Para los que no lo vieron, aca esta un resumen que hizo un user que participo de la meeting mas reducida con William Li.
Preferi no traducirla:
On the afternoon of September 3, I was sitting in a meeting room at NIO’s headquarters in Caohejing, Shanghai—the same place where NIO held its closed-door meeting following the release of its first-quarter financial results.
It felt familiar, but somehow, the atmosphere seemed a little more relaxed.
Just two days earlier, NIO had released its financial results for the second quarter of 2026.
RMB 32.14 billion in revenue, RMB 5.91 billion in total gross profit, an 18.5% vehicle gross margin, RMB 56.7 billion in cash reserves, and three consecutive quarters of profitability.
These numbers may look light on paper, but against the backdrop of the automotive industry this year, they carry considerable weight.
Domestic passenger vehicle sales fell by roughly 20% year-on-year during the first half of the year. Both Seres and Li Auto saw their quarterly financial results swing from profit to loss.
NIO, by contrast, delivered a set of results against the trend: first-half revenue reached RMB 57.67 billion, up 85.8% year-on-year, while total gross profit surged 282.2% to RMB 10.77 billion. Revenue growth outpaced vehicle sales growth, while gross profit growth outpaced revenue growth. This progression—from sales to revenue to gross profit—is perhaps the clearest illustration of how NIO has changed.
Around the same time last year, NIO was still stuck in a quagmire of quarterly losses of around RMB 5 billion. Going from losses to three consecutive quarters of profitability in just one year is clearly more interesting than the financial figures themselves.
At the beginning of the closed-door meeting, William Li said that he now genuinely enjoys managing the business. For him, it is a process of seeking the truth.
1. The Quality of NIO’s Financial Results: It’s Not Just About Profit, But How You Make It
More meaningful than profitability itself is the quality of that profitability.
During the closed-door meeting, William Li repeatedly emphasized one particular perspective: don’t just look at sales rankings. Look at revenue rankings, total gross profit rankings, and operating profit rankings.
“People used to rank companies by vehicle sales. But why did nobody rank them by sales revenue? Why did nobody rank them by gross profit?”
In his view, high-quality growth cannot simply be measured by sales volume.
“In the past, it was about increasing volume, then increasing revenue, and then increasing profit. Today, we need to look at increasing profit, then revenue, and then volume.”
The shift in this sequence reflects a fundamental change in the way the company is being run.
The traditional logic of the automotive industry was that scale is king: increase sales, spread fixed costs across more vehicles, and profits would naturally follow. But the reality in 2026 is delivering the opposite answer: sales can rise while costs rise even faster.
“In the past, we used to say that as volume increased, costs would come down. Now, as volume increases, costs can also go up. Take memory, for example. If you buy less, suppliers may still give you a discount. But if you buy more, they may say, ‘Sorry, the price goes up.’ That’s quite interesting. In the past, the larger the volume, the bigger the discount. Now, sometimes, the smaller the volume, the bigger the discount. This is the real situation. We are seeing diseconomies of scale.”
This is not a problem unique to NIO. It is a supply-side shock affecting the entire intelligent EV industry.
William acknowledged that the industry is now facing a situation where scale no longer automatically brings economies of scale.
He broke down the problem in considerable detail:
“In Q2 this year, the cost per vehicle was RMB 14,000 higher than at the end of last year. That means we lose RMB 1.5 billion every quarter on average. Q1 was still relatively manageable because we still had some inventory from last year. By Q2, most of that was gone. We expect costs to continue rising in the second half of the year, by another RMB 2,000–3,000 on top of that. So costs could be RMB 16,000–17,000 higher in the second half. These are things we simply cannot negotiate. We have absolutely no negotiating power.”
There was a sense of resigned realism in his tone.
“Our negotiation is basically begging suppliers: ‘Brother, please give me some supply.’ That’s pretty much what it looks like.”
The surge in memory prices is being driven by massive purchases from AIDC—AI data centers. When a PCB manufacturer can get five or even eight times the price from an AI company, why would it continue selling to an automaker at the original price?
Batteries, copper and aluminum are being caught up in the same dynamic.
Intelligent EV makers such as NIO happen to be right at the epicenter of this cost shock, because the amount of memory used in an intelligent EV is far greater than in a traditional ICE vehicle.
Even so, NIO’s vehicle gross margin remained stable at 18.5% in Q2, up 8.2 percentage points year-on-year. This was not accidental.
In Q1, NIO’s vehicle gross margin was 18.8%. The two quarters were therefore almost identical. Holding gross margin steady despite an increase of more than RMB 10,000 in per-vehicle costs required both an optimized product mix and discipline in pricing.
“We have not pursued volume through price cuts,” William said. “Even if costs continue to rise in the second half of the year, our guidance still assumes that NIO should basically be able to maintain a stable vehicle gross margin.”
Whatever happens externally, the priority is to strengthen the foundation of profitability.
During the media Q&A following the closed-door meeting, William was asked how NIO would balance gross margin against sales volume.
His answer was straightforward:
“Ultimately, we want to protect total gross profit. Total gross profit comes first. Directionally, we don't think there is much point in doing anything with pricing right now, unless there is a very significant price change. Small changes don't mean much, while big changes simply aren't feasible—the cost structure doesn't support them. There would be no gross profit left.”
