The World’s Highest-Earning Airline Route | Hacker News Reader

The World’s Highest-Earning Airline Route

forbes.com

90 pointswallflower88 comments

The article is critically flawed.

This author is merely regurgitating another blog post [0], and wrongly, conflating profits with revenues. The original post didn't make the same representations.

[0] https://www.oag.com/blog/billion-dollar-route

In fairness, only the headline uses "profit", and was conceivably changed by an editor. The Forbes article refers to it appropriately as revenue throughout.

Does makes $1 billion implies it has a profits or $1B, or could it also mean $1 billion revenue.

I had a little thoughts about this the when commenting on Reddit revenue post, someone said make $1 billion means $1 billion profits. But I often sees Financial Newspaper refer Making x amount of money being revenue.

Make (without context) is ambiguous because you don't know what's made.

See the "Revenue Per Hour" paragraph.

The author uses earnings (not earned revenue), which implies income or profits.

That difference makes the article rather uninteresting. It's clear that the highest grossing flights are just the ones which are the most common and most costly. In that kind of situation, airlines might even be competing as loss leaders: strategically operating at unprofitable levels in order to win market share and make up the difference on other flights travelers take. That is, there's a good chance in my mind that not only are these route not the most profitable, they could be the least profitable.

In my tiny little part of the world it would take me 3 1/2 hours to drive 328km to Halifax.

Or pay anywhere from $500 to $900 to fly the 180km as-the-crow-flies distance for a few minutes in the air.

One flight flies 1,300km the wrong way then back, two connections, a a layover and then to Halifax for $1,100.

Ever considered learning to fly and flying it?

Have any friends with airplanes that can drop you off?

Flying private airplanes is probably the most dangerous common thing you can do.

I believe that motorcycling might be more dangerous.

Compare https://aviation.stackexchange.com/questions/198/how-dangero... where depending on the definition you use, "Private aircraft have a fatality rate about 19 times greater than driving" and 'General Aviation has "about one-sixth as many accidents on a per-vehicle-mile basis" compared to driving.'

while http://www.iihs.org/iihs/topics/t/motorcycles/fatalityfacts/... says "the number of deaths on motorcycles was nearly 29 times the number in cars. "

29x (motorcycles) > 19x (general aviation).

Are there any small enough for the one-parachute-for-plane-and-contents technology that can go closer to the speed of sound (like, go about Mach .8 or so), and rely on pulling into stall to slow down enough for the parachute to deploy without ripping out it's part of the airframe? That should be much safer than the current way when operating small aircraft at speeds where that parachute could not deploy directly.

So, they're not nearly as fast as you're describing, but Cirrus does have parachutes on a lot of it's planes: https://en.wikipedia.org/wiki/Cirrus_Airframe_Parachute_Syst...

I know, the issue there is that if you deploy a parachute at over about Mach 0.4~0.5, you will either rip the airframe apart or reduce survival chances of large vertebrae due to the massive acceleration. The solution there would be going into a stall, as that usually serves to provide the drag needed to slow down to safe parachute speeds, but not so much that the airframe can't be reasonably build strong enough to not suffer damage that matters if one considers the plane scrap metal once it lands with the parachute.

This is just too slow to get the main benefit of air travel, which is being able to go at barely subsonic speeds without handling being a massive headache. Of course you will have worse handling at barely subsonic speeds, due to any large deflection going transsonic and having the potential to result in loss of control. But that can easily be handled electronically, especially if the plane is at least somewhat aerodynamically stable, i.e. won't require constant corrections to not go into stall (like some modern fighter jets do).

If you only have to go a couple hundred miles, the cruise speed isn't a significant factor to the overall trip time. Let the airlines handle the 1000+ mile trips (with all the associated hassles); slower airplanes don't give up near as much time on short legs.

Door to door, I can beat the airlines on most trips under 1000 miles. I don't need to get to the airport 90 minutes ahead of departure, go through security, wait for boarding to complete, wait for pushback, connect through a hub, wait for bags, and can often land at an airport closer to where I want to be and have a rental car or cab waiting. The fact that cruise speed is only 215 mph/350 kph is made up for by all the slippage inherent in airline travel.

Yes, the one thing missing that would be being able to start from a normal road, and land on something similar, which could allow pretty much door-to-door is too hard to make work for how many could afford it.

In Alaska, Montana, Idaho, North Dakota, and South Dakota, it's legal to land on rural roads.

Even in places where you can't land on roads, there are WAY more airports than you probably think in the US. It's about 19K, with a little over 5K of those being public use. (Even the private ones are often accessible with just a coordinating phone call.)

Of those, fewer than 400 are "Primary" airports (defined as having scheduled commercial service and more than 10K enplanements [commercial passengers outbound] per year).

