The Cheap Flight: Technology, Oil, and the End of Spirit Airlines

And the technology piece is that this only works at scale if you can track every single fee, every passenger's purchasing behavior, and optimize continuously.

You need serious software infrastructure for that.

Spirit had it.

Y también usaban los datos para saber qué rutas eran más rentables.

They also used data to know which routes were most profitable.

Si una ciudad pequeña no daba suficiente dinero, Spirit dejaba de volar allí.

If a smaller city wasn't generating enough money, Spirit would stop flying there.

No había fidelidad a los pasajeros, solo a los números.

There was no loyalty to passengers, only to the numbers.

Which is ruthlessly logical and also genuinely damaging to communities.

There's research showing that when a low-cost carrier exits a market, fares on that route jump immediately, sometimes by 40 or 50 percent.

The incumbent carriers had been keeping prices artificially low just because Spirit existed.

Eso es muy importante.

That's very important.

Spirit era mala, es verdad.

Spirit was bad, it's true.

Los pasajeros la odiaban, el servicio era terrible.

Passengers hated it, the service was terrible.

Pero su presencia en el mercado obligaba a American, a Delta, a United, a mantener precios bajos en ciertas rutas.

But its presence in the market forced American, Delta, United to keep prices low on certain routes.

Cuando Spirit desaparece, esas rutas van a ser más caras.

When Spirit disappears, those routes are going to get more expensive.

Spirit had the worst customer satisfaction scores in American aviation for years running.

And yet people flew it.

Over and over.

Because the alternative was a bus that took fourteen hours or simply not going.

Exacto.

Exactly.

Y eso es lo que me parece más interesante de esta historia.

And that's what I find most interesting about this story.

Spirit no era una aerolínea para personas ricas que querían ahorrar un poco de dinero.

Spirit wasn't an airline for wealthy people who wanted to save a little money.

Era una aerolínea para personas que, sin Spirit, no podían volar.

It was an airline for people who, without Spirit, couldn't fly at all.

Muchas de ellas eran trabajadores, inmigrantes, familias con poco dinero.

Many of them were workers, immigrants, families with little money.

The democratization argument.

And it's real.

I've been on Spirit flights where half the plane was people who were clearly traveling to see family they hadn't seen in years.

Fort Lauderdale to Bogotá for $80.

That flight matters.

That's not a luxury purchase.

Sí.

Yes.

Y hay otra parte de esto que es tecnológica: las rutas a Latinoamérica.

And there's another part of this that is technological: the routes to Latin America.

Spirit fue muy importante para conectar ciudades de Estados Unidos, no solo las grandes, sino también ciudades medianas de Florida, de Texas, con destinos en América Latina y el Caribe.

Spirit was very important for connecting U.S.

Eso no era un accidente.

cities, not just the big ones, but also mid-sized cities in Florida and Texas, with destinations in Latin America and the Caribbean.

Era una decisión basada en datos demográficos.

That wasn't an accident.

They looked at where immigrant communities were concentrated and built their networks accordingly.

Which is, depending on how you look at it, either a cynical exploitation of underserved markets or exactly what a well-functioning market is supposed to do.

Maybe both.

Las dos cosas pueden ser verdad al mismo tiempo.

Both things can be true at the same time.

Ahora, hablemos del petróleo.

Now, let's talk about oil.

El problema con el modelo de Spirit es que necesita dos cosas para funcionar: aviones llenos de pasajeros y petróleo barato.

The problem with Spirit's model is that it needs two things to work: planes full of passengers and cheap oil.

Cuando el precio del petróleo sube mucho, como pasó con la guerra en Irán, el modelo simplemente no funciona.

When the price of oil rises sharply, as happened with the Iran war, the model simply doesn't work.

Fuel is about 30 percent of an airline's operating costs on a good day.

For a ultra-low-cost carrier like Spirit, that proportion was even higher because they'd squeezed savings everywhere else.

The margin for error was basically zero.

Y cuando el gobierno americano rechazó el rescate financiero, era el final.

And when the American government rejected the financial bailout, it was the end.

Pero eso me hace pensar en algo más profundo: ¿por qué el gobierno no quiso salvar a Spirit?

But that makes me think about something deeper: why didn't the government want to save Spirit?

En 2020, el gobierno dio miles de millones de dólares a las aerolíneas grandes para sobrevivir a la pandemia.

In 2020, the government gave billions of dollars to the big airlines to survive the pandemic.

The 2020 CARES Act bailout was enormous.

