Would Stefanie Sun still be a pop superstar if she started out in today’s AI-driven music industry?
Machine learning programs serve up new playlists to China’s 520 million music app users, providing openings for unknown talent but disrupting the traditional system of superstars backed by major labels and distributors
Liu Guannan graduated from medical school in Chengdu but his dream was to become a full-time songwriter. So the 27-year-old lived with his parents for two years and faced criticism for being a “drop out” until NetEase Cloud Music, a Chinese music streaming app similar to Spotify, changed everything.
With no professional music training and no significant awards to his name, Liu now makes a decent living writing songs for his folk duo called Fine and enjoys the fame that comes with having more than 100,000 fans.
“I can’t figure out how our songs got that many followers. We didn’t have an agent, we didn’t have the money to promote ourselves and we didn’t have any connections in the music industry in China,” said Liu, whose band started off as a college student pastime but thanks to its online success had its first commercial live concert in early 2017.
“All we did was upload our songs on as many music streaming sites as possible. Then months later we magically become famous.”
Streaming music made global stars out of previously unknown performers such as Lorde and Zara Larsson, each with billions of streams on Spotify, and now the same phenomenon is set to change the music industry in China, where rampant piracy of online music is giving way to fans signing up for online streaming services that offer free songs. Although digital music revenues in China nearly quadrupled to US$195 million over the five years to 2016, with most of that coming from streaming, the business is still tiny compared with global music sales of US$7.8 billion, according to figures from the International Federation of the Phonographic Industry.
“Now younger generations can directly reach an audience by simply uploading their work online. These new technologies help musicians promote themselves and bring better prospect in turns for them,” said Peter Li Si-Song, a Singapore songwriter who has produced albums for Asian stars such as Jacky Cheung, Sandy Lam and Stefanie Sun.
“If Stefanie Sun was to become a singer today, she would have a very small chance of becoming a superstar,” he added.
A representative for Sun declined an interview request, saying Sun was very busy and that they didn’t think the interview fits in their plan to promote her latest album.
While obscure artists like Liu and Fine’s vocalist Qiao Xi might just upload their music and hope for the best, the real magic happens behind the scenes with use of artificial intelligence to recommend music to individual listeners.
Chinese streaming platform NetEase Cloud Music uses machine learning technology and algorithms to predict what music its 400 million users – equal to more than half of China’s internet population – would like to hear, providing them with personalised recommendations that can boost the fame and fortune of unknown performers.
The technology is changing the shape of China’s music industry, just as it has done in western markets over the past decade.
Once known as a haven for piracy, China is expected to see its first initial public offering in the music streaming sector in 2018. Shenzhen-based Tencent Holdings, which in December formed an alliance with Swedish streaming giant Spotify, is planning to list its music arm in what will be one of the most anticipated IPOs of the year, according to Bloomberg.
“In the old days, people could only listen to a song when they knew the song’s name or the singer. That is why we use algorithms and artificial intelligence to expose people to music they don’t already know about,” said Wang Shimu, vice-president of NetEase Cloud Music, a unit of the New York-listed Chinese gaming major NetEase.
By changing how people discover music, NetEase Cloud Music is revolutionising how music becomes popular and how profits get distributed in China’s music industry.
“This is a revolution much needed by the music industry in China,” said Wang at the company’s headquarters in the Hangzhou campus of parent NetEase.
Under the old business model, profits in the music industry were mostly shared by the distribution channels and major music labels, with lesser known artists receiving what was left over.
“China has hundreds of thousands or even millions of talented musicians, most of them struggling to lift themselves out of poverty,” Wang said, adding that about 50,000 independent musicians have uploaded 1 million songs on NetEase Cloud Music’s platform, accounting for 10 per cent of its total playlist. The company wants to lure more musicians to further diversify its song offerings.
Similar to Spotify’s Discover Weekly playlist, a feature which provides users with a personalised weekly round-up of songs they might like but have not yet listened to, NetEase Cloud Music and its Chinese rivals have taken that to the next level by providing personalised recommendations in real time based on an analysis of user behaviour including listening habits, their profile (age, gender, location), what features of the app they use, the singers they follow and also who their online friends are.
Since 2015, NetEase Cloud Music, has been providing each user with their own homepage with personalised playlists, a feature that Wang said can “bring new music to the masses” and “decentralise” the music industry.
“Through the feature, we want people to find a whole new world,” said Wang.
Most of China’s music streaming sites, including Tencent Music’s QQ Music, Kugou and Kuwo, and Xiami operated by Alibaba Group, owner of the South China Morning Post, have similar features. But NetEase Cloud Music claims it is one of the most aggressive ones in terms of investing in artificial intelligence and algorithms.
The ability to offer personalised song lists is proving to be a strong incentive to hook China’s internet users. Even though there are about 524 million mainland Chinese registered to use music streaming apps, the market is not mature enough to support a paid business model for music but is still in a better place than the previous environment when netizens would try and download pirated songs.
Music streaming companies have been willing to pay for copyrighted content even without charging listeners because of the huge business potential the artists represent.
NetEase Cloud Music users spend an average of 90 minutes a day on the app, according to the company. This is seen as significant user engagement time that is expected to transform into big business in three to five years when more users are ready to pay, Wang said.
While music streaming technology has levelled the playing field, there are obvious downsides.
“The threshold for publishing a song is too low now,” said Li, the music producer.
“When I first started as a singer-songwriter more than 30 years ago, we needed to send demos to record companies by post. Most of the attempts went in vain, you had no idea whether they listened to your songs or not.”
He even attributes the rise in music streaming sites and supporting technology as ruining the chances for some of Asia’s up and coming performers to become superstars over the past several years.
“People are easily distracted by the overwhelming offerings online these days. Navigating through the tens of millions of tracks filled with new and novel songs online is difficult enough, let alone telling the good from the bad,” Li said.
For songwriter Liu Guannan, not achieving superstar status is just “fine” – which is why he named the band that. “I never wanted to be a superstar. I am perfectly satisfied with where I am today. It is much better than never having had the chance in the first place,” he said.