Monetization is the Key for Chinese Short Video Startups Going Global, and AI Could Help
As Chinese short video platforms expand overseas, a gap between customer acquisition and commercialization persists.
Bloomberg reported on September 12 that Kuaishou, the domestic version of Kwai, is planning for an IPO next year, citing people familiar with the matter. Previous investors in Kuaishou include Tencent, Sequoia, Baidu, China Media Capital, DCM Ventures, Morningside and others.
One of the leading short video platforms in China, Kuaishou has been expanding into the overseas market aggressively, in a pattern comparable to Douyin, the domestic version of TikTok. Its recent target Brazil has seen the number of active daily users surpassing 3 million, according to Jiemian.
The overseas market consists of upward of 3 billion users. China’s short video platforms have started to make a move into Southeast Asia, Japan, South Korea and India. TikTok now permeates these markets, leaving few opportunities for Kwai. The timing of global strategies has been crucial in these developments. Currently, the cost of customer acquisition is significantly larger, almost three- to fivefold compared with that of three years ago, which left Kwai in an unfavorable position.
However, there remains a significant gap between customer acquisition and commercialization. Despite the low cost of customer acquisition and the still (relatively) un-competitive market, it’s still hard for Chinese players to monetize the traffic. "TikTok is having a hard time to commercialize and the user number could drop significantly once it stops marketing", people familiar with the matter said to Jiemian. "There's a chance to compete with TikTok if Kwai finds the right vertical markets and achieves monetization."
Just as Douyin labels itself an AI company, Kuaishou also uses AI as a fundamental tool behind the logic of monetization.
How do these video companies utilize AI in making money?
Optimizing user experience. AI applications in this area, with the ability to improve visual quality, select content and recommend how to generate content, are all closely associated with improving user experience.
AI predicts people's interests, recommends and pushes content directly to them, and even automatically generates content to suit their interests.
Better experience brings more traffic which is the key to generating revenue. Traffic enables mass advertisement, user subscription, user interaction and many monetization opportunities.
Personalized recommendations. Traditional Internet companies, such as Baidu, Alibaba and Tencent all develop, acquire or invest in video platforms – iQiyi, Youku, Tencent Video respectively. These Chinese video platforms function in a manner similar to YouTube and Netflix, providing mostly purchased, self-made, and sometimes user-generated series and films. These platforms charge users for membership fees.
Machine learning models personalize video rankings, highlight trending topics, filter irrelevant results, etc. Netflix released a research paper that reveals that 80% of hours streamed on Netflix are based on recommendations (driven by its recommendation system), while the remaining 20% is search-driven.
Personalized content pushing. Machine learning algorithms transform the content industry from "people searching for content" to "content reaching out to people". Technology development generates new consumer products, such as ByteDance's TikTok (Douyin), which went viral from China and spread across the world. This kind of mobile app is wholly enabled by AI. Instead of depending on word-by-word searches with clear targets or channel-by-channel browsing with vague interests, the short-video platforms directly feed below-60-second, user-generated short videos to users.
By capturing 'liking', 'disliking', 'commenting' from users and metrics such as 'response time' and user data collected from other partners, the AI decides which videos to show next, continuously catching users' eyeballs. In China, the average daily use time is 31 minutes (over 120 videos); in the U.S., monthly usage in October 2018 was 6.8 hours. The user traffic generates a large amount of advertisement revenue, with the value transferred from the brands to Douyin.
A similar case is Tencent-backed Kuaishou. The short-video mobile app targets China’s grassroots society, i.e. lower-tier city people. With a similar effect of attracting user traffic, Kuaishou targets a variety of users in rural and relatively lower-tier cities. Kuaishou has less revenue from big brand advertisers, with more coming from the live broadcast function. As consumers watch the live show, they have to pay the platforms for the opportunity to interact with the broadcasters, normally by giving virtual gifts. Kuaishou gains revenue mainly from this 'virtual gift tax', as the middleman connecting audience and short-video casters. The company achieved a revenue of CNY 20 billion in 2018, with over CNY 19 billion coming from the live broadcasting business.
Directing users to E-commerce platforms is common in the short video market. E-commerce commissions represent an important path to profit for short video players. Douyin has closed a deal with E-commerce platform Taobao totaling CNY 7 billion, with a CNY 6 billion advertising fee and a commission of CNY 1 billion.