ByteDance made two mistakes

Text|Cattle knife finance Wang Hailong.

Now, Byte's revenue has surpassed Tencent's, and this sentence is given to Byte. Byte missed two major technical directions, one is SORA and the other is Vision. Since the New Year, these two major technical directions are setting off a technology boom, and unfortunately Byte is not doing well enough in both directions.

In the past three years, the momentum of byte commercialization has surpassed technological innovation. In the past, Tencent was always scolded for plagiarism (not innovating), but now Byte is not the same - business expansion has no borders, relying on Douyin's traffic to expand in e-commerce, local life, tourism, real estate, automobiles and other fields, and the power of commercialization far exceeds technological innovation.

You must know that whether it is Toutiao, Douyin, or TikTok Tok, there is a powerful algorithm support behind Bytes, and it can be said that algorithms are the core competitiveness of Bytes.

Behind the sora is also the algorithm, which should be the most powerful byte, but it has been slow and poor in the field of Wensheng Wen and Wensheng graphs, until Wensheng ** sora came out, and ** began to involve the core business of bytes.

Another area of VR MR, it can only be said that Byte is even more regrettable.

The release of Apple's Vision caused a sensation, and Byte pushed it more than two years in advance through PICO, occupying the opportunity and not taking the high ground. Sora and Vision, one is the content side and the other is the device side, why does Byte lag behind in the two areas involving the most core business?

At the end of 2023, Liang Rubo, the current CEO of ByteDance, mentioned the sense of crisis, mentioning the diseases and efficiency of large companies.

In fact, all this has begun in 2021, and everything has been paved for everything, but now Byte's sense of crisis has "come to the face".

01 PICO fell on the eve of Vision, and the regret of Byte on VR was that he chose the right time, but he didn't do it right.

Just two months before the official release of Apple's Vision, Byte PICO began to lay off the team on a large scale, and in November 2023, Byte laid off the PICO team on a large scale, which means that Byte's VR exploration failed.

A new wave of tech may be kicking off, and the trigger is Apple's Vision. At the beginning of February, the pre-sale of Vision was sold out, and the data showed that it sold 16-180,000 units in a weekend, and soon appeared on social ** with various VISION use**.

PICO, which began to lay out a big push two years ago, does not have such a good life as vision. According to reports, the final sales of the PICO 4 were less than 300,000 units, and the overall sales of the PICO were less than 600,000 units, which is quite short of the target of 1 million. In the past five years, VR has experienced two start-up booms and two bubble bursts.

In the last technology cycle, for example, Xiaomi's success lies in entering the game around 2012 no sooner or later, and entering the game too early and too late will miss opportunities, and entering the market too early means bearing the cost of market education.

Byte also accurately grasped the timing, and a series of actions by Meta, an important benchmark object, as well as the shipment of the core product Quest2, all marked the arrival of the third VR boom.

In September 2020, Meta officially launched Quest2, and in 2021, Facebook officially changed its name to Meta.

It is the Quest2 that has brought about changes in the VR market, and in 2021, Meta's Quest shipments will reach the 10 million level, which is considered to be the threshold for a technology product to move from niche to popular. Byte's series of operations on PICO have a similar layout rhythm to Meta, but slightly later.

In August 2021, ByteDance acquired PICO, and a year later, it released a high-profile new consumer VR product, PICO 4 Pro, in an attempt to occupy a position in the new round of VR competition.

It can be said that bytes are a copy of meta's play. For example, through investment and acquisition, we can quickly enter the hardware field, and support content manufacturers such as games through acquisitions to improve the ecosystem. PICO and Quest are very similar in terms of price, quality, ecology and other aspects.

For example, the highlight of Quest is "Half-Life: Alyx", and Byte spent a huge amount of money to win the broadcast rights of the World Cup during the key promotion period of PICO. But PICO's failure also lies in the over-imitation of the Quest.

At the beginning of the launch of PICO 4, the author was in a PICO experience store, and the clerk also admitted that the technical level was on par with the Quest2, and there was not much difference between the pixels and the content of the Quest 2. Byte's lack of hardware is one of the reasons why PICO fails. According to reports, in the follow-up development of PICO, there is a conflict between the byte faction and the PICO faction in the team run-in, and Byte believes in vigorously producing miracles, using the Internet style of play to advance quickly and iterate in small steps, which is contrary to PICO's original steady and steady approach.

PICO is one of the highlights of Liang Rubo's succession as Zhang Yiming, but the two seem to have different views on this project.

According to a report by Yiou, citing people familiar with the matter, the head of PICO went to Singapore to report to Zhang Yiming, and the feedback was that Byte would gradually abandon the PICO business, because the hardware field in which PICO is located is not what ByteDance is good at, and the results have not met expectations in the past few years, and there is no hope for the future.

02 Byte's "Sora", which lagged behind the resignation of Douyin CEO Zhang Nan, brought more conjecture to Byte's other core technology direction.

At the beginning of February, Zhang Nan, the head of Douyin, announced his resignation as CEO of Douyin Group, and the core direction of the future is to focus on the cutting and screening business, and lead the team to seek a breakthrough in AI-assisted creation, and will soon launch an AI picture and ** product.

