Shulou(Shulou.com)11/24 Report--

On July 20, Beijing time, Tesla released earnings data for the second quarter of this year.

After a year-long "global price war," Tesla's gross margin has fallen from 25% a year ago to 18.2% in Q2 this year, squeezing profitability sharply, but Musk still seems confident.

Musk said at Tesla's Q2 earnings call that he would continue to cut prices during "turbulent times." Musk bluntly said it was reasonable to sacrifice vehicle profit margins for more production, and he also proposed to "raise FSD prices."

Taken together, it may not be cars that help Tesla make money in the future, but autonomous driving software. Musk firmly believes that FSD (Full Self-Driving) software will be an important way to increase Tesla's valuation by 5 to 10 times in the future. At this year's 2023 World Artificial Intelligence Conference, he made it clear that autonomous driving technology will increase car usage, and Tesla is expected to achieve L4-L5 autonomous driving later this year. It is understood that FSD software currently sells for $15000 in the United States, and this number will rise further in the future.

And Musk has always said Tesla can sell cars at zero profit. But is this reliable? What did Tesla do right when it came to software-defined cars? Can software make money?

1. Selling cars is better than selling software? For Tesla, although gross margins have fallen, its overall single-vehicle profit is still at a high level. Frequent price fluctuations are also due to their high bicycle profits. Earlier at the China Auto Forum, Xu Changming, deputy director of the State Information Center, pointed out that BYD's single car profit was only 8854 yuan, while Tesla's single car profit reached $10426, 8.5 times that of BYD. In other words, Tesla sells one unit, BYD sells 8.5 units to catch up. In addition to profit growth from car sales, Tesla's software profitability is also worth learning from BYD.

In 2012 Tesla delivered the world's first truly smart car, the Model S. When it comes to Model S, many old car owners will first think of the 17-inch central control large screen, full LCD dashboard, touch screen operation... These have subverted the previous understanding of the car interior, which depends on its software system Version.

Tesla's software system Version is deeply modified based on the Linux kernel, enabling it to achieve OTA rapid correction and upgrade, opening a new era of "software-defined cars" and providing the possibility for software subscription.

After ten years of exploration, Tesla has formed Autopilot FSD fully autonomous driving option package, OTA paid upgrade and advanced car networking function three software subscription services, which have become important profit growth points.

First, Tesla makes software profitable through the Autopilot Suite option. Tesla's different levels of autonomous driving have been developed and improved over the years and are now quite mature. However, these features are not activated by default, users have to pay for subscriptions to activate these features, and as the number of features increases, subscription prices continue to climb.

In April 2022, Musk said in an interview that the number of test users of FSD software had exceeded 100,000. By September of that year, the number had exceeded 160,000. Crucially, FSD software has begun charging, and FSD is currently priced at $15,000 in North America or $199/$99 per month (monthly subscription).

That's about the price of a mid-range car. This means Tesla can make a profit equivalent to selling a car by selling software. That's what Musk said,"In the long run, autonomous driving revenue will make all these numbers (traditional car sales profits) look stupid." "

Second, Tesla changes vehicle performance and configuration through OTA paid upgrades. Users can pay for upgrades to get better acceleration, endurance and comfort profiles. For example, in 2019, Tesla launched an "acceleration performance enhancement package" that can further release the redundancy of the Model 3 all-wheel drive dual-motor through OTA, reducing 100-kilometer acceleration from 4.4 seconds to 3.9 seconds for users to pay $2000.

In addition, it is the advanced car networking function in the cockpit. Tesla allows users to upgrade MCU 1.0 to MCU 2.0 for a fee to get smoother, richer car apps.

In short, Tesla can achieve higher profits, in addition to high-end vehicle positioning and higher prices, its innovative software sales model is the key. Three years ago, 5% of Tesla's revenue structure came from innovation, the software business, and that number could rise to 30% by 2025. What did Tesla do right compared to other car companies? Synthesizing the views of experts at the 15th China Auto Blue Book Forum, there are the following points:

1. Independent and clear strategic positioning and top-level design

Hu Yongli, general manager of Lianyou Technology, believes that Tesla has planned its business model for the next four or five years in advance when making enterprise strategy, including some software and hardware integration functions, etc., and has already had a layout in the strategic system. Other companies may need to think more strategically as they learn and follow through.

2, the market reaction speed is fast and strong

Black Sesame CMO Yang Yuxin believes that many new features are Tesla's first to market, through user feedback, continuous improvement and expansion in areas where there is room for development. "After Tesla gets these two things right, it can really collect money with its new app or software. In this process, we attract talents to do new things and innovate step by step through word of mouth. "

3. The relationship between the original automobile enterprise and the user has been converted from a one-time system to a lifetime system.

From the traditional car profit model, when the car is sold, the profit basically ends. But in the era of software-defined cars, Musk has successfully integrated the relationship between car companies and users throughout the entire cycle of the car, even during use. A virtuous circle of "users continue to provide data and manufacturers continue to expand services" has been formed.

