Hacker News new | past | comments | ask | show | jobs | submit login
A live ranking of airlines by how much luggage they are losing (luggagelosers.com)
113 points by lopkeny12ko 9 months ago | hide | past | favorite | 53 comments



Wow, it can be way more common to lose your bag than I would have thought -- on Aer Lingus, a 1 out of 57 chance! (While Delta is 1 in 497, and Air France is 1 in 1,256 -- more what I expected.)

But the national aspect is seemingly even more interesting -- in India you have a 1 in 97 chance, while in the US it's 1 in 497, and in Japan it's 1 in 7,734.

Now I'm incredibly curious to know what the real, actual culprits are. To what extent is is about the check-in airport or connecting airport that loses it, to what extent is it about airline policies around how luggage is handled, and to what extent is it national regulation that sets standards for airline performance, airport performance, or both?

Because the amount of variation here is just astounding and far, far, beyond anything I would have guessed. I would have naively figured that lost luggage was at a relatively "economically efficient" level and would therefore be pretty similar across airlines and countries... when clearly that is not the case at all.


I think you're missing that the source of the data is social media posts. So garbage data in, garbage data out.


The US DOT publishes "Mishandled Baggage" statistics. These lump together lost and damaged, but for example states that Delta was around a 1 out of 256 chance to get "mishandled"... so, I'm not sure if the stats on this page are really all that far off - they seem all within the range of believable to me.


It says it is normalized against actual published lost baggage statistics.

You can calculate a ratio of social media complaints to lost bags for airlines you have data on and then infer a loss rate for airlines that don't release data.


Someone can just create a Twitter account, tweet that $airline lost their luggage, and skew the statistics.. considering there's some magic maths in the background...


Let's be honest, how likely is that going to happen?

"There's 102 people who die in a car accident everyday. Someone could have just been struck by lightning while driving and skewed the statistics so that number isn't entirely accurate"

cmon bruh


If the 102 number is based on social media mentions... then that's just garbage statistics.


afaik in Japan it's pretty common to use the takkyubin system while traveling where you forward your luggage, so I wonder if that has anything to do with it.

https://en.japantravel.com/guide/takkyubin/31203


Maybe. Or maybe Japanese travellers don't complain about their lost luggage on social media like Indian travellers do?


there is also the whole, do Japanese people even use Twitter to the same extent? Particularly in East Asia you usually see usage of homegrown platforms (LINE in Japan, KaoKao in Korea, the networks behind the great firewall in China)


Or just plain theft at the airport?


Since the source seems to be social media posts, there's a small possibility that theft may be some kind of portion of the number, but I'm not sure due to the questionable accuracy of source data.

Semi-related: I work for a homeless service agency and a few of our clients were arrested for a tiny luggage theft ring at LAX a couple of months ago. One of the detectives said that some of the luggage they found was reported missing and it was luck that someone else had airtags hidden in their luggage to be able to track the item which allowed them to discover more items from others.


> By cross referencing [social media posts] with actual lost luggage data it estimates very closely how much luggage is constantly being lost.

> Using social media as a data source has limitations but seems to be a good proxy indicator when combined with historical lost luggage data.

I'd like to understand the methodology better here. What's the "historical lost luggage data" referenced here?

I'm skeptical of how accurate scraping twitter posts is. Many (if not most) incidents don't get posted about, so I assume they must be doing some extrapolation.


Worth remembering that airlines don’t handle your luggage themselves and instead contract it out to ground handling companies.

It would be more interesting to map flight carriers/numbers to handlers (e.g. Menzies) and regions to give a more reasonable blame/availability overview.


I know that’s true for non-hub airports but is it also true for hubs?

Is Delta really contracting out luggage handling at Hartsfield-Jackson?


I would think it's more likely that luggage handling is centralized at major hubs. For example, there was the infamous Denver Airport Baggage System that was plagued with issues and has been the subject of much analysis [0].

