Li Xiang: Lowering the price doesn't necessarily increase sales, but it can be a way to strike against others.

Pressure is High on Ideal Car in the Face of Industry Price War

Author: Zhan Lianyi
Editor: Qiu Kaijun

“At the top of this industry, two competitors can both engage in a price war, which is something you won’t see in the smartphone industry or any other industry,” said Li Xiang, Chairman of Ideal Car, during the Spring Media Communication Conference held in Beijing on March 2nd. He admitted that the pressure is higher this year, especially in the first quarter which he called it “challenging.”

On one hand, many factors, such as the reduction of new energy subsidies and halving of the purchase tax incentives for traditional gasoline cars, as well as Shanghai’s decision to stop issuing new energy special license plates for extended-range electric vehicles, have caused some customers to advance part of their vehicle purchase demand from January to December of last year.

On the other hand, top carmakers led by Tesla initiated a price war, which caught the industry off guard. Li Xiang believes that cutting prices may not necessarily help increase sales but can certainly “strike against” other competitors as everyone, including consumers, is adopting a “wait and see” attitude assuming other companies will inevitably cut prices, too.

However, Ideal Car will not reduce its prices. Li Xiang argued that Tesla’s price cut would not impact Ideal Car significantly for two reasons. First, Ideal Car’s products are highly competitive. Second, Tesla’s price reduction actually reduced its intersection with Ideal Car in the market. Meanwhile, he expressed concerns that selling medium-sized cars at ¥300,000 RMB or mid-sized SUVs below ¥200,000 RMB would raise customers’ suspicion of low quality.

Li Tie, CFO of Ideal Car, added that the entire market would be under pressure this year, and that in 2023, Ideal Car would challenge a 20% market share of luxury SUVs priced between ¥300,000 RMB and ¥500,000 RMB.

Of course, during the media communication, complaining is not the focus. Li Xiang spent more time sharing his understanding and progress on pure electric vehicles, intelligent driving, and production supply. He also revealed his thoughts and logic of development from “0” to “1” and the challenges and solutions from “1” to “10.”

Ideal Car plans to maintain both hybrid and pure electric vehicles until 2030. When it comes to why Ideal Car was founded, Li Xiang said it is based on trends. Energy and information technology are the two most significant factors that influence society’s development, and intelligent electric vehicles intersect these two areas flawlessly.

Therefore, at the beginning of the entrepreneurship in 2015, although many things were still unclear, he still believed that making electric vehicles was an important opportunity.

Once the target is determined, it is necessary to find the core contradiction and solve it. Li Xiang believed that from the perspective of energy, the core problem for new energy vehicle companies is to solve the difficulty of charging and high battery costs.

Li Xiang analyzed and believed that there are three paths to solve this problem.

The first path is battery swapping. This is a very good way because the user experience can be infinitely close to refueling, and the cost of purchasing a car can also be lowered. But there are pros and cons. Building battery swapping stations means that more batteries need to be prepared, and it also involves the risk of fluctuations in battery prices. Therefore, at that time, with insufficient funds, Li Xiang decided to put as much money as possible on the product side, rather than relying on the service side, and the extended-range electric vehicle was determined.

The second path is to combine user experience with travel efficiency. The specific landing method is to choose to make large batteries (starting from 40 degrees). At that time, whether it was the extended-range model Chevrolet Volt, or the mainstream plug-in hybrid models in China, the pure electric endurance was generally around 50-60 kilometers, while all the models under Li Xiang’s Ideal Automobile currently have an endurance mileage of more than 200 kilometers. Moreover, the private installation rate of charging piles for Ideal car owners is high, which provides a better experience. Li Xiang also added that the total delivery volume of Ideal Automobile is nearly 290,000 vehicles, and the number of charging piles installed is 200,000.

The third path is high-voltage fast charging. According to Li Xiang’s calculation, when public charging piles can be charged 6-7 times a day, they can achieve revenue and expenditure balance. If charging 10 times a day, the profit situation of the charging pile will be better than that of a gas station. However, the current charging piles generally only have 2-3 charges a day and are basically losing money. If the user’s charging time can be shortened to 20 minutes, it can greatly increase the “table turnover rate” of the charging pile. If it is shortened to ten minutes, the charging experience can be basically the same as refueling for gasoline vehicles.

To achieve the above results, Li Xiang said that a high-voltage platform must be used, which not only improves the charging speed but also greatly reduces the cost of batteries by collaborating with silicon carbide and a better drag coefficient.

He gave an example that the best-selling mid-to-large SUV currently has about 100 degrees of electricity and can reach 600 kilometers. However, by using the 800V high-voltage platform, along with better drag coefficient and the optimization of silicon carbide and overall efficiency, 80 degrees of electricity can achieve the same endurance mileage. By reducing the lithium investment after the overall vehicle becomes lighter, the cost can be reduced by 30-40,000 yuan. “In the future, our 800V high-voltage platform electric vehicles can achieve the same price as the current extended-range products.”Meanwhile, in Li Xiang’s opinion, the cost of supercharging stations is not as high as everyone imagines, and it has a spread period of five years, which is much lower in cost than people’s imagination.

