the ones the government cares about. If an airline is able to coax you away from your flight with your consent, the regulators don't give a damn...they take the data, but it will never impact regulation. A volunteer is someone who got an acceptable vvalue proposition. The IDB's are the ones that get weighted heavily by regulators as they pooint to a failure where a passenger is not given either what he/she bought, or what they were willing to accept. I will tell you now, most carriers planners don't see VDBs as an issue. It allows them to give out $$ that must be used for a future purchase in exchange for real cash money into the revenue stream by overbooking. An IDB is real cash out of pocket issued by the airline. It (the IDB process) has many regulations that go along with it that VDBs don't have.

Now the operations folk aren't the same bean counters as the revenue folk...they see both as an issue as it both VDBs as well as IDBs put a stress on the airines operational resources (delays=loss of use of gate/plane/crew and CS agents are unable to work the nect flight when processing both the VDBs and IDBs.)

Better software, better economic model (significantly more complex though.) UA's system isn't designed to optimize revenue on a flight by flight basis, but to optimize revenue on a systemic basis. Maxing out the prerated flight revenue of 1 plane may not be the best move for the revenue of the system as a whole. Selling a full fare seat (the last seat on the plane) for ORD-STL at $600 sounds great, but if in so doing in, you miss out on a FRA-ORD-STL booking worth $2k, you have hurt the system revenue at the cost of segment revenue. UA's system=system revenue based, not leg revenue based. Leg based revenue systems is so 1970s/1980s, while network based systems are late 90s/today. Prior to UA buying "Deep Blue" from IBM in the late 90's there were no computers capable of doing large scale network revenue modeling in a real world aplication time frame. Thanks to Moore's law, it is a possibility now that didn't exist 15 years ago. If you've ever been to UA's old HQ (not the new downtown place,) I BELIEVE (not sure) that the pond there isn't so the geese can play, but was used for cooling, as there is a huge bank of compressors and fans with next to the big pond with coils running to the pond. I'm betting the mainframes were part of the reason for the heat exchange, but that is only conjecture on my part.

That's cool. :) Thanks for the informed perspective, fastair!

I don't disagree with any of that, but it doesn't make my point "incorrect" - the reality is that IDBs on any US airline are few and far between - we are really dealing in the margins of insignificance.

The new "standard" offer as being defined in the training going out to GAs is to offer based on the lowest fare in the market, rounded up to the nearest $50 and no lower than $100 nor higher than $400.

I've also heard that they are possibly going to be less inclined to accept refundable fares for VDBs or to accept repeat customers, though I didn't see that in the documentation I was able to sneak a peek at.
They care in the sense that they report on them, but I've never seen the government take action against an airline for IDBs, other than when they neglect to properly inform customers of their rights in such cases.

Where's the correlation? Can you show that IDB rate is related in any way to on-time performance?

Unless the GAs "nickel-and-diming" customers actually has any impact on the dispatch times of flights - something that has not been shown in any way - then this is a straight red herring. It is like saying that the company has one of the industry's poorest on-time records because they don't put olives on the salads.

Show something resembling a connection, other than just a personal hunch, and it might be an interesting topic to analyze further. As it stands right now it is just a joke.

Here's the actual data, IDBs per 10,000 customers and on-time rates for the past 4 years. There's no correlation at all that I can find. Perhaps someone else sees some.


If the correlation premise is correct then we should see carriers with lowed IDB rates have higher on-time numbers, right? Or that YoY the same carrier should have a higher on-time rate when they have lower IDB rates, right?

What the data does show is that AS, with their very high on-time numbers, also has middle to bad IDB numbers, while B6, with virtually zero IDBs has pretty poor on-time numbers. F9 cut their IDB rate by more than half YoY from 2010 to 2011 but also dropped a couple points in on-time performance. And UA's major carrier-leading on-time performance in 2010 that they bragged about so much occurred commensurate with an IDB rate 20% higher than 2011's and nearly the same as 2009's, but the on-time performance those years was markedly worse. WN's best on-time performance in the past 4 years came with their highest IDB rates, too.

There's no correlation. There's no causation. There's just FUD being spewed.

I always wonder - has anyone use a full Y to earn VDB whatever they want?

Since DL is doing some of these already, I won't be surprise if these happen.

But rejecting a previous VDBed customer from another VDB will resolve oversale problem?

Actually the numbers above neither prove or disprove any correlation related to the GA software being used. The only way those numbers would accurately show correlation is if the carriers in question had the same VDB procedures and training, the same overselling patterns, the same boarding procedures, etc., etc., etc., while having the GA software as the primary difference. All of those other factors vary greatly from one carrier to another in the list so you can't use the numbers to isolate on GA software as the deciding factor one way or the other.

Nothing in the numbers either supports or refutes anyone's claims about the effect of GA software used during VDB situations on delayed departures.

Yes. Some have even been banned from flying on carriers due to abuse of said behavior.

Who said anything about it relating to the software? I'm simply trying to establish whether there is a correlation between IDB rates and on-time percentages, akin to this line:
Regardless of how the agents process it, the data does not seem to support that there is any relationship, correlation or causation between on-time performance and IDB rates. If there is, I'd love to see an explanation of it. But the data doesn't seem to suggest any correlation that I can track.

In the interest of keeping this discussion on ArmchairAirlineExecutiveTalk (TM) somewhat coherent, I'd like to point out that the word "software" does not appear anywhere else in this thread. In other words, nobody has (so far) suggested the existence of "any correlation related to the GA software being used". Inasmuch as correlation has been discussed at all, it's purely the [putative] correlation between IDB rates and on-time rates. Not whether such correlation might be due to GA software.

So, in this thread, I don't believe there were any "claims about the effect of GA software used during VDB situations on delayed departures," even implicitly. It all started (this time) with channa's suggestion that CO should improve their on-time performance instead of worrying about VDB compensation. That implies a policy causation (late departures because managers have spent too much time worrying about VDB policy, instead of optimizing on-time departure), rather than a software causation.

That said, I'm provisionally convinced by sbm12's data suggesting no observable correlation at all. (To remove "provisional" I'd have to scatterplot the data points and do a regression, which doesn't seem especially fun or worth the effort).