Tesla's success has also inspired other car companies to step up research and innovation to meet consumer demand for advanced software features. However, OEMs who want to make profits through software like Tesla need to have many qualities, such as vehicle OTA capability and embedded hardware capability.

OTA of the whole vehicle can constantly bring fresh experience to users. On the basis of the current hardware, the freshness cycle of automobile products can be increased to the maximum extent through software, so that it has the ability of continuous improvement. Of course, the complete OTA process should also cooperate with the advanced electronic and electrical architecture of the whole vehicle.

Hardware embedding with sufficient capacity is also very important. Take Tesla as an example. Although the model was delivered with version 2.0 hardware, the function was unusable. With the change of software, AP HW 2.5 and AP HW 3.0 were gradually introduced. Not only automatic driving, but also other functions, such as battery, seat heating and other hardware can be unlocked through background software.

3. Is software reliable for making money? Ouyang Minggao, academician of Chinese Academy of Sciences and vice chairman of China Electric Vehicle 100 People's Committee, said that traditional fuel vehicles can rely on brand and supply chain to practice value recovery, but new energy vehicles "may be difficult to realize by these." He pointed out that "the great value of new energy vehicles is to be generated in the later services, so this gives us a very new topic, that is, how to tap the value of services from an intelligent perspective." "

McKinsey China Auto Consumer Insight 2023 also confirms this view. Among the reasons consumers consider buying electric vehicles, 60 percent of respondents chose "high intelligence" in 2022, up from 44 percent in 2019 and 45 percent in 2021, according to the report.

According to relevant forecast data, the global automotive software market size in 2018 is only 6.54 billion US dollars, and it is expected to exceed 35 billion US dollars by 2025, when the code volume of a smart car is expected to reach 700 million lines, an increase of about 2.3 times compared with 2022.

But software will not disrupt the auto industry overnight.

During the China Electric Vehicle 100 Forum (2023), Cao Bin, general manager of Neusoft Ruichi, said in an exclusive interview with many authoritative media that under the trend of "software-defined cars," the automobile industry is facing the most dramatic period of change since its birth. In the future, the coupling between automotive IT and the Internet will lead to business model innovation.

Cao Bin, general manager of Dongsoft Ruichi, said again at the 2023 China Auto Forum that with the continuous improvement of the importance of software in the whole vehicle, it will gradually become an important direction of the "inner volume" of the automobile industry. The "software-defined automobile" will present the stepped landing characteristics:

The first phase is to build intelligent infrastructure to enable the possibility of software-defined vehicles. For some time now, both upstream and downstream in the automotive industry have been working toward this goal, gradually building domain-oriented SOA software architectures.

The second stage is to create new features, new experience of the product power. With the gradual introduction of the new architecture of the whole vehicle into mass production, the application software provides new functions, and the new experience becomes an essential element of market competition, and the trend of building mass scenarios and personalized innovative functions of the whole vehicle will become more and more obvious.

The third stage is the software value-added service stage. With the increasing investment in software, how to realize it and how to continuously obtain software revenue through value-added services after the vehicle is sold will be the main task.

In this process, due to the more and more extensive application of software on the whole vehicle, the development scale and overall complexity of the driving software have been greatly increased, and higher requirements will be put forward for the software development capability of the vehicle factory and related parts suppliers.

Cao Bin pointed out: "This not only refers to the software development organization system and development method of the enterprise, whether it can support such large-scale software development, but also includes whether the software development cost is enough to amortize the marginal gross profit margin in the new model, and whether the gross profit can cover the annual increasing software development raw material cost." "

In the past, the development system of many automobile enterprises was very closed, and the software and hardware of automobile function development were often deeply coupled. In this mode, once the subsequent demand changed, it might drive the whole system-level change, and the change cost was very high.

However, the update iteration speed of new products and new technologies is accelerated, the price competition is fierce, and higher requirements are put forward for R & D efficiency and cost, and even rapid replication between different models is required. For example, chips, in the past a chip may be used in the market for a few years before it is eliminated, but now, just come out of the chip may soon become a transitional product. This situation presents new challenges for software development. When we develop a large software application on a completely new architecture, system, or chip platform, it is likely to need to be updated iteratively in just one or two years.

To sum up, software-defined cars are driving the automotive industry into a new era of intelligence. In the future, automobile manufacturing may not be the highest barrier, but software is the field that can generate premium space.

[Full text reference]

[1]Tesla Software Iteration History: Three Phases, Six Versions, and Countless Innovations, Geek Park

[2]What did Tesla do right? Can other car brands make money from software?, Auto Business Review

[3]Tesla Q2 earnings report: net profit broke 2.7 billion US dollars, up 20% year-on-year, Xinhua News Network

This article comes from Weixin Official Accounts: Chebai Think Tank (ID: EV100_Plus), Author: Zhou Shuangjiang

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.