Yes, I know there are many airports. This is about starting from "The next road with sufficiently little traffic that has suitable clearance and curvature to start from", and maybe something like folding wings inwards or so to get it small enough for maneuvering it as a large car.

This would have been my reply 5 years ago—when I lived in California. Winter flying in Halifax or anywhere with seasons is not particularly safe, nor practical (ice wrecks havoc on small planes)

Fair point, though on bad winter weather days, a 3.5 hour drive is still a viable option. (Flying on clear winter days is great; airplanes love the cold weather.)

Learning to fly in the USA costs about $10k to $15k. Then about $200 per hour for actually flying the aircraft ( fuel, hire, insurance, airport fees ) so about $600 for that return trip.

I guess the maths might work for very frequent trips or if he could take a passenger who would split the costs.

I've taken flying lessons but it's been 25 years since I've done it. The cost even back then was crazy something like $150/hour. I took off and landed all with an instructor not solo. The instructor also pointed out one day a giant purple balloon I would probably have run into if not for him pointing it out. And barfing, lots of barfing.

This comment is amazing. Is it a ‘yes’ for you? Would that be common anywhere?

Yes - Alaska. Urban legend is there are more pilots per capita than licensed drivers.

I don't know about that stat, but:

Alaska has six times as many pilots and 16 times as many aircraft per capita as the rest of the nation.

https://m.facebook.com/alaskapilot/photos/a.427738310641446....

(I've seen the 6x as many pilots stay previously. This is the first thing I found that repeated it.

/lazy cite)

Looks like this is a re-hash of the data from https://www.oag.com/blog/billion-dollar-route

Bit stupid crunching it by revenue. Obviously the long routes generate the most rev, but they also have the highest fuel costs.

Sydney to Melbourne is #2 and is barely a 1 hour flight.

Even a bad analysis can yield interesting datapoints

> Obviously the long routes generate the most rev,

No, why? If the trip is twice as short, airlines will be able to sell twice as many tickets per day.

Considering the revenue per trip doesn't make much sense from their perspective.

It really surprises me that the Australian domestic flight with Qantas is so high on the list.

Syd-Mel air route is the 2nd busiest in the world, with @55,000 flights/yr.

https://www.statista.com/chart/12598/the-worlds-busiest-air-...

Judging from the chart at the end, it looks like monopolistic price gouging explains it. It's the third-highest revenue per hour of the entire list, but the total hours travelled is fairly low compared to others on the list. Compare the three North American transcontinental flights: half the revenue per hour margins, but far more hours traveled, or simultaneously the lowest-margin and most-travelled routes on the list.

The morning and evening business traffic drives the majority of the profits. You can get to Sydney for ~$100 at the right time of day but I've done MEL-CBR a lot – same direction, not as far – and it's $800 if you want to go at 6:30am and come back at 5pm.

The departures board at the Qantas terminal at 6am is funny. There's a flight to Sydney – a fully loaded 737-800 – every 15 minutes. Over in Jetstar it's the same, Virgin the same.

Not quite Japan levels, where from Haneda to various other cities you can have flights that frequent, but all 777s/787s.

Ohare to one of the nyc airports operates on about a 15 minute schedule. Midway is more like 35 minutes.

I'm not an air expert, but I would guess that they're only including air hours, not taxiing, loading, unloading, etc. On a short-haul, 50k flights per year route, that's going to be a larger %age of each day.

Doesn't surprise me too much. It's a very busy route -- with a smaller selection of both airlines and airports. Plus Qantas is the perrenial corporate airline.

Notice that SFO to Newark makes the list further down. If you summed all Bay Area airports to all New York serving airports. Across all airlines... You might see a closer contest.

They have a virtual monopoly on the corporate market. At the peak times of Monday morning and Friday evening seats can go for A$300-$A400 or more one way for an economy class seat on a 737 for a 55 minute flight.

I think membership of the Qantas club has a lot to do with this, and the fact that it's a Oneworld membership. See my comment in this thread about doing MEL-CBR; I was doing that weekly, got Platinum, and am I loyal to Qantas? Sure as shit I am, I can buy an economy ticket to see my family back in the UK and get access to the international First Class lounges.

Also the fact is that the Business lounge at CBR is practically an extension of the office from 5-7pm. Everyone you know is there.

I wonder how much of their business is a result of QF selectively giving membership to the Chairmans' Lounge? It would surely mean a lot of revenue is steered / kept their way through travel policies.

Do any other airlines have an equivalent to CL? I assume the greater availability of private jets in the US rules it out there.

I don't know, I think the general preferences of a lot of people count for more. I mean, you only have two realistic choices for business: Qantas or Virgin. A friend of mine chose Virgin, for whatever reason, and he tells us their lounges are as nice if not more so than Qantas'... but still he wishes he'd chosen Qantas.

