Qutoutiao Stock

Qutoutiao Inc. operates mobile content platforms in China.

The company’s eponymous mobile application, Qutoutiao, meaning ‘fun headlines’ in Chinese, applies artificial intelligence-based algorithms to deliver customized feeds of articles and short videos to users based on their unique profiles, interests and behaviors. Qutoutiao has attracted a large group of users, many of whom are from lower-tier cities in China. They enjoy Qutoutiao’s fun and entertainment-oriented content as well as its social-based user loyalty program. Since its beginning as a news aggregator, Qutoutiao has evolved and developed significantly on the content side and also has a rich offering in short video content, online games and live-streaming.

Midu, first launched in May 2018 as Midu Novels and with an alternative version Midu Lite launched one year later, pioneered provision of free online literature supported by advertising. The company has also been developing standalone short video applications and mobile games among the many other initiatives in motion since early 2019 to further diversify the range of products its users can enjoy and the ways in which it can bring value to its users. The company’s mobile applications have gained a certain level of success and popularity among Chinese mobile internet users since launch, with combined average MAUs (the number of unique mobile devices that accessed the company’s relevant mobile application in a given month) of approximately 97.6 million, combined average DAUs (the number of unique mobile devices that accessed the company’s relevant mobile application on a given day) of approximately 25.0 million and average daily time spent per DAU of approximately 43.2 minutes in the three months ended December 31, 2021.

Mobile Applications

The company primarily delivers content through Qutoutiao, which is its flagship mobile application and means ‘fun headlines’ in Chinese. Qutoutiao aggregates articles and videos uploaded from content providers and presents real-time customized feeds to users based on each user’s profile, behavior and social relationships.

Feeds are presented to users on both the main page of Qutoutiao and topic pages. Both the main page and topic pages are customized for each user using the company’s content recommendation engine. Topic pages include, among others, videos, entertainment, humor, anecdote, relationship, family, health, food and pets. A user may also search content or follow specific content providers. Users may save their favorite content pages, as well as indicate the content pages that they dislike.

The company promotes social interaction among users to engage them more closely with the content they have viewed, as well as with each other. Users may post comments and engage in discussions with other users by responding to comments. A user can also share content through a variety of means, including emails, messaging applications or social networks.

The company launched its mobile literature application, Midu Novels, in May 2018, which offers users free literature supported by advertising. Unlike the traditional paid-only model in the online literature industry, which charges users fees for most content offered, users of Midu Novels can enjoy their favorite literature under an innovative free-to-read model. The company’s users have access to a comprehensive selection of literature covering a wide range of genres, including romance, fantasy, science fiction, history and other genres for free. The company classifies the content genres with multi-dimensional reading tags, which its users can choose to follow. The company is able to offer its users free literature as it primarily monetizes through advertisements that its advertising customers place on Midu Novels. The company’s AI-powered content recommendation engine coupled with its strong data analytics capabilities also enable the company to improve user experience and increase the time its users spend on Midu Novels by making personalized recommendation and delivering to them literature that caters to their interest.

The company launched Midu Lite in May 2019, which combines a loyalty program with the standard offerings from Midu Novels. The differentiated product design has led to Midu Lite attracting an incremental reading population whose needs have not been well addressed by Midu Novels previously. As a result, the company has observed minimal user overlap between the two versions.

The company has also been exploring opportunities in short videos, which it offers through both the Qutoutiao application and standalone short video applications. It is a large and attractive market and an important part of any content ecosystem.

User Account Systems and Loyalty Programs

The company offers user loyalty programs on its mobile applications. Registered users can earn loyalty points if they become active users, refer others who later register and become active users, or engage in various activities while logged in.

Referral-based Loyalty Points

The company’s registered users earn loyalty points when they invite others to download and register on its mobile applications. After an invited user registers with the company, the existing registered user is eligible to receive loyalty points or cash credits. The company is thus able to leverage the embedded social relationships of each user and prompt its users to voluntarily invite their families and friends to become its registered users.

