This article is from the WeChat public account: Yiyu Observation (ID: yiyuguancha), text/big entertainer.
When the short story adaptation of "Writing", a short story adaptation that broke the popularity record of Tencent Video's micro-short drama station, appeared, Zhihu breathed a sigh of relief, and the IP film and television of the online text that it had insisted on in the past two years finally found a "blowing" object.
I don't know when it began, "storytelling" and "engaging in online literature" have become the focus of Zhihu's publicity, and the paid novel business has become an urgent commercialization path under the leadership of the government overnight.
Led by paid online articles, various business attempts such as live broadcasting, knowledge payment, and e-commerce, Zhihu has opened the road of commercialization, and profitability has become the primary goal at present.
On March 26, Zhihu released its financial reports for the latest quarter and the whole year of 2023. According to the financial report, the revenue in the fourth quarter of 2023 was 1.14 billion yuan, a slight increase of 2.2% year-on-year, and the adjusted net loss was 91.3 million yuan, a year-on-year decrease of 31.9%.
Zhihu's 2023 financial report
Although the scale of losses continues to narrow, Zhihu's revenue growth continues to slow down, and the revenue growth in the fourth quarter of 2023 is the lowest since 2020.
In the past year, in order to boost its performance, Zhihu once again restructured its revenue composition, first merging its advertising and content commercialization business into the "marketing services" segment, and redefining paid membership and vocational training as new growth engines.
In fact, this is what Zhou Yuan, the founder of Zhihu, has repeatedly emphasized in the past year, in addition to the community, Zhihu also has two major brand businesses, "Yanyan Story" and "Zhihu Zhixuetang", the former is actually similar to the paid online reading service provided by China Literature, and the latter is adult vocational education and training. For Zhihu, it has become the norm to try different business models, and it seems to have become a "convention" to get used to chasing the wind and then give up quickly.
On the earnings call, Zhihu CEO Zhou Yuan and CFO Wang Han answered more about when they will be able to turn losses into profits. Zhou Yuan listed achieving profitability as the top strategy for 2024, and gave an expectation to achieve a single-quarter non-GAAP profit in the fourth quarter of this year.
Zhihu CEO Zhou Yuan
Judging from the current core business composition of Zhihu, it is not easy to achieve this.
On the one hand, advertising revenue continues to decline, on the other hand, whether it is online literature or adult education and training, its revenue ceiling is quite obvious, especially in the past year or so, Zhihu has placed high hopes on online articles and IP adaptation, there is still no breakthrough development.
In the era of AI, Zhihu has accumulated a huge Chinese corpus for more than ten years can and should actually discover greater value, not necessarily need to start from scratch to build a large model behind closed doors, look at the Reddit next door can earn $60 million a year just by providing training materials for Google's large model, Zhihu should rediscover its own value, rather than losing itself in blindly imitating and following.
From short stories to "small readings"
In 2023, Zhihu's revenue growth will show a quarter-by-quarter slowdown trend, with year-on-year growth rates of 33.8%, 24.9%, 12.1% and 2.2% in the four quarters, respectively. Since the completion of the dual listing in April 2022, Zhihu's revenue growth rate has gradually slowed down from the high growth rate in 2021, especially in the fourth quarter of 2023, when the growth rate dropped to only 2.2%.
From a full-year perspective, Zhihu's total revenue in 2023 was RMB4.2 billion, up 16.5% from RMB3.6 billion in 2022, but this growth rate was still lower than the previous two years. From 2021 to 2022, Zhihu's revenue was 2.96 billion yuan and 3.6 billion yuan, respectively, achieving growth of 118.9% and 21.8%, respectively.
Zhihu revenue trend
After business adjustment, Zhihu's main revenue source has become more prominent.
In terms of revenue composition, Zhihu reclassified its business segments in 2023, merging the revenue from advertising and content commercialization solutions into the marketing services segment, while continuing to develop paid membership and vocational training businesses. These changes reflect Zhihu's efforts to adapt to market changes and seek new growth points.
It also shows that in addition to advertising and other business segments, paid membership and vocational training business have become the main driving forces supporting Zhihu's revenue growth.
In terms of performance for the full year, the growth of paid membership and vocational training business was more significant. For the full year of 2023, the revenue from the paid membership business was RMB1.83 billion, up 48.4% year-on-year, and the revenue from the vocational training business was RMB570 million, up 127.8% year-on-year. In comparison, the revenue of the marketing services business in 2023 was RMB1.65 billion, down 15.5% year-on-year, and the revenue of other businesses was RMB150 million.
Zhihu's main business composition
In May 2023, it launched a new story brand, "Yanyan Story", an independent app that focuses on the original short story platform based on Zhihu's original story business. Although it is in the name of stories, the essence of this business is actually shorter online literature as a whole, and it is no wonder that some people will ridicule Zhihu as a new generation of "story clubs".
