Volcano Video, Bytedance trying to do Kuaishou


The difference in origin between Kuaishou and Douyin comes from the difference in fault tolerance between single and double columns.


In Kuaishou’s dual-column display, the content has a title cover, which is selected by the user on their own initiative. Even if the user clicks the wrong button, the probability of mentally blaming the platform is much smaller. That is, the dual-column mode allows users to see more content on a single screen, and more traffic can be used for sample tests to discover user interests.


In Douyin’s single-column immersive underwriting logic, the content that users see is pushed to me by default, and users are more picky in single-column full-screen mode. They have psychological expectations of seeing high-quality content on Douyin. The way to use more traffic to try and error, the more time the product will push more certain content.


The result of the difference in traffic distribution caused by this interaction and fault tolerance is that the content of fast hands is richer. The innovative medium of short video and the characteristics of low threshold and strong content can accommodate most ordinary people’s daily life content. TikTok’s head is more concentrated, suitable for pushing some national content, and it is more likely to be liked by Internet celebrities, media and institutions. TikTok has relatively strong content, and Kuaishou’s community atmosphere and live broadcast are relatively strong, and competition has created different products.


There is no right or wrong for single-column and double-column. This UI difference is essentially a trade-off. If you want a single column, you want high-quality content and header content. You want to double column, that is, to be more personalized and diversified. When you choose a different route, it means you gain some benefits or bear some losses.


Kuaishou now has two core apps, a dual-rank master terminal and a single-rank ultra-fast version. The ultra-fast version is developing rapidly and has to bear 60 million of the Kuaishou’s 300 million DAU target.


In the Bytedance IES interactive entertainment system, Douyin is a single row, while the amount of Douyin’s super speed version is relatively small. The only dual-row products with large-scale users are the volcanoes that are benchmarked with Kuaishou.

  • (Of course, Volcano is not a double-column product in the complete sense, because the drop-down on Volcano’s video details page is to watch the next video instead of the comment area like Kuaishou, which may cause the inferences below the article to be wrong. And follow Kuaishou The difference between the speed version is that Volcano has its own user ecology. Strictly speaking, Bytedance does not have its own dual-line product.)

From the perspective of the competition between the two armies of Kuai Shao, the volcano video can be regarded as a new “Tik Tok Speed ​​Edition” against the Kuaishou Speed ​​Edition. Because the original Tik Tok Super Speed ​​Edition and Volcano Super Speed ​​Edition are similar in product form to the Kuaishou Super Speed ​​Edition, they are single-line short video products similar to Douyin. The TikTok/volcano speed version’s secondary stay/7 stay are much lower than the fast-hand speed version, and there is also a big gap in user activity indicators such as DAU/MAU.


The greater value of Kuaishou Speed ​​Edition to Kuaishou lies in helping Kuaishou understand the ecological differences of Shake Kuai more deeply, find ways to adapt to the community, and explore the advantages of Kuaishou. Starting from the original difference in single and double row fault tolerance, Kuaishou complements it. The chain of individual content consumption has been established, and at the same time, Shuanglei’s determination to continue to be a community resolutely is strengthened.


To deal with the fast-moving fast version, Douyin needs a different solution.


On January 8, the volcano video was renamed Douyin Volcano Edition, and a new icon was launched. Zhang Nan mentioned that the purpose of this brand integration upgrade is to optimize user experience. In the future, the content of Volcano and Douyin will gradually realize interoperability.


In summary, there are two points:

1. Open up the data on both sides, so that Douyin users can browse the works of Volcano, and Volcano users can browse the works of Douyin.

2. Open up the brand, let the two products use the same brand, hoping to use the Douyin brand to drive Volcano to gain new user recognition.

Shake Kuai has always been a gentleman’s competition on the bright side, there are backs and forth. In late June 2018, Kuaishou announced that the annual goal of the K3 campaign was 300 million. As a result, Douyin released 320 million DAU in early July. Now before the Spring Festival Gala, it announced that DAU has exceeded 400 million and the Douyin Volcano is going to the Kuaishou Express Edition. Version.