He did some calculations: if a high-speed supercharging station can reach about 640-800 kilowatt-hours of electricity, and since the general station has a 3+1 configuration, which is one 480 kilowatt and three 250 kilowatts chargers. If 3000 supercharging stations are established by 2025, the total cost will be 10 billion. Spread it out yearly, the cost is controllable for businesses with annual revenue in the hundreds of billions.

Therefore, Li Xiang stated that by 2030, two strategies will be pursued simultaneously, one further improving the efficiency of extended range, and the other making improvements in high-voltage pure electric cars to give customers two options.

L7 still follows the extended range route

Of course, this is also owing to the accumulation from the “0” to “1” stage. The current revenue scale and growth rate of Ideal Automobile are sufficient to support investment in the pure electric platform, including research and development and deeper supply chain layout.

2024 will usher in the era of intelligent electric vehicles

The above layout is primarily based on the energy sector, and in terms of information technology, Li Xiang summed it up as artificial intelligence.

Ideal Automobile has four in-house artificial intelligence algorithm teams that respectively solve problems in the intelligent cabin, intelligent driving, intelligent factories, and intelligent retail.

“We have actually been using algorithms to solve quality problems,” said Li Xiang. From a quantitative perspective, the cost allocation for Ideal per vehicle’s quality is about 50% of the same-priced BBA (BMW, Benz, Audi) products. The entire algorithm system behind it is crucial, including analysis of various accidents. Therefore, Ideal car insurance is also more favorable than most new energy vehicles priced around 100,000 yuan.

Li Xiang believes that only AI can change the physical world, and the core of AI is learning, not programming, but rather imitation and learning from humans.The reason for this conclusion is that software 1.0 cannot solve problems in the physical world offline.

What is software 1.0?

As Li Xiang claimed, it refers to human beings making rules and coding themselves, and the best program is used by human beings themselves. Whether using the Internet or other terminals such as APPs, the principle remains the same.

“Its core mainly runs in a digital world, and changes in the physical world cannot affect it.” As Li Xiang illustrated with the example of ride-hailing, we can quickly get a taxi, but all the changes that occur between point A and point B cannot change the total ownership cost, which is even higher than that of taxis due to higher idle rates. Therefore, when ride-hailing companies want to grow rapidly or increase market share, they will lose money. The only way to make money is to “exploit drivers,” such as taking a 30% cut in the United States and a 25% cut in China.

Extending to software 2.0, the AI era mainly includes behavioral learning and cognitive learning, which is also the case in the field of autonomous driving.

The autonomous driving behavioral learning stage does not require strong computing power. Based on this, vehicles under the Ideal Cars brand can complete the work by carrying Mobileye chips. In fact, this stage of assisted driving relies on rules written by humans, rather than AI algorithms. It was not until the release of Tesla FSD that the industry underwent a fundamental change and entered the cognitive learning stage.

Li Xiang claimed that the future competition in the field of autonomous driving mainly involves three parts.

The first is how to reduce the cost of vehicle computing platforms and sensors. The cost for Tesla is only $1,500, while the cost for enterprises using dual Orin chips is more than $4,000, resulting in a big gap.

The second is who can have more end-to-end closed-loop data, and only getting part of the data is useless. This means that enough vehicles need to run on the road.

The third is to challenge the higher cost and greater difficulty of large model training. Li Xiang claimed that in addition to solving how to make the model more efficient, it also includes difficulties in China’s training chip development, which has been restricted by the United States since last year, and restrictions on Nvidia selling high-bandwidth training chips to China.

Restricting the export of Nvidia's high-bandwidth chips to China“For companies determined to develop artificial intelligence, there are three essential parts to invest in and address. As Li Xiang said, besides producing high-quality products, platforms, and algorithm applications, the real competition lies in the underlying component, including AI operating systems.

Li Xiang believes that the era of truly intelligent electric vehicles will arrive in 2024, especially for mid to high-end models, based on large models and BEV technology that will cover navigation assistance in all cities, making the user experience completely different.

Data shows that the overall usage rate of Ideal Auto’s Advanced Driver Assistance System is over 13%, and in high-speed scenarios, the usage rate has already exceeded 50%. Li Xiang said that if a technology is used by a user every day, and its usage rate exceeds 60% of the mileage, it will never be replaced.

In Li Xiang’s view, car companies based on NVIDIA’s Orin platform will basically launch urban navigation-assisted driving by the end of this year. Ideal Auto also plans to open the “early bird” user beta testing of urban NOA on Ideal AD Max in the fourth quarter of this year. Ideal Auto will not calibrate one city at a time but will cover all cities based on large models when releasing urban NOA.