Of course, using online texts to attract users to pay is only the first step, and Zhihu actually attaches more importance to IP adaptation, wanting to increase the thickness of revenue. The IP of "Washing Lead Hua", which was born in Zhihu, has been adapted into the online drama "For the Dark Fragrance", and was successfully launched on the long video platform in October last year.
But there are very few such cases, especially on Zhihu, where the content of the long story line is not strong.
And at this moment, the outbreak of the short drama gave Zhihu another layer of hope, and settled on the short story. In terms of subject matter and content, the short stories on Zhihu are highly consistent with popular short dramas, such as rebirth, revenge, true and false identity, violent beating of scumbags, suspense reversal and other cool points, both of which correspond.
This aspect can indeed be said to be the outlet that Zhihu's online text business unintentionally bumped into, and therefore, on the road of online text IP film and television, it seems to be a good choice to target micro-short dramas with short stories, and the popularity of "Writing" has also carved a template for Zhihu.
Stills from "Writing".
However, at present, the production of short dramas that rely more on capital operation, how many are willing to spend extra money to buy online copyright adaptations are actually questionable, unless Zhihu personally participates in the adaptation, the vice president of Zhihu also said in his public relations speech at the beginning of this year that he will continue to add short story IP film and television investment, and launch the first batch of self-made short dramas within the year, but when the short drama is really launched, it is unknown how much the short drama competition that has become a red sea can give Zhihu, and in the absence of the advantage of the broadcast channel, it is quite difficult to simply attract users on the site to pay for the adapted short drama。
What's more, whether it is paid reading, the ceiling of the industry is already here, and China Literature, which exists in the form of a "Big Mac", has long stood here.
The success of China Literature is not only based on its years of hard work in the online literature market, but also on the support of Tencent's entire entertainment business system, whether it is the acquisition of Xinli or Tencent Animation, or the direct output of IP to Tencent Video, these are all difficult for Zhihu to obtain systematic resources.
If you want to become a "short drama industry reading article", you really have to weigh how much determination you can pay for entering the film and television ecology.
In the era of AI, Zhihu must find its own value anchor
With the popularity of large models, data has actually become one of several important AI resources in short supply.
AI large language models are built using text taken from the internet, including scientific research, news reports, and useful or unused line content from various websites. This material is broken down into lemmas, i.e., models that are used to learn how to construct words and parts of words that resemble human expressions.
For AI models, the larger and more complex the trained data, the stronger its capabilities, and OpenAI is a typical representative.
Reddit, a recently listed U.S. community site, gave an additional source of revenue in addition to advertising in its prospectus: selling content to AI companies that build ChatGPT-like chatbots to train AI.
Source: Pexels
How much revenue can this service bring, according to Reddit, its three-year contract with Google can earn more than $200 million in revenue.
You must know that this part of the income can be said to be "pure profit" in the real sense, because the content on Reddit is all UGC creation, and as a platform, all it has to do is open the data interface to Google, and it can lie down and collect money. And Reddit's revenue in 2023 will only be $800 million, and the net profit income of more than $60 million a year is undoubtedly a huge amount of money for Reddit.
This is the direction that Zhihu should actively explore.
After all, compared with text communities such as Tieba and Hupu, Zhihu originally had the best part of the corpus content on the Chinese Internet. In particular, its question-and-answer method that it focused on in the past is completely the best form of training AI.
Even though the content of Zhihu has been polarized and hydrated in the past few years, a lot of the precipitated content still has extremely high value. And there are still some users who are constantly updating high-quality content on Zhihu, and instead of constantly bothering these users with commercial advertisements and paid stories, Zhihu should actually rethink its position on the Chinese Internet as a community that originally wanted to provide high-quality content.
Source: Internet
For such a business form as an AI large model with multiple high thresholds from capital to technology to infrastructure, Zhihu, as a company that has not yet made a profit, does not need to follow the trend to chase the hot spot of the "large model" itself, but should seize this opportunity to actively connect with those large model companies with real technical potential, and use a large number of accumulated high-quality Chinese corpora as resources for commercial realization.
One of the challenges of a good AI large language model is to get information in a timely manner. At present, many large models are actually collected for free on websites such as Zhihu to train models through crawlers, but the greatest value of the community is that a large number of users are creating new content every day.
In addition, a reliable new source of information that distinguishes it from other AI-created content may increase in value over time, because if AI information is trained alone, the result is a tragedy caused by "inbreeding", and the phenomenon of fabrication will only get worse.
Source: Internet
For Zhihu, they have realized that expanding the scale of users will not bring them more revenue, but will continue to dilute the quality of content. But whether the entire community atmosphere can return to the original state is obviously a question mark, if you can seize the data dividend period brought by this wave of AI large models, so that Zhihu can achieve commercial realization without excessive harm to the user experience, then, after changing countless tracks, perhaps Zhihu needs more to return to the original intention of "there is a problem, go to Zhihu", but this time it is necessary to let the AI in the Chinese world give the best answer through Zhihu.