Recently, it has been rumored that Byte has been developing a "Chinese version of Sora" - an innovative ** model boximator before the emergence of Sora.

Byte's response is that it cannot be implemented as a perfect product at present, and there is still a big gap between it and the leading foreign generation model in terms of picture quality, fidelity, and duration. The algorithm and ** are the world's top bytes, and they are lagging behind in the field of Wensheng**. In fact, this is caused by Bytes' "indifference" to the large model.

Bytes have been slow to make progress with large models.

In the past year, technology giants have taken turns to launch large models, but Byte is late, and in the ranking of domestic large models, Byte has not been able to get a high position.

In fact, the entire byte is not sharp enough for large models.

Liang Rubo mentioned in the year-end summary that Byte's large companies are seriously ill, including himself, and his sensitivity to new things has declined, and he will not pay attention to GPT until 2023 in the internal technical report, while the earlier ones in the industry will start to invest from 2018 to 2020. The release of SORA was a wake-up call for bytes.

The market began to worry about whether the ability of Wensheng represented by SORA would have an impact on the short ** represented by Douyin.

Why Douyin's advantages in the field of ** have not been translated into technical capabilities, you know, Douyin began to promote **search as early as 2019, and the relationship between text, pictures, and ** is constantly strengthening.

In addition, Douyin's algorithm itself is a strong point, but it has not accumulated in the fields of Wensheng**. An important factor is that Douyin is busy commercializing.

From 2021 to 2023, Douyin's monetization rate will continue to increase in the fields of live streaming and **. An intuitive feeling is that the proportion of ads on Douyin has also increased, and at the same time, there are also information such as ** and bringing goods in non-advertising. Bytes attach much more importance to commercialization than to technology.

In the 2023 open letter, Zhang Yiming mentioned that ByteDance could not miss AGI (General Artificial Intelligence), and pointed out that AGI is an indispensable partner for Douyin and TikTok to discover new growth opportunities in the world, "It can solve ByteDance's second curve growth dilemma." ”

In fact, Douyin is currently facing a growth bottleneck.

As a blockbuster product that once swept the information industry, Toutiao has also faced the problem of growth bottlenecks in the transformation of information flow on various platforms.

The development of Internet products is so fast, and today's young people may be exposed to short** longer than information flow products, forgetting the 700 million monthly live Toutiao today, but at that time, the products were in full swing, even if they were industry practitioners, they were still shocked to see this data at that time.

There has never been a product with more than 1 billion monthly active users, except for WeChat, but in fact, in addition to WeChat, Toutiao has never had a product with more than 700 million monthly active users before. Look at it now, less than 5 years, Toutiao is already a previous generation product.

At present, Douyin's traffic growth has gradually peaked, but it has begun to exert its strength through the short-term transformation of ** account, Xiaohongshu, etc., and there is a trend of being on an equal footing with Douyin.

03 Bytebytes, who do not "fight", do suffer from the disease of large companies, and Liang Rubo's conclusion is that Bytes began to be inefficient, sluggish, and have low standards.

He gave an example, a simple requirement of an internal system, pulled a bunch of people, at the beginning of the assessment to 1000 man-days, that is, 1 person to do 1000 days, after in-depth questioning, it was found that in fact, some people misunderstood, and finally found that only 1 person to do 15 days can be completed. After an employee left to join a start-up company, he told a colleague he knew that he had worked in the start-up company for 6 months in 1 month. ”

From 2019 to 2021, it expanded from 10,000 to 100,000 people, and the sheer size of the organization brought about a reduction in efficiency.

But from another point of view, the byte business layout also brings inefficiencies. One is that in the market, the bytes that don't fight wars are becoming less efficient.

In 2020, Byte basically laid a key battle, and the gap between Douyin and Kuaishou laid a decisive gap this year, and the pattern of the two was basically fixed. Since then, Tencent Microvision, etc., which have entered the game one after another, have no opponents in Douyin before the launch of the WeChat ** account, including for a period of time after the launch. On the other hand, the diversified layout of bytes has led to the dispersion of combat power.

In the past few years from 2021 to 2023, Douyin has successively deployed in Feishu, PICO, and Volcano Engine, resulting in the dispersion of troops. If bytes are still a miracle, then now they are scattered in three areas.

At the end of 2022, also at the all-staff meeting, Liang Rubo mentioned that due to the slowdown in ByteDance's revenue growth in 2022, DAU growth did not meet expectations, so it will continue to "lose weight and gain weight", "significantly reduce the 2022-2023 HC plan" and "take stock of retreats and redundant organizations".

For a whole year, the market has continued to spread the news of Byte** Mutong Technology, and Byte has made cuts in games, VR and even Feishu and other projects.

With more than 100,000 employees, there is no way to avoid the disease technology of large companies.

Previously, Tencent and other Internet giants of the previous generation were also troubled by the illness of large companies. At the same time, Ma Huateng and other Internet founders of the older generation are always vigilant.

Shen Nanpeng, an investor in Bytes, once said, "If you lose completely in your own turf and in the field you are familiar with, this is a huge mistake." ”

By this standard, Byte has already made two mistakes.