American, Delta, United, Southwest, they all got billions in grants and loans.

The political logic was: these are essential infrastructure, too big to fail, too many jobs.

Spirit wasn't quite in that category.

It was big enough to matter to millions of passengers, but not big enough to be politically untouchable.

Hay algo irónico aquí.

There's something ironic here.

Las aerolíneas grandes sobreviven con el dinero del gobierno.

The big airlines survive with government money.

La aerolínea barata, que servía a las personas con menos dinero, no recibe ese apoyo y cierra.

The cheap airline, the one that served people with less money, doesn't get that support and closes.

No sé si eso es el mercado libre o algo diferente.

I'm not sure if that is the free market or something else.

That's a pointed observation and I think it's correct.

The history of American aviation is actually a history of the government picking winners, through bailouts, through postal contracts in the early days, through the whole regulatory structure.

The fiction that it's a purely free market is just that, a fiction.

Y ahora, ¿qué pasa con el futuro?

And now, what happens with the future?

Porque la pregunta tecnológica más importante para mí no es por qué cerró Spirit, sino: ¿puede alguien construir un modelo mejor?

Because the most important technological question for me is not why Spirit closed, but: can someone build a better model?

¿Hay una manera de usar la tecnología para hacer los vuelos baratos de forma más sólida?

Is there a way to use technology to make cheap flights in a more stable way?

That's the question I keep circling back to.

There are a few directions this could go.

Electric aircraft for short routes, which would dramatically reduce fuel dependency.

Hydrogen propulsion, which is further out.

And then there's the software side, where companies are experimenting with subscription models for flying, like a Netflix for short-haul routes.

Los aviones eléctricos son interesantes pero todavía son muy pequeños para las rutas largas.

Electric planes are interesting but they're still too small for long routes.

Para volar de Miami a Bogotá, necesitas muchísima energía.

To fly from Miami to Bogotá, you need enormous amounts of energy.

Los aviones eléctricos que existen ahora solo pueden volar distancias cortas, como dentro de un país pequeño.

The electric planes that exist right now can only fly short distances, like within a small country.

True, but that's a technology curve that is moving fast.

Ten years ago, the idea of electric vehicles dominating city streets seemed like a niche fantasy.

And those Latin American routes, the ones Spirit served, many of them aren't that long.

Fort Lauderdale to Nassau is barely 200 miles.

That's absolutely within the range of near-future electric aviation.

Y hay otro aspecto tecnológico que no hemos mencionado: las reservas.

And there's another technological aspect we haven't mentioned: reservations.

Spirit fue una de las primeras aerolíneas en vender casi todos sus billetes por internet, sin agencias de viajes, sin intermediarios.

Spirit was one of the first airlines to sell almost all its tickets online, without travel agencies, without intermediaries.

Eso redujo mucho los costes.

That reduced costs enormously.

Hoy parece normal, pero en 2005 era una revolución.

Today it seems normal, but in 2005 it was a revolution.

Direct distribution.

Cut out the Global Distribution Systems, which are these enormous old technology platforms that airlines had to pay commissions to for every ticket sold through a travel agent.

Skipping that layer saved Spirit maybe 15 to 20 dollars per ticket, and at Spirit's volume that was the difference between profitable and not.

Entonces, si pensamos en la historia completa, Spirit es el resultado de tres cosas: la desregulación en 1978, el internet en los años noventa y dos mil, y los algoritmos de precios dinámicos en los años dos mil diez.

So, if we think about the complete story, Spirit is the result of three things: deregulation in 1978, the internet in the nineties and two thousands, and dynamic pricing algorithms in the twenty-tens.

Cada tecnología nueva hizo posible el siguiente paso.

Each new technology made the next step possible.

And what ended it wasn't a better technology.

It was a geopolitical shock, a war that sent oil prices through the roof and snapped a business model that had almost no cushion against external shocks.

There's a lesson in that somewhere about fragility and efficiency being two sides of the same coin.

Eso es muy cierto.

That's very true.

En el mundo de la tecnología, hay una idea que se llama 'optimización extrema': cuando eliminas todo lo que no es necesario, tu sistema funciona perfectamente en condiciones normales.

In the technology world, there's an idea called 'extreme optimization': when you eliminate everything that isn't necessary, your system works perfectly under normal conditions.

Pero cuando hay un problema, se rompe completamente.

But when there's a problem, it breaks completely.

Spirit era un sistema extremadamente optimizado.

Spirit was an extremely optimized system.

Lean to the point of brittle.