Engagement-based Loyalty Points

A user is eligible to receive loyalty points for engaging in various activities on the company’s mobile applications. Such activities include viewing and sharing content, providing valuable comments and encourage inactive users to continuously re-engage with the company’s mobile applications. The company also creates fun tasks, such as daily missions to tap into the competitive reward psyche of users.

Content

The company strives to become the light entertainment content platform of choice for a growing base of users. The company primarily delivers content that can be viewed by users during a short period of time. The articles on the company’s platform generally contain both images and short texts and can be read within one minute; and a majority of the videos on the company’s platform run less than two minutes. In the three months ended December 2021, the average daily time spent per DAU for the company’s mobile applications was approximately 43.2 minutes. The company aims to deliver quality and relevant content to users, and content sourcing, management and recommendation are among core focuses of its operations.

Content Sourcing

The company sources content from approximately 4,995 professional media outlets under a licensing arrangement and from more than two million freelancers registered on its platform. The company operates an online content upload system for content providers to prepare and upload content. Fees paid to content providers relates to the amount of views associated with such content.

A content provider that is new to the company’s online content upload system is required to go through a registration and approval process. Each content provider is required to sign an agreement electronically in the registration process. The company has the right to freeze an account for any violations of the rules, such as plagiarism or submission of inappropriate content.

After its registration with the company, a content provider can prepare and upload content electronically through the online content upload system. The system also allows each content provider to track its performance on a real time basis, including information, such as the number of views, comments, shares and saves for its content.

The company’s mobile literature applications, Midu Novels and Midu Lite, primarily source content from traditional PC-based online literature platforms, which grant the company permits to publish their literature content on the company’s platforms for a fee. The company has also built an in-house editor team that works with authors directly, which combines human experiences with data analytics in guiding and producing quality literary works. The company curates quality literature content that caters to its users’ interests based on the company’s analysis of users’ profiles and their reading histories. As of December 31, 2021, the company offers more than 206,000 pieces of literature on Midu Novels and Midu Lite.

Content Management

As a gatekeeper for the company’s platform, its content management system is designed to ensure both the quality and appropriateness of information presented to users, which include content and comment postings. Content may be declined for quality reasons, such as videos or pictures of low resolution or duplicative content. The company also declines content and comment postings that appear to violate relevant laws and regulations or are otherwise inappropriate for its platform. The company undertakes an efficient and thorough screening process that involves both algorithm-based screening and manual review. The company has also implemented a complaint procedure that enables it to identify inappropriate content utilizing its users’ feedback.

Algorithm-based Screening: The company applies algorithms to screen texts, as well as images and videos. The company’s system screens texts based on pre-set keywords, and it utilizes artificial intelligence to identify inappropriate images and videos. The screening system automatically declines content that does not meet the standards of its platform and flags suspicious content for manual review by its content management team.

Manual Review: The company’s content management team, which consisted of 314 employees as of December 31, 2021, was responsible for monitoring all information before delivery through its platform. The content management team reviews suspicious content identified in the algorithm-based screening process and makes the final decision as to whether to decline such content. Given the complexity and diversity of information submitted to the company’s platform, its content management team also reviews all content that has not earlier been flagged up in the algorithm-based screening process.

Complaint Procedure: A user may submit a complaint about a particular piece of content through the company’s mobile applications. The user is prompted to identify the basis for the complaint, such as duplication to pre-existing content, violation of law, factual mistake, low quality or plagiarism. The user also needs to provide a written commentary to support the complaint. The company removes the relevant content if it concludes that the complaint is valid. In addition, while the complaint is under review, the company may also temporarily block the relevant content from being further delivered until the company can investigate the complaint and reach a conclusion.

Content Recommendation

The company’s platform intelligently delivers personalized light entertainment content to users. The content recommendation process involves the following components.

Content Tagging: Each piece of content is multi-tagged with references to the key topics it covers. Before submitting an article or video, the content provider may provide up to six tags. Such tags range from general topics, such as ‘entertainment news’ to specific topics, such as the name of an actor. The company utilizes both algorithm-based screening and manual review to further refine such tags. Furthermore, the company’s smart video tagging technology uses deep learning that further increases the accuracy of tags. The company’s technology also automatically selects and displays to users the most appropriate ‘cover images’ for videos. The company’s technology has the potential to improve the quality and relevance of video content shown to its users, thereby enhancing user experience.

