United Miles are not that bad a value
Sep 12, 2014, 1:09 pm
#1
Original Poster
Join Date: Aug 2004
Location: Round Lake, IL
Programs: UA GS, Marriott Titanium
Posts: 1,064
United Miles are not that bad a value
So I have been on this form for at least a decade now and keep seeing the "devaluation of miles" posts here and there. While I agree miles are being somewhat devalues, now that I have a Sabre data feed I thought I would play around with numbers and see how bad the devaluation really is.
Method of testing:
I do not want to give away too much of my algorithm here (as I am planing to incorporate it as a new page on my site - http://faremechanic.com - to show on a weekly basis what the current value of various miles are), but basically I ran a large set of near in trips within the major 4 legacy US carriers (UA, AA, DL, US - note, this is an alpha test not a prod sampling). I used near in trips because they would have the highest prices, least chance of saver awards, and would probably serve as the most reliable way of testing dollars per mile. After running a sampling, I would then apply how many miles it would cost for each program for an unrestricted plane ticked (standard award). From there I can up with the following mile rates:
AA: $0.012087
DL: $0.011137
UA: $0.016614
US: $0.011112
From there, UA looks like you actually get the most "value" for your mile. Now I will state again, this is an alpha test using only domestic US close in tickets. Applying to a much larger set would give far more accurate results, but the numbers were all within the accepted range (about $0.01 per mile).
So, a few questions for the crowd:
1. Do you see any major flaws in my theory?
2. If I ran this same concept with a larger sample and over the top 50 airlines world wide, would you see value in it on a weekly basis?
3. I considered adding upgrades and C/F tickets in the future, but decided upgrades are too hard to add (not a guarantee that you can get an upgrade - whereas you are guaranteed a standard award in almost all cases) and probably will add a sampling of F - thoughts?
4. Do these numbers fall in line with what you would expect?
My theory to why UA is better than the rest is?
1. 50,000 flat miles for standard US regardless of peak/non-peak (compared to 60,000 at the other 3 with peak)
2. Several close in tickets were more expensive than the other competitors
Love to hear any thoughts and whether this is worth pursuing on a regular basis for a large group.
Gary
(note - this is not a post to advertise my site in any way. Its not even a live site - its still just in alpha and not ready for the real world.)
Method of testing:
I do not want to give away too much of my algorithm here (as I am planing to incorporate it as a new page on my site - http://faremechanic.com - to show on a weekly basis what the current value of various miles are), but basically I ran a large set of near in trips within the major 4 legacy US carriers (UA, AA, DL, US - note, this is an alpha test not a prod sampling). I used near in trips because they would have the highest prices, least chance of saver awards, and would probably serve as the most reliable way of testing dollars per mile. After running a sampling, I would then apply how many miles it would cost for each program for an unrestricted plane ticked (standard award). From there I can up with the following mile rates:
AA: $0.012087
DL: $0.011137
UA: $0.016614
US: $0.011112
From there, UA looks like you actually get the most "value" for your mile. Now I will state again, this is an alpha test using only domestic US close in tickets. Applying to a much larger set would give far more accurate results, but the numbers were all within the accepted range (about $0.01 per mile).
So, a few questions for the crowd:
1. Do you see any major flaws in my theory?
2. If I ran this same concept with a larger sample and over the top 50 airlines world wide, would you see value in it on a weekly basis?
3. I considered adding upgrades and C/F tickets in the future, but decided upgrades are too hard to add (not a guarantee that you can get an upgrade - whereas you are guaranteed a standard award in almost all cases) and probably will add a sampling of F - thoughts?
4. Do these numbers fall in line with what you would expect?
My theory to why UA is better than the rest is?
1. 50,000 flat miles for standard US regardless of peak/non-peak (compared to 60,000 at the other 3 with peak)
2. Several close in tickets were more expensive than the other competitors
Love to hear any thoughts and whether this is worth pursuing on a regular basis for a large group.
Gary
(note - this is not a post to advertise my site in any way. Its not even a live site - its still just in alpha and not ready for the real world.)