Same thing happens in supply chains, in hospital systems, in electrical grids.

The pandemic taught us that lesson with just-in-time manufacturing.

Nobody had spare capacity because spare capacity costs money and the algorithms told you to eliminate it.

And then the moment something goes wrong, there's no buffer.

Exacto.

Exactly.

Y para mí, la pregunta final es esta: ¿quién va a llenar el espacio que deja Spirit?

And for me, the final question is this: who is going to fill the space Spirit leaves behind?

Porque esas rutas a América Latina, a las ciudades más pequeñas de Florida y Texas, ahora van a ser más caras.

Because those routes to Latin America, to the smaller cities in Florida and Texas, are now going to be more expensive.

Las familias que viajaban con Spirit van a pagar más o no van a volar.

The families who flew with Spirit are going to pay more or not fly at all.

Frontier Airlines is still there, Allegiant is still there.

But neither of them has Spirit's network.

And the big three, American, Delta, United, they have every financial incentive not to replace Spirit.

Why would you fill a market with cheap seats when you don't have to?

Actually, I just looked this up while we were talking, and Frontier already announced they're adding routes on 14 of Spirit's most-traveled corridors.

Eso es interesante.

That's interesting.

Pero Frontier también tiene problemas financieros.

But Frontier also has financial problems.

El año pasado intentó comprar Spirit y no pudo.

Last year it tried to buy Spirit and couldn't.

Entonces la pregunta es si Frontier va a ser más estable o si va a tener los mismos problemas.

So the question is whether Frontier is going to be more stable or whether it's going to have the same problems.

El petróleo sigue siendo caro.

Oil is still expensive.

The whole sector is walking a tightrope.

And it makes me wonder whether the era of ultra-cheap flying that technology made possible in the 2010s was actually a temporary condition, a window that opened during a period of unusually low oil prices and unusually high digital efficiency, and is now closing.

Puede ser.

Maybe.

O puede ser que la próxima generación de aviones eléctricos o de hidrógeno abra una nueva ventana.

Or maybe the next generation of electric or hydrogen aircraft will open a new window.

La historia de la aviación es siempre así: una tecnología termina y otra empieza.

The history of aviation is always like this: one technology ends and another begins.

Pero en el tiempo que hay entre las dos, la gente con menos dinero es la que sufre más.

But in the time between the two, it's the people with less money who suffer most.

That's the thing I'll take away from this story.

Spirit was imperfect in every measurable way.

The seats, the service, the fees that appeared from nowhere.

But it solved a real problem for real people, and the market and the war and the government between them couldn't keep it alive.

That's worth noticing.

Oye, una cosa antes de terminar.

Hey, one thing before we finish.

Antes dije que Spirit usaba datos demográficos, y tú dijiste que era correcto.

Earlier I said Spirit used demographic data, and you said that was correct.

Pero creo que en español, cuando yo dije 'datos demográficos', usé una palabra que muchos estudiantes no conocen bien.

But I think in Spanish, when I said 'datos demográficos', I used a word that many students don't know well.

Actually, yes.

'Datos' caught my attention because it's one of those words that seems obvious from English but has a much broader life in Spanish.

'Data' in English is almost exclusively technical.

'Datos' in Spanish shows up everywhere.

Exacto.

Exactly.

En español, 'datos' significa información en general, no solo información de computadoras.

In Spanish, 'datos' means information in general, not just computer information.

Puedo decir 'dame tus datos', que significa 'dame tu información de contacto'.

I can say 'dame tus datos', meaning 'give me your contact information'.

O 'necesito más datos para entender el problema', que significa 'necesito más información'.

Or 'necesito más datos para entender el problema', meaning 'I need more information to understand the problem'.

Es una palabra muy flexible.

It's a very flexible word.

So when I hear 'datos personales' in Spanish, that's personal data in the tech sense, like GDPR, privacy policies, all of that.

But 'dame tus datos' is just 'give me your details', like your phone number.

Same word, completely different register depending on context.

Perfecto.

Perfect.

Y si algún día alguien te pide tus datos en España, no pienses que es un hacker.

And if someone in Spain ever asks for your 'datos', don't think they're a hacker.

Solo quiere tu número de teléfono.

They just want your phone number.

Aunque con Fletcher, uno nunca sabe.

Although with Fletcher, you never know.

I'm going to pretend I didn't hear that last part.

Thanks everyone for listening.

We'll be back next week, with presumably fewer fees than Spirit Airlines and slightly more leg room.