Interest and Social Graphs: Through an automated process, the company develops interests and social graphs for each user based on the user’s profile, behavior and social relationships. User profile data are provided by users when registering for an account on the company’s mobile applications. In addition, the company is able to gain a fairly accurate picture of a user’s profile, including age, gender and location, based on the user’s behavior on the company’s platform. The user’s behavior also provides the company with a granular view of the topics and content characteristics that could be of interest to the user. In addition, the interest and social graphs take into account the user’s social relationships with other users and such other users’ interests and behaviors.

Recommendation: The company’s content recommendation engine suggests content based on each user’s interest and social graphs, and continuously receives behavioral data inputs to update and refine its recommendations in real time to identify content that is most likely of interest to each user.

Monetization

The company places advertisements on its main pages, topic pages, and content pages. The company’s differentiated user base represents an attractive demographic target for businesses.

The company has an in-house advertising platform, which is a full-fledged and technology driven advertising exchange capable of not only monetizing internet media traffic generated in-house but also that originating from third-party media platforms, which operates a programmatic advertising system. It has a programmatic advertising system that utilizes a bidding system for advertising customers to bid for the targeted audience on mainly its platform. The company’s programmatic advertising system considers a wide range of parameters to determine which advertisement to show, including price bid, predicted click-through rate, predicted user conversion metrics and content relevance, to dynamically maximize the company’s advertising customer’s return on investment (ROI), as well as its revenue potential. The company’s advertising technology aims to maximize its revenue potential by rewarding the more relevant advertisement with a more prominent position, despite the potentially lower price bid of such advertisement. The company actively monitors the advertisements placed to help ensure their relevance.

Customers for the company’s programmatic advertising system are consisted of advertising agents and end advertisers. The company has its own sales personnel who are responsible for supporting and monitoring the performance of advertising agents and to attract advertising customers to use its programmatic advertising system directly. The company enters into standard agreements with advertising agents generally for a term of one year. The company’s advertising agents are responsible for identifying end advertisers, confirming payments and setting up accounts on its programmatic advertising system for advertising customers. The company provides ongoing training to advertising agents to help them become familiarized with the functionalities and capabilities of its programmatic advertising system. These advertising agents are responsible for collecting and submitting the relevant documentation and licenses from advertising customers for the company’s approval to open accounts on its programmatic advertising system; and are also liable for any infringement of third-party rights or violation of regulatory requirements caused by advertisements placed by their end advertisers.

Through collaboration with third-party merchandise suppliers, the company offers an online marketplace on Qutoutiao in which users merchandise offered by third-party merchandise suppliers. The company does not carry any inventory, and each merchandise supplier merchandise directly to users. A user pays the purchase price for a merchandise to the company. The company deducted its commission related to the merchandise and remit the remainder to the relevant merchandise supplier.

Technology

The company has focused on and will continue to invest in its technological infrastructure. Its business is supported by the following key technologies.

Interest and Social Graphing: Through an automated process, the company develops interest and social graphs for each user based on such person’s profile, behavior and social relationships. The company assesses the user’s desired content characteristics through technologies, including natural language processing, image analysis, and content tagging. The interest and social graphs take into account the user’s social relationships with other users and such other users’ interests, including their behaviors. The company continuously refines each user’s graphs based on the user’s behavior over time through artificial intelligence.

Content Recommendation Engine: The company’s content recommendation engine recommends content based on user behavior, and continuously receives behavioral data inputs to update and refine its recommendations in real time to identify content that is most likely of interest to each user. The company’s content recommendation engine is capable of processing large quantities of data, and can handle several billion inputs per day. In addition, new content is aggregated and recommended in real time from among millions of new contents added, reducing the possibility that the company’s users missing the content that may interest them when they next update their view in the company’s mobile application.