They've got just enough of an edge, plus the whole tie-in to the global Oneworld program. Our company's Chairman is a member of the CL, but AFAIK if I requested a Virgin flight that wouldn't be an issue.

This article talks about "routes" but only shows revenue by individual airlines. For example LAX <=> JFK has tons of different airlines flying between them while Qantas has near monopoly on SYD <=> MEL route (for corp travel, which is a big chunk).

I'd much rather see overall revenues between destinations summed across all flights operating between them.

BA tops the list at 1B revenue NYC to Heathrow. Virgin runs 80% as many flights on this same route, so they should be making in the ballpark of 800M revenue on that route. This is more than a number of other entrants on the list.

So I can only assume that the metric on display is something other than what the title and blog post communicate. It is something like,

    - Select all airline-routes pairs.
    - Group by endpoints and airline.
    - Order by SUM(revenue) DESC.
    - Drop duplicate routes.

That is, for any given route, they report the revenue of the highest revenue airline on that route. They select the top ten routes by the single airline making the most revenue on that route. This is an utterly zany metric and I can't imagine what use there is knowing it, but there you have it.

Qantas does not have a "near monopoly". There's plenty of alternatives.

However, Qantas is the preferred carrier for businesses though, so their tickets during the morning around 8 AM and early evening tend to be fairly expensive, as they are the most popular flights for business travelers.

Qantas is certainly the highest but no way you could describe it as a "near monopoly". Competition is very healthy indeed, even given that Jetstar is owned by Qantas.

Not surprised about the Vancouver to Toronto flights, its cheaper for me to fly from San Fran to Toronto and back versus flying from Vancouver to Toronto and back as indicated by my corporate flight expenses on Air Canada which usually have some kind of discount applied. This gets worse in the surge seasons when everyone is on holidays.

This is unfortunately because Toronto is Air Canada's hub and Calgary is WestJet's hub.

You can get directly to a few places from YVR. But eventually you have to fly through Calgary, Toronto, SFO, SEA, LAX, ORD, JFK, etc to get to many cities.

http://www.yvr.ca/-/media/yvr/documents/air-services/destina...

I am flabbergasted that Qantas has the #2 slot for Melbourne-Sydney! No wonder players like Compass and OzJet tried to shoulder into that market (with little success) 10 or 20 years ago. I know of a couple of friends who live in one city and literally commute to the other every day/week for work, but I didn't realise how many people must to it daily to make this such a cash cow route.

Its the second busiest route in the world. It's a very easy flight, compared with driving and train, and is heavy with business travellers, hence the revenue.

It could still use more services, the flights should be cheaper than they currently are.

It’s not the second busiest by passenger numbers. All of the world’s busiest routes are in Asia.

It’s the second “highest earning” in terms of fare revenue, however.

It is number two in the world by total number of flights per year: http://www.traveller.com.au/worlds-busiest-airline-flight-ro...

It's somewhere around #5 by total number of passengers.

Indeed. That’s because Qantas and their competitors use a large number of small aircraft on the route, maintaining a high frequency.

If you look at the real busiest routes (by passenger count), like Singapore to KL, they tend to use larger wide body aircraft which carry a lot more passengers per flight.

https://www.statista.com/chart/12598/the-worlds-busiest-air-...

The article has some intriguing detail: contrary to my bias, it seems to be that the routes with the most earnings are highly contested (and not the fringe ones with an immediate monopoly).

But I wonder: has this more to do with a) existing demand and b) high optimization capacity (since the routes have dozens of flights/day and include at least one of the airline‘s hubs.

Sounds like a good short-list of places where you might first build supersonic or hypersonic transport hubs:

https://www.youtube.com/watch?v=zqE-ultsWt0

https://boomsupersonic.com/

SFO - FRA has to be incredibly profitable . Only Lufthansa and Unites fly it (codeshare partners ). Regular economy for the rest of the year is around 2300 USD round trip and closer to the flight it’s around 3500. It’s completely ridiculous compared to other more competitive routes like London.

In my experience inter-continental flights to and from the U.K. are often a lot cheaper than from mainland Europe.

As an example, flying with Emirates to Dubai is often a few hundred euro cheaper from London that any other north and central European destinations they fly directly to.

Is that price without any connections? Because I suspect that flight gets most of its passengers from connections.

nor sure I understand the question. it's for a economy return flight (direct). funnily enough you can sometimes get a Lufthansa flight to Copenhagen for around 1000 and that connects through frankfurt. that's because copenhagen is more competitive from SFO. with carry-on that's a way to game the pricing but I heard they cancel the return if you don't check in in CPH.