Last edited by gldwebs; Sep 12, 2014 at 1:10 pm Reason: Disclaimer added
Sep 12, 2014, 1:57 pm
#2
Join Date: Oct 2012
Posts: 638
I think it is important to search for possible bias. Are the flights you chose really representative? Are the people who use miles for domestic tickets likely to buy close-in tickets if they don't have the miles? Maybe those people would otherwise purchase N tickets, in which case the value of a corresponding award is not so good. What I'm trying to say is, spend a lot of time playing devil's advocate before you spend a lot more time perfecting your algorithm.
Sep 12, 2014, 2:02 pm
#3
FlyerTalk Evangelist
Join Date: Aug 2002
Location: Bay Area, CA
Programs: UA Plat 2MM; AS MVP Gold 75K
Posts: 35,067
So I have been on this form for at least a decade now and keep seeing the "devaluation of miles" posts here and there. While I agree miles are being somewhat devalues, now that I have a Sabre data feed I thought I would play around with numbers and see how bad the devaluation really is.
Method of testing:
I do not want to give away too much of my algorithm here (as I am planing to incorporate it as a new page on my site - http://faremechanic.com - to show on a weekly basis what the current value of various miles are), but basically I ran a large set of near in trips within the major 4 legacy US carriers (UA, AA, DL, US - note, this is an alpha test not a prod sampling). I used near in trips because they would have the highest prices, least chance of saver awards, and would probably serve as the most reliable way of testing dollars per mile. After running a sampling, I would then apply how many miles it would cost for each program for an unrestricted plane ticked (standard award). From there I can up with the following mile rates:
AA: $0.012087
DL: $0.011137
UA: $0.016614
US: $0.011112
From there, UA looks like you actually get the most "value" for your mile. Now I will state again, this is an alpha test using only domestic US close in tickets. Applying to a much larger set would give far more accurate results, but the numbers were all within the accepted range (about $0.01 per mile).
So, a few questions for the crowd:
1. Do you see any major flaws in my theory?
2. If I ran this same concept with a larger sample and over the top 50 airlines world wide, would you see value in it on a weekly basis?
3. I considered adding upgrades and C/F tickets in the future, but decided upgrades are too hard to add (not a guarantee that you can get an upgrade - whereas you are guaranteed a standard award in almost all cases) and probably will add a sampling of F - thoughts?
4. Do these numbers fall in line with what you would expect?
My theory to why UA is better than the rest is?
1. 50,000 flat miles for standard US regardless of peak/non-peak (compared to 60,000 at the other 3 with peak)
2. Several close in tickets were more expensive than the other competitors
Love to hear any thoughts and whether this is worth pursuing on a regular basis for a large group.
Gary
(note - this is not a post to advertise my site in any way. Its not even a live site - its still just in alpha and not ready for the real world.)
Method of testing:
I do not want to give away too much of my algorithm here (as I am planing to incorporate it as a new page on my site - http://faremechanic.com - to show on a weekly basis what the current value of various miles are), but basically I ran a large set of near in trips within the major 4 legacy US carriers (UA, AA, DL, US - note, this is an alpha test not a prod sampling). I used near in trips because they would have the highest prices, least chance of saver awards, and would probably serve as the most reliable way of testing dollars per mile. After running a sampling, I would then apply how many miles it would cost for each program for an unrestricted plane ticked (standard award). From there I can up with the following mile rates:
AA: $0.012087
DL: $0.011137
UA: $0.016614
US: $0.011112
From there, UA looks like you actually get the most "value" for your mile. Now I will state again, this is an alpha test using only domestic US close in tickets. Applying to a much larger set would give far more accurate results, but the numbers were all within the accepted range (about $0.01 per mile).
So, a few questions for the crowd:
1. Do you see any major flaws in my theory?
2. If I ran this same concept with a larger sample and over the top 50 airlines world wide, would you see value in it on a weekly basis?
3. I considered adding upgrades and C/F tickets in the future, but decided upgrades are too hard to add (not a guarantee that you can get an upgrade - whereas you are guaranteed a standard award in almost all cases) and probably will add a sampling of F - thoughts?
4. Do these numbers fall in line with what you would expect?
My theory to why UA is better than the rest is?
1. 50,000 flat miles for standard US regardless of peak/non-peak (compared to 60,000 at the other 3 with peak)
2. Several close in tickets were more expensive than the other competitors
Love to hear any thoughts and whether this is worth pursuing on a regular basis for a large group.
