Many of us spend hours on the internet researching flight deals,

許多人花幾小時在網路上查機票價格，

trying to figure out an airfare pricing system that seems random.

試著搞懂看似隨機的機票定價系統。

Fees appear to fluctuate without reason, and longer flights aren't always more expensive than shorter ones.

費用似乎是毫無來由地波動，而且較長程的班機不見得總是比較短程的貴。

But behind this is actually the science of dynamic pricing,

但這背後其實是變動定價的科學，

which has less to do with cost and more to do with artificial intelligence.

它和花費較無關，而是和人工智慧較有關。

It's been more than a hundred years since the first scheduled flight.

第一班經過規劃的航程距今已超過一百年了。

With the Comet scheduled by British Overseas Airways Corporation

由英國海外航空公司制定航程的彗星號，

to start the world's first jet passenger air service—

開始世界上第一班以噴射機載客的服務了。

In the beginning, commercial aviation was a tightly regulated marketplace,

一開始，商業航空是個價格受到嚴禁管制的市場，

where airfares were based on distance traveled rather than passenger demand.

機票價格都是依據飛行距離而非乘客需求量。

Most international routes were operated by a single national carrier,

大多數國際路線都是由單一國內航空公司營運，

and the lack of choice resulted in uncompetitive fares for consumers.

缺乏選擇導致價格無從競爭。

Deregulation, however, changed all that.

然而，管制的取消改變了這一切。

In 1978, the U.S. government began the process of removing government controls

在 1978 年，美國政府開始移除

over routes and market entry for new airlines.

對於路線和新航空業者進入市場的管制。

In 1983, fares followed.

在 1983 年，價格也放寬了。

And by the early 1990s, the liberalization of the aviation industry had spread around the world,

而到了 90 年代早期，航空業的鬆綁已散播至世界各地，

making pricing more competitive.

使得價格更有競爭性。

Critics also say there was too much competition in the first years after deregulation.

評論家也指出，在取消管制的頭幾年競爭太高了。

Airlines merged and became more dominant,

航空公司合併、變得更有支配性，

changing the industry from a regulated cartel into an unregulated cartel.

改變了這項產業，將它從受限的同業聯盟變成不受限的。

In the U.S., just 4 airlines control 68% of the domestic airline capacity.

在美國，僅僅四間航空公司就占了 68% 的國內航班承載量。

While ticket prices have been falling between major hub airports,

當主要機場間的機票價格降低時，

the lack of competition and a reduction in flights has resulted in higher average airfare for smaller, less competitive cities.

飛往較小、較沒競爭力的城市的平均票價會因為競爭力低和班機縮減而提升。

But this hasn't halted the demand in air travel.

但這並未減少飛航的需求。

The number of airline passengers across the globe has continued to grow,

全球飛機乘客的數量逐漸在增加，

particularly in the last 10 years, with the amount of people choosing to fly skyrocketing by nearly 1.8 billion.

尤其在過去十年，選擇飛行的人數飆升至 18 億。

So what does this increasing volume of passengers mean for pricing tickets?

所以乘客數的增加對於機票定價意味著什麼？

Despite steep surcharges and baggage fees tacked on during the '07-'08 oil shock,

即便在 2007 至 2008 年石油危機期間有許多誇張的額外收費和附加行李費用，

ticket prices aren't actually focused on the seats' combined cost such as taxes and fuel.

票價其實並非聚焦在座位的合併費用，如稅負和燃油。

Instead, airlines price tickets using a strategy called airline revenue management.

航空公司反倒利用一種稱為「航班收入管理」的策略制定票價。

Its end goal? Make as much money as possible.

最終目的為何？盡可能賺更多的錢。

This is working in real time.

這會在即時的情況下運作。

So when a customer goes to book a seat, the airline determines the price they see

所以當一名客戶去訂機位，航空公司藉由分析一系列的因素，

by analyzing a wide range of factors including the status of their entire network.

包含他們整個網路的狀態，來決定他們看見的價格。

Of course, these decisions aren't being made by humans.

當然，這些決策並非人類所為。

Algorithms adjust fares by using information like past bookings, remaining capacity,

透過演算法利用各種訊息，包含過往訂購、剩餘座位、

average demand for certain routes, and the probability of selling more seats later.

特定航線的平均需求、和之後銷售更多位子的可能性，來調整價格。

One example of this process being used is for pricing connecting routes.

定價轉接航班就是利用這種方法的其中一個例子。

Let's say you want to fly London to Miami.

假設你要從倫敦飛到邁阿密。

The flight you want to book also serves as the first leg of a connected flight to Bogota.

你想訂的航班同時是去波哥大的第一班轉機航班。

While you're traveling less distance to Miami,

即使你飛到邁阿密的距離較短，

you may end up paying more than passengers flying through to Bogota.