Advertising: The company’s advertising technology enables advertising customers to bid for audience and automatically deliver relevant, targeted promotional links to users. The company’s system rewards more relevant advertisements with more prominent positions, despite the potentially lower priced bids of such advertisements. The company’s audience segmentation technology helps ensure the relevance of advertisements shown to users by analyzing their interests through browsing activity, viewed content and commenting history. In addition, the company has the ability to predict click-through rates for advertisements using logistic regression, gradient boosting decision tree and linear and nonlinear modeling algorithms. Enhanced precision of these click-through rate projections can help maximize the cost effectiveness of customers’ advertising budgets. The company’s oCPC system further takes into consideration customers’ overall return requirements by assessing their desired end results in the context of their budgets and the availabilities of relevant advertising inventories, providing a holistic solution to their marketing needs rather than a simple product amid their marketing strategy.

Content Screening Technology: The company’s text screening system screens information based on pre-set keywords. The company utilizes artificial intelligence to identify inappropriate or objectionable content from images, speeches and videos, significantly increasing efficiency over manual review. The company also applies deep learning methods to analyze complex visual content. Through big data and continuous training, the company’s system is able to monitor and identify objectionable visual content with a high degree of accuracy. The screening system automatically declines content that does not meet the standards of the company’s platform and flags suspicious content for manual review by its content management team.

Fraud Detection: The company’s fraud detection technology focuses on ensuring that its user loyalty programs are not abused by fraudulent users or activities. The company’s anti-fraud system employs self-encoding and deep learning methods, such as sophisticated neural networks in analyzing users’ behaviors, as well as the locations of and the data from their devices to detect fraud. The company’s anti-fraud system assigns a fraud score to users based on the inputs it receives, with higher scores signifying a greater suspicion of fraudulent activity. The company also utilizes artificial intelligence methods, including decision trees, random decision forests, naive Bayes classifier, Gaussian mixed model, and logistic regression to maximize the predictive accuracy of the company’s anti-fraud system while minimizing the need for manual review.

Marketing and Promotion

The company grows its user base through advertisements and paid-for traffic on mobile channels. The company intends to continue to implement innovative and cost-effective marketing initiatives.

Intellectual Property

As of December 31, 2021, the company had registered 920 trademarks, including trademark for ‘Qutoutiao’, and owned 17 patents in the PRC. The company was the registered holder of 297 domain names in the PRC, and had been granted 495 software copyrights and 30 artwork copyrights as of December 31, 2021.

Regulations

The company engages in business activities that are value-added telecommunications services as defined and described by the Telecommunications Regulations of the PRC and the Catalogue of Telecommunications Business.

The company is required to obtain an Internet news license from CAC for the dissemination of news through the company’s mobile application.

As of December 31, 2021, the company completed the registration on the National Internet Audio-visual Platforms Information Management System with its Qutoutiao app.

The New Advertising Law, the SAIC Interim Measures and Draft Internet Advertising Measures require the company to monitor the advertising content shown on the company’s mobile applications to ensure that such content is true, accurate and in full compliance with applicable laws and regulations.

As of December 31, 2021, the company had registered 920 trademarks in the PRC, including the trademark for ‘Qutoutiao’, and filed 1,337 trademark applications in the PRC.

As of December 31, 2021, the company had owned 17 patents and filed 81 patent applications in the PRC.

As of December 31, 2021, the company had been granted 495 software copyrights and 30 artwork copyrights in the PRC.

As of December 31, 2021, the company was the registered holder of 297 domain names in the PRC.

Research and Development

The company’s research and development expenses were RMB551.6 million (U.S.$86.6 million) in 2021.

Competition

The company’s primary competitors include content aggregators, such as Jinritoutiao (operated by Bytedance), Kuaibao (operated by Tencent), and Yidianzixun (an affiliate of Phoenix News). To a lesser extent, the company also competes with mobile news portals, such as Tencent News, SINA News, Sohu News, NetEase News and Phoenix News. The company also competes with other mobile literature applications, such as iReader, QQ Reading, Qimao Free Novels and Fanqie Novels.

History

The company was founded in 2016. It was incorporated under Cayman Islands law in 2017. The company was formerly known as Qtech Ltd. and changed its name to Qutoutiao Inc. in 2018.