Yes, that's what I meant (and you exemplified), they charge more for SFO-FRA than for a flight that has one of its legs as SFO-FRA

And I suspect most of the passengers on SFO-FRA are from European connections and don't have FRA as their final destination (or initial origin)

I just checked and you can get a flight with a 2 hour layover for $500. I'm not sure why anyone would pay 3x the price unless their company is covering the cost.

No China locations, which is interesting. Are high speed trains dominating their domestic market?

Definitely the case in Japan at least; Tokyo to Osaka on the Shinkansen is a $150 or so ticket and gets on the order of 500k passengers a day.

There are 10-15 high speed trains between each way per day Beijing and Shanghai (4.5 to 5.5 hour journey time).

Planes are nominally faster (2.5 hours), start earlier (6:30am) and end later (last departure is ~10:30pm, whereas last fast train is ~7pm). There are several airlines serving that route, meaning planes leave roughly every half an hour, and tickets can be had for as little as $110 one way (more at more sane hours or if you buy closer to the date of travel).

There are ~10,000 flights in each direction each year.

> There are 10-15 high speed trains between each way per day Beijing and Shanghai

Are you kidding? There's 50 or more each way per day. Well more than 10,000 a year, more like twice that, and the number will only go up.

I was incorrect, and you were almost correct.

There are 44 high speed trains from Beijing to Shanghai each day.

Of those, 10 have a journey time of <5 hours. 23 have a journey time of >5 but <6 hours. 11 have a journey time of >6 hours.

Huh. I stand corrected. I am surprised it's so few. Possibly a shortage of rolling stock as I hear the line is reaching capacity on weekends.

Nonetheless, the capacity is certainly there with low marginal costs now the infrastructure is built. They could run one every 5 minutes if they wanted - think Tokyo-Osaka. Give it time!

I understand why they don't feel the need to increase frequency (it's actually easy enough to buy a tickey for next day travel) but I don't understand why they don't run all the 'fast' trains at the faster sub-5h speed.

Nominally means “in name only”. Did you intend to use a word with a different meaning?

http://www.dictionary.com/browse/nominally

Sorry, I intended to explain that later in the comment, but got distracted and then just submitted without re-reading.

The flight time is nominally ~2.5 hours, but many/most planes leave late, meaning the actual flight time for the passenger (time between scheduled departure, and actual arrival) is >3 hours.

And airport security and boarding takes longer than train security and boarding. So the time saving isn't all that much vs. the fast train.

If you look only at the stated (=named) departure and arrival times, the plane looks much faster (nominally). If you consider that you need to be at the airport at least 1h before departure, and frequently airports are outside the city centre while train stations are in the city centre, that nominal advantage shrinks considerably.

The usage seems quite right here.

https://en.wiktionary.org/wiki/nominal#Adjective

>Insignificantly small; trifling.

Do you think that accurately describes the difference between 2.5 hours and 4.5-5.5 hours?

A flight time of 2.5 hrs is realistically 4.5 hours when you add in security, boarding, deboarding, getting a taxi on the other end...

Yup, except:

- (in plane's favour) taxi lines at train stations in Beijing are much longer than those at the airport

- (in train's favour) domestic flights in China are often delayed by 30+ mins

I'm perplexed why the trains don't run all through the night for a journey of that duration. How does someone get to Beijing for a morning meeting?

There are trains that run all through the night and can get you from Shanghai to Beijing for a morning meeting.

The D312 leaves Shanghai at 7:10pm, and arrives at Beijing South station at 7:07am. Even if it takes 45 mins to get a taxi, you can still make it to financial street or the central business district for a 9am meeting :)

Seriously, though, would you prefer to take a 6:35am plane, or (if it were available) a 4:10am train scheduled to arrive at the same time? Going to bed at 9:30pm in order to leave for the airport at 4:45am isn't terribly disruptive. But taking the train would likely force you to split your night's sleep into two chunks.

That, or fares are lower?

http://www.anna.aero/2016/08/31/london-to-new-york-market-an...

Number of passengers between London and New York

One of the interesting data points not mentioned is these routes usually use larger aircraft like B747 (BA) and A380 (QF, EK). So the main reason why they are so profitable is primarily down to passenger numbers more than anything else.

What would be more interesting to the layperson is the actual profit the airline makes per flight for each of the top routes.

Is there any data on this anywhere?

I would love to know SFO > Tokyo as I make the flight 1-2 times a month (Tokyo being a hub in some cases to the rest of Asia).

I wonder if it makes sense to compare this to the Shinkansen route from Tokyo to Osaka.

I want to know how BA gets away with their ridiculous fuel surcharges.

They get away with whatever customers are willing to pay for.

What's the SF -> Newark connection I'm missing?

United hubs.

Newark’s proximity to NYC.

Reply on news.ycombinator.com