Currently, in terms of performance, only Ideal AD Max can handle this type of urban NOA with large models, while AD Pro cannot due to AD Pro’s performance being designed from the beginning to only meet high-speed NOA. Li Xiang predicts that most leading enterprises will be able to reach the level of Tesla by the end of 2021, and at this time in 2024, everyone will be able to achieve the level of Tesla in North America by the end of 2022 or early 2023. “In the future, at least for mid to high-end vehicles, if they cannot offer urban NOA, consumers will have differences in their buying behavior. At this time, mid to high-end cars will completely enter the era of intelligent electric cars based on Software 2.0.”

“A self-produced rate of 30% is a healthy standard,” Li Xiang said as he discussed the future of Ideal Auto after sharing the journey of the company’s past and present.”As early as 2015, when Li Xiang founded Li Auto, he mentioned that it would become a global leader in artificial intelligence by 2030, and this still holds true today.

A more specific plan is to capture 35% of the market share for passenger cars priced above 200,000 RMB by 2027.

“Whether there are three or five leading companies in the head of the industry, we hope that we will definitely be one of them,” Li Xiang said. Regardless of whether it is the product level, platform application level, or underlying system level, Li Auto must have the same capabilities as Tesla and Apple to compete.

Li Xiang discussed the specific implementation of the goal planning through a “cross” model, which includes five dimensions: brand, product, gross margin, R&D, and delivery.

Regarding the brand dimension, Li Auto must have a very clear positioning. In Li Xiang’s view, only 1% of the new automakers have a qualified brand, and most of them don’t even have basic common sense. Talking about family cars today and 3-second acceleration tomorrow, all the logic is chaotic and can affect subsequent products and business processes.

In terms of products, both software and hardware are important. For software, Li Auto must self-develop centralized domain controllers, intelligent driving controllers, and other technologies. For hardware, product development should be based on brand positioning.

Li Xiang admitted that gross margin is crucial. If Li Auto wants to be a long-term player in the automotive industry, it needs to have hundreds of billions of RMB in revenue, and relying solely on financing is not sustainable. Therefore, how to manage the business becomes crucial.

As an intelligent electric vehicle enterprise, Li Xiang believes that a gross margin of 20% is a healthy threshold. Currently, in the industry, Tesla’s gross margin exceeds 20%, and if BYD includes dealer fees in the calculation, its automotive gross margin also exceeds 20%.

Li Auto insists on maintaining a gross margin of 20% or higher, which is where the company’s blood-making ability lies.

Regarding R&D, Li Auto divides it into three levels: product development, platform development, and system development.

Li Auto range-extending platformFrom 2019, Ideal Motors decided to develop its R&D in a platform-based way. Inspired by the development model of Apple phones, the entire platform R&D is divided into four teams: extended-range electric platform, high-voltage pure electric platform, intelligent driving platform, and intelligent cockpit platform. This greatly improves the efficiency of OTA, but the investment is also particularly high. To continuously reduce platform R&D costs, systematic development support is needed to determine which technologies to self-develop, which ones to cooperate on, and which ones to outsource.

In Li Xiang’s view, only after achieving basic product development, platform-based R&D, and system R&D, can it become an enterprise like Apple, Huawei, and Tesla, and the competitive procurement advantage will be at a higher level.

In terms of delivery, in order to rapidly increase production capacity, Ideal Motors divides all components into three major parts. The first part is traditional components, such as rearview mirrors and bumpers, which are directly purchased from suppliers. The second part includes new electronic chips, domain controllers, etc. China purchases 40% of the chips from the world, and the manufacturing efficiency is very high. The third part is related to the three electric components, which are limited in supply from suppliers due to their own judgement standards for new forces orders. Ideal Motors is also looking for partial self-construction to achieve self-sufficiency. Ideal Motors believes that a self-reliance rate of 30% is a relatively healthy standard.

Considering various factors, Li Xiang said that Ideal Motors’ sales system is becoming more rational. In February, each store sold nearly seven cars per person, and if the number of management personnel is deducted, it exceeded 10 cars. This is probably more than three times that of new forces peers.

Of course, Li Xiang also pointed out that there is still much room for improvement in the enterprise, including commercial, supply, and organizational capabilities, etc.

Summarizing his entrepreneurial experience, Li Xiang said that as he challenges the goal of reaching 10 billion RMB in revenue, he has experienced five stages. The first stage is feeling unique and never looking outside. The second stage is looking outside but not understanding. The third stage is finding some high-level consultants or companies to help you understand. The fourth stage is doing immediately after understanding and encountering problems. The fifth stage is finding the problem and working hard to solve it.

“We learn and avoid from our rivals’ strengths and weaknesses.” Since starting his business, what Li Xiang has learned the most may be how to learn better.

Ideal Motors stores are highly efficient# 这是一个标题



以下是一个 HTML 的表格:

姓名 年龄
张三 18
李四 20


  1. 第一项
  2. 第二项
  3. 第三项


  • 第一项
  • 第二项
  • 第三项

以下是一个 HTML 的图片:








This article is a translation by ChatGPT of a Chinese report from 42HOW. If you have any questions about it, please email