Product integration operations similar to the Douyin Volcano version have actually occurred overseas in ByteDance.


The parties concerned looked back and found that Musical.ly did two things wrong in 2016 before it was acquired: it was too late to return to China and the algorithm was too weak.


After Musical.ly was acquired by ByteDance, it took less than a month to quickly replicate the recommendation algorithm on the distributed machine learning platform of the joint strip and the core product indicators soared.


This shows that the key to making a short music video is:

1. Think of and agree with the idea of ​​short music videos before others

2. Certain product and interactive design capabilities, able to polish the prototype of this product

3. Efficient recommendation algorithm

4. Commercialization team + user growth


3 and 4 are the key elements to improve ROI. Musical.ly just finished 1 and 2, and it became a phenomenal product in Europe and America. And bytedance added 3 and 4, so you can use high ROI, run out of traffic to realize, and then further purchase traffic.


Since the acquisition of Musically at the end of 2017, ByteDance has begun to advance its merger plan. However, there is a transition period when the North American region still retains Musically, and other regions adopt Tiktok. Later Musically merged into Tiktok also from the front end to change the UI, but there are still two APP IDs.


ByteDance chose Douyin to replace Volcano and Kuaishou at the beginning of 2018. The core market hypothesis is: Douyin is in the city and Kuaishou is in the sinking market. Based on the same assumption, when ByteDance started its overseas short video development strategy in 2018: there is a dual structure of urban and rural areas and a huge rural market pushes volcano Vigo, and countries with huge cities push TikTok.


Vigo once topped the Google Store in Brazil, Indonesia, and Kuwait, Qatar, Saudi Arabia, and Morocco in the Middle East, and performed well in relatively sinking countries.


In 2019, Byte finally established TikTok’s global strategy, merged Vigo users into TikTok, and stopped the operation of Vigo products. On December 31, Vigo ceased operations in the Middle East.


However, Vigo did not merge with Tiktok at the APP level. Instead, starting from the spring of 2019, in countries where Tiktok and Vigo exist at the same time, Vigo will gradually weaken and strengthen TT. Tiktok will target Vigo users and make some targeting on the Vigo app. Here to induce users to download Tiktok.


The merger of overseas products may become a template for domestic integration.

In the short video market, ByteDance has adopted a very aggressive carpet covering method: watermelon video is benchmarked against YouTube, volcano video is benchmarked against Kuaishou, and Douyin is benchmarked against Musical.ly.


From 16-17, the daily activity of Kuaishou in short videos was much higher than that of Musical.ly. During the same period, Musical.ly encountered bottleneck growth in the United States. Therefore, for a long time, ByteDance has an impact on the volcano’s resources/talents. Investment is far more than Douyin. Volcano is a strategic product implemented by the serious headline play. Even the advertising title for the Spring Festival Gala originally expected in 2018 is Volcano instead of Douyin.


From the beginning, Douyin’s product was human intervention, and Volcano is a very data science end to end.


But the volcano is a highly similar product to Kuaishou, with basically the same products and positioning, and it is destined to only take the rest of the Kuaishou market. In 2018, Hunan Satellite TV’s New Year’s Eve party was still named after the volcano. I originally hoped to bring the volcano back when the urban workers returned home, but it was actually not established. Half of the people in his home family may already be playing Kuaishou. A lot of money, but never see the opportunity to catch up with the fast hand.


The fast-growing product that seizes the opportunity of connecting the people across the country during the Spring Festival is Douyin. During the 2018 Spring Festival holiday, Douyin increased its daily activities by 10 million, from 30 million DAU to 100 million DAU in 4 months. Since May, it has become the number one short video track.

