Airplane Pricing Algorithms: Outsmarting Flight Ticket Discrimination

The Lunar New Year holiday is among the priciest time to travel in China, but this year, Chinese internet user Lan He managed to score a one-way flight ticket from Kunming, Yunnan province, to her parents’ home about two hours away in Jian, Jiangxi province, for less than 500 yuan (US$68).

The trick? Fool the algorithms, according to Lan.

Before Lan booked her flight, she unleashed a string of social media posts complaining about costly holiday airfares. “I’m not going because flight tickets are too expensive,” she wrote, singling out some of China’s most widely used travel booking platforms. “Look at me, Qunar, Trip.com, Tongcheng, Fliggy: I’ll uninstall you all if you keep increasing the price.”

Five days after, Lan said she found the cost of her desired ticket went down by more than half from 1,200 yuan.

Qunar, Trip.com, Tongcheng and Fliggy did not respond to requests for comment. Fliggy is operated by Alibaba Group Holding, owner of the South China Morning Post.

Lan is among a growing number of Chinese internet users who are looking for ways to fight what they describe as Big Data-enabled discrimination. They believe that online platforms charge existing customers and frequent visitors higher prices, and occasionally withhold discounts from them.

In 2020, researchers led by Fudan University professor Sun Jinyun found that iPhone users, presumed to be wealthier consumers, received fewer subsidies and discounts on the ride-hailing app Didi Chuxing. They were also offered a more costly vehicle for the same journey, compared to those using Android devices.

Didi’s then-president, Jean Liu Qing, denied that the company’s differentiated pricing was based on the type of phone passengers had. She said the fares of two trips with the same starting point and destination could vary due to differences in the routes taken, travel speed and coupon availability.

That same year, a blogger said he found delivery fees for the same restaurant tripled on Meituan after he became a paid member. The on-demand services platform said at the time that the difference stemmed from a location error and had nothing to do with the user’s membership status.

While some users believe they are able to trick algorithms into offering lower prices with certain online behaviour, such as complaining on social media, industry experts say such practices are unlikely to work in the long run.

Algorithms “provide a comprehensive portrait of each user through Big Data, and a single post online cannot represent all the dimensions of a user,” according to Nankai University law professor Chen Bing, who did research into several internet firms and was quoted by state-backed newspaper, China Youth Daily.

“Counter-training” algorithms might “achieve price reductions in the early stage, but if the user continues to buy and use the service, the algorithms will look at factors such as the order volume, and may increase prices eventually,” Chen said.

Still, Beijing has begun to crack down on the perceived abuse of algorithms.

In December, the Beijing branch of the Cyberspace Administration of China (CAC), the country’s top internet watchdog, summoned representatives from 11 online platforms based in the city – including Meituan and Didi – to discuss issues related to algorithms.

Meituan promised to hold quarterly meetings with experts, users, delivery workers and merchants to refine its algorithms. The company also vowed to improve its treatment of delivery workers by optimising their routes and introducing a mechanism to alert riders or stop them from taking orders if they were working overtime.

That followed the CAC’s announcement in November that it would address “typical issues with algorithms” on online platforms, including unfair pricing that targeted different demographics.

Several domestic Big Tech companies pledged to improve their algorithms. ByteDance-owned Douyin, the Chinese counterpart of TikTok, said it would establish a safety centre this year to make its recommendation system more transparent.

RedNote, the Shanghai-based social media app that has recently gained popularity among TikTok users in the US, sought public comments on how to improve its algorithms. It also reminded users that they could disable personalised recommendations any time.

Pinduoduo, the popular budget-shopping app run by Temu owner PDD Holdings, said it was “actively building a healthier ecosystem” to prevent “Big Data-enabled price discrimination.”