0: https://www5.in.tum.de/~huckle/DIABaggage.pdf


Forgot to answer your actual question: Delta does directly handle its ramp operations at ATL, as well as at BOS, CVG, DTW, MSP, MIA, and SEA. But in the US there are state and federal labor laws to consider and where possible, they, like the other major airlines, offload a lot of staffing generally to Unifi or similar specialist staffing agents. Unifi used to be part of Delta until spun off and sold around 2018 if I remember correctly. Beforer then they were called Delta Global Staffing and primarily handled temp hiring. Now it is 49% owned by Delta but covers hiring for United and Alaska and a few others where possible, and provides the training and certs and all that as well, temp or full time, where possible, including some ramp positions at ATL. SLC, LAX, LGA, JFK are all Delta hubs and positions for ramp agents are entirely absent on Delta's career's page, but on Unifi there are pages and pages that cover almost everywhere in North America where Delta flies. Compare Unifi's open careers page: https://unifi.avature.net/careers to Delta's: https://delta.avature.net/en_US/careers and it should be pretty obvious as to the difference.

This is a bit of generally not very useful osint info but a subdomain search on HR platforms can reveal at least to some degree the labor situation at a lot of entities. avature has over 4000 subdomains in their dns records on securitytrails alone. Not all are for staging or VPN access or SMTP. Some are no longer active, but most do resolve. That's medium potatoes at best compared to the 10k+ at bamboohr. Although if you are doing OSINT you're probably not going for a job at a lot of these staffing agencies anyway. Ever since I was asked to leave a Cutco presentation in 2011 I haven't applied for a job since and I've been poached multiple times and am basically retired at 37. But this is a neat trick and I like dataset gathering for its own sake, and it definitely gives off an interesting view of the economy that may or may not reflect how any one feels about how things are going.

(If I need a job, I prefer fangraphs or baseballprospectus anyway, but they aren't hiring anyone who filed tax returns as professional gambler, sadly)


I know that it at least differs between the US and Europe. I've known ramp agents in a few countries and in the US the contracts in the end are handled by the airlines and training/regulation is governed by the FAA. The couple of European ramp agents are ultimately contracted to the airport (or really, the entity that operates the airport). This is a small sample, but there definitely exists more than one model of how ground staff is managed that is the norm. To further complicate things, the company that pays you may be a subsidiary of the airline you work for or sometimes a different airline or group of airlines, operating under a different name.

I'll give you an illustrative example that hopefully demonstrate how convoluted things can get. Swissport is one of the largest ground services providers in the world. It was first split off from Swissair as its in-house ground service department and became an entity under the holding company SAirGroup in the mid 90s, after plans to merge several of the flag carriers of smaller European countries fell apart when Swiss citizens voted not to join the EEA and therefore, denying Swissair of access to both 5th freedom rights in the EEA and potential cabotage rights. A few years later shortly after 9/11 the airline collapsed as did the holding group which resulted in Swissport being sold off to private equity (Swiss International took over operations on the airline side, sort of, by virtue of a Swiss government bailout and the acquisition of Crossair, the regional arm of Swissair that was divested earlier, by the creditor banks, but in a ton of debt and had to recoup as much as possible. Swissport continued operating after it was sold off and maintained relationships with, well, eventually just about every major airline through mergers and acquisitions, Crossair became Swiss International and was taken over by Lufthansa in 2005. Swissport has maintained relationships with Swiss through Lufthansa but only provides full service to Swiss at ZRH and cargo service at Basel (which Swiss pulled out of in 2015 on the commercial side, and also, is actually located in France, at least airside). It also handles regional operations for Lufthansa out of Munich. ZRH is the main Swiss hub and also a Lufthansa hub, but the fact that Swissport provides full ground service there is almost happenstance since Swiss serves over 100 other destinations and Swissport over 200, many of which overlap, but they do not directly serve the airline elsewhere except cargo operations to Basel. But there is another major airport in Switzerland, and Swissport operates there, but Swiss and Lufthansa Group uses dnata as their handling agent, a subsidiary of Emirates. dnata has no presence in Basel and is a minority operator in ZRH, where Swissport handles something like 80%+ of all ground operations. If you fly Swiss to just about anywhere else in the world, you'll see Swissport, but they aren't likely, save for some prior arrangement where airlines with limited presence may agree to contract through another airline, to essentially provide service in a separate agreement, but these are more ad hoc and subject to change and affects relatively few flights in comparison.