最後你卻可能要比直飛到波哥大的旅客付更多錢。

That's because airlines are trying to discourage you from buying seats

那是因為航空公司在試圖阻止你買

that they want to keep available for the full journey.

那些他們想保留給飛全程乘客的位子。

This entire thought process is done by an algorithm,

這整個思考過程都是由演算法而來，

which predicts those seats will sell for a higher price in the future.

它預測那些位子在未來可以賣得更高價。

Airlines also profile their customers to help them adjust prices.

航空公司也會將他們的客戶建檔，藉此幫助他們調整價格。

This often means placing passengers into one of two groups: leisure or business.

這通常意味著把客戶擇一歸類：休閒型和商務型。

And the way each group is priced is very different.

而兩種類型的定價方式非常不一樣。

An airline offering a flight from London to Bali can assume that the people on that route are largely holidaymakers,

一家提供倫敦到峇厘島航線的航空公司可以假設這趟航班大多數乘客都是去度假的，

so it places them in the leisure bracket.

所以把他們歸到休閒型。

These passengers usually book months in advance,

這些乘客通常會早幾個月訂票，

so airlines tend to start the price for these seats relatively high.

所以航空公司通常會讓這種位子一開始起價相對地高。

It then adjusts the price according to market response, making sure it's high enough to maximize profit,

然後接著根據市場反應調整價格，確保足夠高價，能夠使收益最大化，

but low enough so that it doesn't result in unused capacity.

但也要夠低價，才不會滯銷座位。

While on a typical business route, such as London to Hong Kong,

而一個典型的商務路線，比如從倫敦到香港，

airlines usually start with low prices to fill a minimum capacity.

航空公司通常會從低價開始，好達到最低承載量。

Then it increases prices steeply for business travelers who book last minute.

然後劇烈提升價格，為的是那些在最後一刻才訂的商務旅行人士。

That's because airlines know business travelers tend to book later

那是因為航空公司知道商務旅行者通常晚點才會訂，

and are far less price sensitive than their leisure counterparts.

而且相較於休閒型的客戶，他們比較不會在意價格。

In fact, business travelers are willing to pay 60% more on average to secure a seat.

事實上，商務旅行者願意付多於平均價 60% 的價格好確保座位。

Some analysts believe airlines use consumers' internet browser cookies

有些分析師相信航空公司利用客戶網路瀏覽器的記錄存儲 (cookies)

to determine what flights they've been looking at.

來判別他們在找些什麼航班。

This way, they can increase those prices to encourage them to buy.

如此一來，他們可以提升那些價格好促進買氣。

But solid proof of this practice is hard to come by,

但是這類行為的確實證據很難獲取，

and it's something airlines and price comparison sites strongly deny.

而且是航空公司和比價網站極力否認的。

Targeted dynamic prices however can increase customer satisfaction and encourage consumer bookings.

然而，鎖定的變動價格能夠提升顧客滿足感，並且鼓勵顧客訂票。

For example, a member of an airline loyalty program could receive

舉例來說，一家航空公司忠誠計劃的成員

'just for you' price displays based on their purchasing history.

可以藉由他們的購買記錄收到「獨享」票價選項。

Technology has also allowed some airlines to create a 'basic economy fare'

科技同樣使得一些航空公司能夠適度地提供「基本的經濟票價」，

with limited amenities to compete with minimal service, low cost carriers.

好和提供最少服務、最低價的航空業者競爭。

That lower fare is key if full service carriers want to appear on the first page of search engines like Google Flights.

若是全服務航空業者想要出現在類似 Google Flights 搜尋引擎的第一頁，較低價格就是關鍵。

But it's not just airlines that are using AI technology to their advantage.

但並不是只有航空公司因使用人工智慧技術謀利。

Consumers now have access to sites which monitor fares, using their own algorithms and past data,

現今顧客能夠使用各網站監測票價，利用它們自己的演算法和過往記錄

to predict the lowest price a seat will reach to then alert their customers.

來預測一個座位能達到的最低價，並得到通知。

The threat airfare search sites pose to airlines' dynamic pricing system

票價搜尋引擎對航空公司的變動定價系統帶來威脅，

even saw United Airlines suing the website Skiplagged, which helps passengers find loopholes for cheaper tickets.

美國航空甚至控訴網站 Skiplagged 因它幫助乘客鑚漏洞買較便宜的票。

With AI having such an important role in the field of air travel pricing,

有了重要的人工智慧進入機票定價領域，

it's likely that complex algorithms will continue to fight the airfare war.

複雜的演算法很可能會為了票價繼續爭鬥。

That may not be a bad thing, as airlines manage a growing number of passengers

那或許不是壞事，因為隨著航空公司必須掌控日益增多的乘客量，

and consumers push for better choices and easier ways to book their trips.

乘客也在爭取更好的選擇和更容易的訂票方式。