Gary
(note - this is not a post to advertise my site in any way. Its not even a live site - its still just in alpha and not ready for the real world.)
Yes, it's flawed.
1. Not all mileage is used for last-minute purchases, so the cost of the ticket at that point may or may not be relevant.
2. Close-in purchases, particularly on UA, are when UA is more likely to have Saver awards because of how they (mis)manage their inventory and dump it close to departure.
3. Even if one were to book a ticket last-minute, the cost of that specific flight is not terribly relevant. If you're booking a last-minute SFO-JFK, and UA wants $1,200, but DL and AA want $900, the cost of the UA flight may be more, but the value of the flight may be less, since other carriers have similar options at lower fares.
If you wanted to do a real test, you would need short-term, mid-term, and long-term options, and compare the mileage cost to the reasonably best market fare for a similar trip (e.g., UA is not worth $300 more if DL has a flight in 10 minutes for less).
Also, if your pricing examples are showing that UA is most expensive, that's also not indicative of UA's sales. UA has a lower RASM that most of its competitors.
Sep 12, 2014, 2:16 pm
#4
FlyerTalk Evangelist
Join Date: Dec 2003
Location: LAX
Posts: 10,906
I agree with #3 as redemption should compare to last minute price on any airline.
I disagree with your comments on #2 since UA inventory management style is irrelevant for this evaluation. If they indeed make more saver awards available they are ahead in this category.
(BTW by your logic LH and CX constantly mismanage inventory as they routinely release plenty of award seats close to departure)
I disagree with your comments on #2 since UA inventory management style is irrelevant for this evaluation. If they indeed make more saver awards available they are ahead in this category.
(BTW by your logic LH and CX constantly mismanage inventory as they routinely release plenty of award seats close to departure)
Yes, it's flawed.
1. Not all mileage is used for last-minute purchases, so the cost of the ticket at that point may or may not be relevant.
2. Close-in purchases, particularly on UA, are when UA is more likely to have Saver awards because of how they (mis)manage their inventory and dump it close to departure.
3. Even if one were to book a ticket last-minute, the cost of that specific flight is not terribly relevant. If you're booking a last-minute SFO-JFK, and UA wants $1,200, but DL and AA want $900, the cost of the UA flight may be more, but the value of the flight may be less, since other carriers have similar options at lower fares.
If you wanted to do a real test, you would need short-term, mid-term, and long-term options, and compare the mileage cost to the reasonably best market fare for a similar trip (e.g., UA is not worth $300 more if DL has a flight in 10 minutes for less).
Also, if your pricing examples are showing that UA is most expensive, that's also not indicative of UA's sales. UA has a lower RASM that most of its competitors.
1. Not all mileage is used for last-minute purchases, so the cost of the ticket at that point may or may not be relevant.
2. Close-in purchases, particularly on UA, are when UA is more likely to have Saver awards because of how they (mis)manage their inventory and dump it close to departure.
3. Even if one were to book a ticket last-minute, the cost of that specific flight is not terribly relevant. If you're booking a last-minute SFO-JFK, and UA wants $1,200, but DL and AA want $900, the cost of the UA flight may be more, but the value of the flight may be less, since other carriers have similar options at lower fares.
If you wanted to do a real test, you would need short-term, mid-term, and long-term options, and compare the mileage cost to the reasonably best market fare for a similar trip (e.g., UA is not worth $300 more if DL has a flight in 10 minutes for less).
Also, if your pricing examples are showing that UA is most expensive, that's also not indicative of UA's sales. UA has a lower RASM that most of its competitors.
Sep 12, 2014, 2:19 pm
#5
FlyerTalk Evangelist
Join Date: Aug 2002
Location: Bay Area, CA
Programs: UA Plat 2MM; AS MVP Gold 75K
Posts: 35,067
A better metric would be to measure each of short-, mid-, and long-term (e.g., a week out, 2 months out, and maybe 10 months out).
Sep 12, 2014, 2:33 pm
#7
FlyerTalk Evangelist
Join Date: Apr 2001
Location: MEL CHC
Posts: 20,987
I am 100% sure gldwebs will be sending money to all who make a recommendation for improvements to his/her commercial web site. ^
Sep 12, 2014, 2:39 pm
#8
Original Poster
Join Date: Aug 2004
Location: Round Lake, IL
Programs: UA GS, Marriott Titanium
Posts: 1,064
First off, I have no plans to profit off this. I am a data geek and now that I have a data feed for another project I can use it to play with numbers. If I do this it will be freely available for everyone to look at.