Lufthansa had its own ground services arm that it sold off - LSG Sky Chefs. Gategroup, which owns Gate Gourmet, was the Swissair catering service until being spun off in the same mid 90s expansion that ended up in the collapse. Swissair is no longer an active brand, but it is still valuable to an extent, so it is licensed out, although I have no idea where - like how Pan Am got licensed out to a freight rail carrier, it's likely stamped somewhere random and unrelated to Switzerland or air traffic, but brings in revenue so, why not?

I rarely fly now and when I do, it's domestic US and by charter. That's a whole other crazy mess but somehow, less messy than the scheduled airline industry and far less problematic than the duopoly both having supply chain issues, as both Airbus and Boeing currently does. That's on the news, if you haven't been paying attention, and it's not going away for a bit, by the looks of it. Air travel, meanwhile, remains safer than driving.


Wow, I don't know about the data quality but this ranking pretty much exactly matches my perceived quality ranking of airlines that I have flown, even though I've been lucky enough to never lose luggage. Maybe it's not just luck since I always prefer Alaska or Southwest over United or American where possible.


Cool project, but aren't luggages handled by the airport crew?

Attribution to airlines doesn't make much sense to me. Country ranking is better but anecdotal experience tells me there are some cursed airports.


Yep but getting your luggage back is the responsibility of airlines (and usually whoever they outsource to).

That means airlines that cheap out on that will rank higher for lost luggage because they'll get more mentions on social media for lost luggage (usually after days of not getting it back!).

The airlines that return luggage fast won't show up on social media as much.


It was very unexpected to me that there are differences of 10x or more. Max/min is 187!

Super interesting and thank you for sharing!

    1.87% Air Lingus (1 in 50!  wtf)
    0.71% Spirit (well, it's spirit)
    0.45% Frontier (ditto)
    0.43% United (ditto)
    0.32% American
    0.16% Ryanair (low cost carriers can be good if they want?)
    0.19% Delta
    0.14% Southwest
    0.01% All Nipon (showing off)


I wouldn't trust much on that site.

Notice the stat "Lost Bags (Over 30 Days) By Country.

11 Countries have the same EXACT number of bags lost over 30 DAYS.

922

What is the chance that TWO countries would be exactly the same let alone 11.

I am not a data scientist but I call BULL.


You're certainly right about the 922 number showing up way too often -- that has to be a mistake.

It says:

> My robot scours social media platforms 24/7 for people talking about their lost luggage and which airlines they flew, in 100+ different languages. By cross referencing that with actual lost luggage data it estimates very closely how much luggage is constantly being lost.

I think it needs more transparency on how this is working. There are going to be major cultural differences in terms of how often people complain on social media about lost luggage, and also just differences in frequencies of different terms ("lost luggage", "lost bags", "lost bag", "lost my bag", "lost my luggage") in different languages.

I'll trust actual reported lost luggage data, I don't know what the social media robot thing is doing.


Due to birthday paradox with 46, 2 is reasonably possible. But the numbers are suspect also notice 461 and 1843... Either there is no actual data or they are doing some weird percentage calculations...

EDIT: Too many number are divisible by 461...


In fact, every number in the "Lost bags (last 30 days)" column is an exact multiple of 461 or within 0.04 of a multiple of 461.


nice catch, thank you


Very cool. Looks like 5 of the airlines all lost the same number (461) of bags though, and 12 more airlines lost 922 bags. Data acquisition error?


And 922 is twice 461


This is cool, it’s also not the most accurate.

The US DOT requires airlines to submit reports on mishandled luggage.

One such is available here; https://www.transportation.gov/sites/dot.gov/files/2024-05/M... Edit: see page 39

Point to point carriers such as Allegiant and Spirit typically do much better than carriers that operate a connection like Delta and United.