The rapid growth of Douyin has changed Byte’s corporate strategy, and Volcano has shifted from offensive positioning to strategic defense to hold back Kuaishou. Since March 18, the volcano’s firepower input has plummeted. In April, it encountered regulatory problems and was removed from the shelves with Kuaishou. After reaching the peak of nearly 60 million DAU, the volcano’s DAU traded sideways at around 50 million DAU for the entire 18 years, and gradually entered a slow decline channel. In March 2019, the official announcement of daily life was 50 million +.


Commercial monetization has become the core of volcano’s small video at this stage. From the two dimensions of live broadcast and advertising, live broadcast income has risen sharply, becoming the product with the highest live broadcast income in the byte system, and a concept of “new line crowd” is proposed in commercialization. , That is, an incremental market with a wide range of consumption fields, strong consumer demand and great potential, and through a series of data to promote the high proportion of volcanoes in the “new line crowd”, and successfully established benchmark commercialization cases such as KFC.


DAU has not risen for a long time, and the business direction is lack of imagination. Development resources and manpower have also begun to turn to Douyin and Tiktok. By 2019, Volcano will begin to further serve the strategic goals of IES.


In order to capture the fast-growing live broadcast revenue of Kuaishou, ByteDance began to prepare for the live broadcast platform, and merged and merged the live broadcast technology and operation teams of the three products of Volcano Video, Douyin Video and Watermelon Video to form a new “Live Broadcast Business Center” “, supports all ByteDance’s businesses. Among the three products, the live broadcast business of Volcano Video is the most mature, which is equivalent to that the soldiers trained by Volcano are now reused in the entire system. The head of Volcano Video is Han Shangyou as the head of the live broadcast center.


Regarding this integration, I quote a former colleague from the volcano:

  • If I were the boss, I would think the first is whether the user overlap is OK; the second is whether the expectations of the volcano project have been met, which is obviously not achieved now; the third is whether the volcano is likely to continue to grow, now it depends on the problem Serious; the fourth is that Volcano’s profitable business (live broadcasting) has also merged with Douyin. The fifth is that after the merger, it can also reduce internal struggles and work together to compete with Kuaishou.

The combination of shaking fire can increase the recommendation pool and allow the interoperability of technical capabilities.


Because recommendation is unified, content is inherently vertical and fragmented, and vertical content consumption can be connected to a very mature recommendation system at a lower cost. The transformation of this algorithm is costly (the reuse framework is very Fast) and high returns (the framework is very mature).


When ByteDance tested the app to get through the project, it tried to add Douyin content to Volcano, and the benefits were obvious. However, if the content of Volcano entered Douyin, the indicator fell more miserably, making the Volcano content into Douyin but failed to push it down.


In theory, if the product expands the content pool, the effect will rise, and the friction may be the algorithm’s uncoordinated adaptation. At that time, the unsatisfactory result of Volcano content entering Douyin was related to Douyin’s user structure and product style control at the time, but today’s headline small video, which showed a single row of results similar to the Volcano UI, introduced small videos of Volcano and Douyin. It can directly help Today’s Toutiao increase traffic by a few points, and at the same time, Douyin Volcano’s own product content is also evolving. The commercial help brought by the small video on the main end of Toutiao is even better than the introduction of watermelon video content.


The product integration of Douyin and Volcano depends on whether the algorithm can be compatible with the user tastes of the two products. The practical level depends on how the algorithm and data are adjusted. In the case of algorithm compatibility, merge producers and consumers, and after product integration, there will be more collaborative data for content production and consumption. The business model of content has scale effects. For the same short video category that is partial to content consumption, the stronger the scale effect, the better the head product data, and the more difficult it is for competing products.


The scale effect coefficient is related to the factors of production that the platform can monopolize, that is, what you have is not available elsewhere. The more creative UGC community products are mainly monopolized on the creative side, and waist users who prop up the consumer market will basically not carry their own content everywhere. In terms of information flow products that are more distributed, self-media is a source of water for the world, and the core creative ability that the platform can monopolize is insufficient.


Therefore, it is easy to make mobile information flow products, while Tencent’s short video micro-vision is difficult to make, but once short video products are made, there will be a high threshold and a moat.