As for the short/mid/long term comparison, here was my issue. Since I cannot query saver availability, it's hard to know how many miles are needed to replace that trip with an award trip. By using 100 domestic trips (that's what I did in my sample) combining hubs, large city non hubs, and medium sized cities I tried to get a good variation of destinations. To compare price to miles needed, you would have to use something comparable. While Ua does sometimes open up last minute saver, it's not always a guarantee. I did my search with consultant type trips - mon to thurs - to try to get what I would believe to be a price that would equate to a standard award. Sure you can sometimes book awards for cheaper tickets, but isn't the real value the maximum you could get for your money? By averaging 100 itins, I figure I have at least removed outliers to get something more accurate.
I'm sure I have holes, but I do not see any reason to put in saver comparisons. If you can get a saver then you got a discount. But the real value, imo, is what you know you can definitely do.
I'll send those that make recommendations a percentage of the zero dollars I make 
As for the short/mid/long term comparison, here was my issue. Since I cannot query saver availability, it's hard to know how many miles are needed to replace that trip with an award trip. By using 100 domestic trips (that's what I did in my sample) combining hubs, large city non hubs, and medium sized cities I tried to get a good variation of destinations. To compare price to miles needed, you would have to use something comparable. While Ua does sometimes open up last minute saver, it's not always a guarantee. I did my search with consultant type trips - mon to thurs - to try to get what I would believe to be a price that would equate to a standard award. Sure you can sometimes book awards for cheaper tickets, but isn't the real value the maximum you could get for your money? By averaging 100 itins, I figure I have at least removed outliers to get something more accurate.
I'm sure I have holes, but I do not see any reason to put in saver comparisons. If you can get a saver then you got a discount. But the real value, imo, is what you know you can definitely do.
I am 100% sure gldwebs will be sending money to all who make a recommendation for improvements to his/her commercial web site. ^
Last edited by FlyinHawaiian; Sep 14, 2014 at 3:22 am Reason: Merge
Sep 12, 2014, 2:43 pm
#9
Join Date: Feb 2008
Programs: 6 year GS, now 2MM Jeff-ugee, *wood LTPlt, SkyPeso PLT
Posts: 6,526
Yes, it's flawed.
1. Not all mileage is used for last-minute purchases, so the cost of the ticket at that point may or may not be relevant.
2. Close-in purchases, particularly on UA, are when UA is more likely to have Saver awards because of how they (mis)manage their inventory and dump it close to departure.
3. Even if one were to book a ticket last-minute, the cost of that specific flight is not terribly relevant. If you're booking a last-minute SFO-JFK, and UA wants $1,200, but DL and AA want $900, the cost of the UA flight may be more, but the value of the flight may be less, since other carriers have similar options at lower fares.
If you wanted to do a real test, you would need short-term, mid-term, and long-term options, and compare the mileage cost to the reasonably best market fare for a similar trip (e.g., UA is not worth $300 more if DL has a flight in 10 minutes for less).
Also, if your pricing examples are showing that UA is most expensive, that's also not indicative of UA's sales. UA has a lower RASM that most of its competitors.
1. Not all mileage is used for last-minute purchases, so the cost of the ticket at that point may or may not be relevant.
2. Close-in purchases, particularly on UA, are when UA is more likely to have Saver awards because of how they (mis)manage their inventory and dump it close to departure.
3. Even if one were to book a ticket last-minute, the cost of that specific flight is not terribly relevant. If you're booking a last-minute SFO-JFK, and UA wants $1,200, but DL and AA want $900, the cost of the UA flight may be more, but the value of the flight may be less, since other carriers have similar options at lower fares.
If you wanted to do a real test, you would need short-term, mid-term, and long-term options, and compare the mileage cost to the reasonably best market fare for a similar trip (e.g., UA is not worth $300 more if DL has a flight in 10 minutes for less).
Also, if your pricing examples are showing that UA is most expensive, that's also not indicative of UA's sales. UA has a lower RASM that most of its competitors.