Since they are souring their data from people on social media I suspect the chance that someone complains about missing a bag varies more than the amount of missing bags.


i would also imagine that Spirit and Allegiant both charge for checked and don't have frequent flier programs that waive that fee.


> Since they are souring their data from people on social media...

I don't know if that's a typo of 'sourcing', 'scouring' or you meant souring, but all work in this case.


There is a regional issue that can come about. We hung out with friends a little over a week ago in Croatia and Italy. We came back via Rome -> Copenhage n -> SFO. No luggage issues on SAS.

Friends flew Split -> London -> SFO on British Airways. Apparently the day they left a large percentage of luggage that transited the BA terminal in Heathrow had issues. One bag was lost (returned to the US) for about 5 days, the other is still location unknown.

A big problem is that gate people and checkin people will suddenly claim "the flight is full" (some times it is) and try and force people to pack bags. When tasked with such, people don't always think about possible valuables in their bags and rebalance accordingly.

In the case of our friends. Their "carry on" was not much bigger than a backpack. The flight was not full. The recovered bag, thankfully, had jewelry and other items the wife of the couple didn't think about when agreeing to check the bag.


Note, the data source is social media posts.


Interesting charts to see! Interesting way to drive some airline accountability with data. I wonder if useful for regulatory bodies as metric for overall operational scorecarding, or correlations with overall airline safety/incidents.


India, UK, Canada being on top of the losers list make total sense. Their countries will also rank similarly in any other arbitrary criteria.

I'm surprised to see Swiss so up there.


Swiss Air is owned by Lufthansa and part of the Star Alliance. Also, being from a small country almost all destinations are outside Switzerland and suffer from whatever care the local handler give to your baggage...


Unsurprisingly point-to-point airlines lose less luggage. I'm guessing that those numbers are greatly influenced by which airports you connect through or how commonly an airline flies through problematic airports etc.

US Airlines are probably also less often going to lose luggage since it's typically (from my experience at least) impossible to check through luggage into a domestic connection.


I wonder how much this is airlines vs. airports. There was a time I needed to fly a lott for work. You got to know never to chech in luggage through Helsinki or Heathrow, as the chaces of entering the lost luggage carousel were >40%, regardless of the airline taken.


How much of luggage handling is the airline vs. the airport crew? I had assumed it was mostly the latter.


And also the whole airport luggage handling system. These are not simple anymore, but instead massive automated systems. With all the usual issues of dealing with real physical objects in addition to identifiers associated...


I wonder, why numbers of lost bags (last 30 days) are repeating for different airlines? For example All Nippon Airways (ANA), LATAM Brazil, Malaysia Airlines and Air Transat lost 461 bags each. This month it was even more popular to lose 922 bags.


Formula seems to be round(num_social_media_posts * 460.85)

Where 460.85 comes from is unsure, but I guess a calibration factor based on some more reputable data set.


What is with the combined 'GB' and 'IE' characters? Why do they exist and why would anyone use them, given that they break the ctrl+F functionality.

Edit: They don't show up when I copy paste them to a comment.


I think those might be the UK and Ireland flag emojis but they are not present on all platforms or fonts so you see the substitute.


They're flag emoji, which Windows doesn't display (likely for political reasons)


British Airways lost my bag but found it again when I boarded my return flight, they paid all of the nice replacement clothes I bought during the trip. So I’m fine with them losing my luggage again!


I had a similar experience. My bag was delayed by a day, so Alaska paid for the ~$2000 to cost to replace the contents. I received my bad via delivery a few days later


What I want to know now is: if a bag has a functioning tracker, such as by Apple or Samsung, does that help facilitate its recovery?


What a fun site and cool idea! Very interesting. A bit more depressing but I’d be a bit curious about rates of lost pets in cargo


Wanna see a plot of this against each company's stock price deviation from the airline industry benchmark.


Yuck. British Airways is right at the top. I just bring carryons only.


Airlines less traveled always have best score in the data.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: