B station "horizontally"

Bilibili changes the playback volume indicator, and long videos are more profitable. Bilibili will use playback time instead of playback volume as the main indicator affecting content recommendation and monetization. The implementation details of the change, user consumption habits, and advertiser acceptance will all affect its significance. Bilibili still relies on short videos to provide traffic, and how to balance the two in the future remains a difficult problem. This article explains the significance of Bilibili's move and the problems that exist, and is recommended for students who are interested in Bilibili's operations.

Bilibili is about to undergo an important adjustment.

At Bilibili’s 14th anniversary celebration on June 26, CEO Chen Rui announced that in the coming weeks the number of playback minutes will replace the current number of external playbacks.

After the revision, the position showing the number of views in the video will be replaced by the playback time, which will serve as an important indicator that affects users' judgment of content quality. The statistics of playback time will be affected by the speed multiple, and the change will not affect popular recommendations.

"Compared to the number of views, the time users actually spend watching the video is a more valuable reference data." Chen Rui said in his speech, "As long as your content is good, the number of minutes of viewing can better reflect the quality of the content than the number of views."

On the surface, it was just a change to a video’s display data, but the announcement was like a stone that stirred up a thousand waves. Under the related hot searches, UP hosts excitedly expressed that the disadvantage of long videos over short videos in the past has finally improved; users began to nominate their "favorite" UP hosts and tracks: "Tongliao Khan is going crazy" and "ASMR won."

Using playback time as an indicator to measure users' content consumption means that Bilibili has begun to rethink its community ecology and business logic, and to promote long videos to strengthen the platform's content barriers.

One year after being “impacted” by short videos, can medium and long videos “change their fate” as they wish?

1. Dunhao’s short videos “incited public outrage”, while its long videos “went to the dark alleys”

In the past two years, Bilibili seems to have fallen into some kind of predicament, and has been swept into the whirlpool of public opinion without paying attention. Interestingly, the reactions of users and creators to this change were surprisingly consistent, and almost all of them gave positive feedback - which is good , which to some extent reflects the users' "resentment" towards short videos.

Since the introduction of story-mode, users and even creators of Bilibili have been arguing over the differences between "horizontal and vertical" and "long and short", and the community's native ecology has also been fiercely rejecting short videos. In the first half of this year, the reduction of creator incentives once again became a hot topic on Weibo, and the "stop updating trend" followed closely, which was also mixed with dissatisfaction with the "screen dominance" of short videos. Many top UP hosts expressed concerns about the future of Bilibili.

B station UP master @-LKs-

After the "stop updating trend" was over, the top UP host @智能路闸 released a 50-minute oral video, pointing out that "Bilibili is not Douyin". This is not the first time he has expressed similar views. In the year-end summary video at the end of 2022, he also mentioned the dilemma of UP hosts, "The number of views of long videos will never compete with short videos" and "UP hosts are all numb. No matter how hard they try, they can't beat the short video bloggers who easily get millions of views."

B station UP host @Smart Roadblock

@鴻诗人劇基德, who has 3.57 million followers, expressed indignation in a post in May this year, "I really suggest that Bilibili separate long and short videos. Put the short video entrance in the second menu below the mobile version, move the dynamics to the third position, and put the membership purchase into 'My', and it will take off immediately." There was a chorus of agreement in the comment section.

Short videos have been criticized a lot. The core issue is not the difference between long and short, vertical and horizontal screens, but it points more to the fact that the traffic competition between short videos and long videos is not at the same starting line, and that low-quality short videos disrupt the platform community ecology.

Due to the advantages of rich viewing scenarios, low trial and error costs, and "short, flat, and fast" that make people addicted, short videos can always quickly become popular with high playback volume and completion rate, which is an advantage that long videos do not have. Knowledge area UP host @智能路闸 For example, the average playback time of his 45-minute video is 10 minutes, and an hour is only 15 minutes.

In addition, as Chen Rui demonstrated in his speech, under the premise of measuring content quality by playback volume, low-quality content can also attract users' attention with gimmicky titles and covers, and obtain 700,000 playback data. For most creators, this number is already unattainable.

In terms of production costs, the high production costs of long videos also put UP hosts in the dilemma of "gambling on traffic". @智能路闸 introduced himself, "I wrote a total of 34,000 words for a Nintendo video. I read three books before writing it. In the end, I changed the copy, added some humorous paragraphs, recorded audio, and finally handed it over to the editor, which took another 15 days to edit."

Similar dilemmas are not uncommon among creators. Bilibili was eager to introduce vertical short videos, but did not properly handle the relationship between the two video modes. In the past year, short videos have squeezed the living space of long video creators, both in terms of traffic and commercial realization.

2. Bilibili “picks up” long videos?

Compared with the previous "judging a video by its playback volume", using playback minutes instead of playback volume as the external indicator of the video will, to a certain extent, bring long videos and short videos to the same level.

As users have complained, "If you want to watch short videos, why come to Bilibili?" Compared with Douyin and Kuaishou, Bilibili's advantage has always been long videos. Stabilizing the long video ecosystem is also its way of gaining a differentiated advantage in the situation of being "surrounded by wolves".

On Weibo, @卢诗翰 said, "The number of minutes played essentially reflects the time users actually spend watching the video. It can focus on the content more accurately and is indeed more suitable for the B station ecosystem than the number of views."

It is worth noting that Bilibili has not announced the weight of playback minutes in the algorithm recommendation. At present, it seems that changes in the video's external data have more impact on users' judgment criteria for content quality.

So, which type of videos are more likely to benefit under the new indicators? Kas believes that the answer may be videos with large amounts of information, high information density, and high replay rates, that is, videos with a good sense of rhythm and that can bring users a sense of value and gain. For example, game guides and teaching videos of niche cultural circles have tool value, and because of their high replay rates, they have a certain degree of potential to break through the circle and win recommendations.

Taking live game videos as an example, in the past, users' viewing behavior made limited contribution to the playback data, but their actual time consumed in the video was extremely high; knowledge courses, for example, have high replay value because they cannot be played in one go and have a lot of content that needs to be watched repeatedly to digest, and they also gain significant benefits under the new indicators; some film and television UP hosts, such as @木鱼水心, also gain continuous attention from users due to the high narrative level of their videos.

Previously, among the 10 videos with the longest cumulative viewing time released by Bilibili, @宋浩老师's "Advanced Mathematics" was far ahead. Among the top ten, knowledge course videos occupied 7 seats, and the rest were e-sports events and game strategy videos.

Image source: Station B

In addition, videos that can arouse the interest and resonance of more people can also obtain higher playback time due to high playback volume even if the replay rate is low. For some vertical videos with a fixed audience, the consumption time contributed by core fans may bring more opportunities for the video to expand its circle.

At the same time, the frequency of recommendations for some low-quality marketing accounts and clickbait videos will be greatly reduced, while some live-streaming slices and water videos will face an unfavorable situation due to their low replay rate.

Xiao A, who registered as a member of Bilibili in 2015, has experienced many revisions and adjustments of the platform, but still maintains the habit of browsing Bilibili for 3 or 4 hours a day and is a heavy user of the platform. In Xiao A's view, within the Bilibili ecosystem, the playback time has another specific benefit, that is, it can protect high-quality videos within a small circle. Xiao A told Kas Data, "Bilibili may return to the early form of browsing the homepage to find treasures that break the circle, and each circle gradually strengthens but still has small-scale exchanges."

This may be the ideal state that the original users of Bilibili yearn for. However, how it will work in reality is still unknown. Although Bilibili said that the revision will be completed in early July, as of press time, Bilibili has not made any adjustments.

3. Long video spring "Tathagata"

Among the popular memes promoted by Bilibili, "Tathagata" is a fashionable abstract "new top streamer".

Professor Guo Jicheng gave a magical interpretation in the video, "Tathagata, has he come? Tathagata. Has he really come? Tathagata. Has he come or not? Tathagata." The subtle semantics triggered a large number of secondary creations.

Under the semantics of this meme, this state of "seemingly coming, but also seemingly not coming" is also very suitable to describe the "spring of long videos" that Bilibili users are looking forward to.

After going through the period of "generating content for love", UP hosts have gradually become full-time and team-based, and the driving force behind making videos is to be able to "make a living".

2023 is already halfway through, and there is less than a year and a half left for B Station to achieve its goal of "breaking even in 2024". According to its financial report, the total revenue in the first quarter of 2023 was 5.07 billion, of which the value-added service business revenue was 2.16 billion, accounting for 43%, and the advertising business revenue was 1.27 billion, accounting for only 25%. To truly "solve" the dilemma of long videos, B Station needs to increase its advertising business and find other replicable monetization strategies.

According to the data released by Bilibili, PUGC (professional user-generated content) videos are still its foundation, accounting for 70% of the total playback volume, with a year-on-year growth rate of more than 26%. Whether it is to stabilize the community ecology or strengthen the differentiated advantages, it is necessary to "give a hand" to long videos and let the UP owners of long videos "eat food".

From this perspective, the change of using playback time instead of playback volume can be said to have increased the commercial competitiveness of UP hosts. On the one hand, it can add another dimension to measure the quality of advertisements and the effect of user reach, and on the other hand, it emphasizes that "user viewing behavior on Bilibili is more valuable" and strengthens the effectiveness of advertising.

"Bilibili is rectifying the pricing power for UP hosts," @岚夕 said on Weibo. "The cost of understanding for advertisers is still too high. When the number of views is listed in the report, it is easy for UP hosts to suffer a loss. If they make heavier content, they will get lower evaluations, which will only make them lose competitiveness in the long run."

However, this change only provides a stepping stone for UP hosts to realize their monetization potential. In an environment where brand advertising budgets are shrinking and the "ROI theory" is dominant, the commercialization of Bilibili needs to gain more competitive advantages before advertisers can accept this set of rules.

Bilibili has also taken corresponding actions: in an internal email at the end of June, Bilibili will integrate multiple teams to establish a new first-level department, the "Transaction Ecology Center", to serve the business needs of users and UP masters.

On the other hand, there are also many problems with the revision. Some people are worried about whether some high-cost but short-playing videos, such as MAD videos, will be affected. Some people also question how to control the behavior of brushing the playing time under the new standard, and whether the playing time will be included in the algorithm recommendation, rather than a trick to "make the data look better."

After more than a year of rapid development of vertical short videos, it is impossible for Bilibili to give up short videos, a fertile land of traffic. In the first quarter of 2023 financial report, the average daily playback volume of vertical videos increased by 82% year-on-year. The traffic and advertising brought by the short video model to Bilibili are still what Bilibili cannot lose. So how to balance the proportion of long and short video recommendations in the future without losing one is another difficult problem facing Bilibili.

Xiao B, another deep user of Bilibili, is concerned about whether using playback time as an explicit indicator will challenge users' previous content consumption habits. He believes that before viewers become familiar with the existing indicators, they will most likely continue to use playback volume as an anchor point and convert playback time to determine the quality of videos in a specific track.

Xiao A has a different view on this, "Users will still subconsciously choose the video with the highest data." Compared with user content consumption habits, he is more worried about technical issues. As Chen Rui said in his anniversary speech, counting the number of minutes played means that user behaviors such as speed, multi-device playback, and dragging the progress bar must also be taken into account, which is much more difficult than the number of plays.

In short, there is no final conclusion before it is "implemented".

Author: Lin Xuan

Source: WeChat public account "CaasData (ID: caasdata6)"