I think you have a good list of the issues, and I would only add that any survey should (and the idea works survey did not):
(1) look at long range (6 months or even 11), medium range (looking a month out) and close in (looking at say a week or 6 days away),
(2) look at both premium (J or domestic F) and Y seats,
(3) look at saver plus "standard" awards, and
(4) then look at both head to head routing direct, and with a connection.
For example, if the search is STL-NRT, or OKC-MAD its a good head to head comparison. Ditto looking at NYC-LHR, or LAX/SFO-NYC on non-stops.
Sep 12, 2014, 2:47 pm
#10
Original Poster
Join Date: Aug 2004
Location: Round Lake, IL
Programs: UA GS, Marriott Titanium
Posts: 1,064
I definitely agree on the need for J and f. I also agree on the exact comparisons.
I need to figure out if there is a way for me to get saver availability. At least then if I do a comparison I know that the price is related to an exact mileage.
Btw, why don't they ever release a last minute seat on the flights I want.
So, doing it would be fun (I just love numbers), but once I have the numbers is there any real use for it. I guess you could use it for a pay vs award calculator concept or for showing programs who is doing better than others.
I need to figure out if there is a way for me to get saver availability. At least then if I do a comparison I know that the price is related to an exact mileage.
Btw, why don't they ever release a last minute seat on the flights I want.
So, doing it would be fun (I just love numbers), but once I have the numbers is there any real use for it. I guess you could use it for a pay vs award calculator concept or for showing programs who is doing better than others.
Sep 12, 2014, 3:16 pm
#11
Join Date: May 2014
Posts: 202
Why?
I think you have a bunch of data but no questions.
I suggest picking one case and analyzing that. Maybe it is a family if four near JFK who want to go on a long distance vacation in economy. Or maybe it is a retired couple who want to fly business class from MCO to see the grandkids in the Midwest.
You will find yourself picking different sets of cities and advance purchase for each case.
I would also value a flight by choosing the least expensive flight for a trip, not necessarily the carrier that provides the miles.
I suggest picking one case and analyzing that. Maybe it is a family if four near JFK who want to go on a long distance vacation in economy. Or maybe it is a retired couple who want to fly business class from MCO to see the grandkids in the Midwest.
You will find yourself picking different sets of cities and advance purchase for each case.
I would also value a flight by choosing the least expensive flight for a trip, not necessarily the carrier that provides the miles.
Sep 12, 2014, 3:33 pm
#12
FlyerTalk Evangelist
Join Date: Mar 2014
Location: 4éme
Posts: 12,012
If I were using your site, I would want to compare flights and days that are specific to my need. i.e. I want to fly on these dates. That's how I look for flights when paying with $$. Why shouldn't I do the same if I am paying with points?
Sep 12, 2014, 3:51 pm
#14
Join Date: Oct 2012
Posts: 638
Why? is indeed the right question.
Let's say that someone is visiting your website in order to determine which airline to fly for the next few years, with the specific objective of getting the best value for the miles.
If such a person wants saver awards, then the only reasonable criteria is which airline offers the most saver awards. I don't think the relative value of saver awards would be a factor.
In other words, focus on standard awards. Don't worry about comparing the relative value of saver awards.
Let's say that someone is visiting your website in order to determine which airline to fly for the next few years, with the specific objective of getting the best value for the miles.
If such a person wants saver awards, then the only reasonable criteria is which airline offers the most saver awards. I don't think the relative value of saver awards would be a factor.
In other words, focus on standard awards. Don't worry about comparing the relative value of saver awards.
Sep 12, 2014, 3:56 pm
#15
Join Date: Apr 2013
Location: SFO
Programs: MP AA
Posts: 33
If a fare for aaa-bbb 500 miles costs $100 on UA and $200 on AA and both cost be 10,000 miles, which one is a greater value?
UA is $.01 and AA is $.02 if you calculate cpm ($pm). Is AA a better value or do they just have higher priced fares?
The value of a mile is dependendent upon how you travel. I use mine for BF and GF. My cpm is going to be significantly higher then someone flying coach.
UA is $.01 and AA is $.02 if you calculate cpm ($pm). Is AA a better value or do they just have higher priced fares?
The value of a mile is dependendent upon how you travel. I use mine for BF and GF. My cpm is going to be significantly higher then someone flying coach.