“When we run internal calculations, it is always about maximizing total gross profit and maximizing EBIT,” William said. “If we look at the current situation, changing prices doesn't help. It won't increase total gross profit, and it won't lead to a significant increase in volume either. A 5% price increase doesn't really make much difference.”
For William, operating profit is the most fundamental measure of a company's health.
“Under GAAP, there are also adjustments related to stocks and options that have nothing to do with cash flow. Looking at operating profit shows whether you have the ability to continuously turn technological innovation, product R&D and service systems into profits. From the perspective of operating profit, it proves that we have the ability to continuously convert technology innovation, product development and our service system into profit. That is also the foundation of healthy and sustainable development.”
This is not profit created simply by cutting costs.
During the first half of the year, NIO continued investing in fundamental R&D. Its charging and battery-swapping infrastructure continued to expand, while sales expenses increased alongside the launch of new models.
The source of its profit is increasingly coming from the precision of its business management.
Since last year, NIO has been implementing its CBU mechanism, breaking the company down into individual operating units and making everyone responsible for EBIT.
2. Forget the Sexy Stories. Focus on the Most Fundamental Numbers.
At the closed-door meeting, William did not talk about disruption or being “far ahead of everyone else.”
Instead, he spent a great deal of time discussing CBU, ROI, efficiency and cost control—topics that may sound dry and unglamorous.
Several times, whenever he talked about “doing the math,” there was genuine excitement in his eyes.
It was the excitement of someone who had found a way to solve the problem.
Just how deeply is NIO committed to operational quality?
According to William, NIO is now working on what it calls “CBU down to the individual component level.”
A car can be broken down into 17,000 or 18,000 individual parts.
“If we sell 500,000 new vehicles a year, that means more than 8 billion components. Actually, that's not too bad. Large models nowadays have trillions of parameters, so 8 billion isn't that much.”
NIO ultimately wants to use this massive amount of data to identify opportunities to improve efficiency and reduce costs.
“How can AI identify opportunities to improve efficiency and reduce costs from all of this? That is what we are working on. The direction is already very clear, but the capabilities are still being built. We often run into dead ends. Sometimes we feel like we've hit a wall and can't get past it. But then, after discussing it, we find a way forward. We estimate that it will take us about another year to fully develop this capability.”
At its core, the CBU—or basic business unit—mechanism is about breaking the company into a series of independently accounted-for “cells,” with clear ROI metrics and performance incentives for each business unit.
William gave an example:
“If our sales are RMB 100 billion this year, improving efficiency by 5% means RMB 5 billion. Everyone gets very excited. Seeing that improvements in our organizational capabilities can genuinely translate into business results—that's very practical, very down-to-earth.”
There is another easily overlooked detail.
NIO no longer looks at its operating metrics simply on a monthly basis.
It looks at them every day.
“Our company hasn't had a monthly vehicle-delivery target for a long time. We stopped looking at that more than a year ago. Whatever the actual number is, that's what it is,” William said.
“But when it comes to our operating objectives, a year ago everyone was already looking at their own business results. Under the CBU mechanism, it's not something you look at every month. You look at it every day.”
Qin Lihong added a very specific example.
Yang Bo, who is responsible for user operations for the NIO brand, would traditionally be called the “national sales chief” at a conventional automaker.
But Yang describes his own role differently—“He is the fellow of all the fellows nationwide.” NIO calls its salespeople “fellows.”
The role is not about managing them, but supporting them.
His job is to help 6,000 fellows improve their productivity—even if only slightly—under the existing conditions.
“There are two things we need to do. First, we need technological empowerment—AI. Second, we need an organizational structure. We have regional companies, general managers and sales leaders, all directly empowering those 6,000 fellows. At the same time, the organizational structure itself cannot simply be bypassed. Finding that balance is extremely difficult.”
“Talking about these things isn't particularly exciting, but they take a tremendous amount of time. We are trying to make sure that every individual can improve their productivity even a little under the existing conditions. Ultimately, that's where the results come from.”
This kind of incremental, almost invisible improvement is unlikely to attract much attention from the outside world.
But it is genuinely changing the way NIO operates.
William also responded to the perception that NIO has become “boring.”
“People may feel that NIO has become a little boring. You don't even have robots dancing anymore…”
“I’m actually really enjoying managing the business now.”
The implication was that the industry itself has fewer and fewer surprises to offer.
“Take Apple's product launches. This year, they may launch a foldable iPhone, but that wouldn't exactly be a huge surprise. It's not that there is no room for progress. Today's iPhone is obviously much better than the iPhone 4. But it doesn't bring the same kind of surprise as the iPhone 4 did back then.”
“That doesn't mean Apple hasn't made progress. A lot of the progress is happening in places you can't see. Take Apple's chips, for example. That's where some of the biggest progress has happened.”
When everything returns to calm, the capabilities that truly change the game are often hidden beneath the surface.
3. The Foundation for Cross-Industry Expansion Is Already There. But NIO Chooses to “Let Others Go First.”
Today, the automotive industry finds itself in a somewhat awkward position.
It is no longer the industry standing at the absolute center of the spotlight.
Ten years ago, new-energy vehicle startups were among the sexiest stories in the technology world. Even just a few years ago, every auto launch was a collective spectacle, and investors were terrified of missing the next new automotive startup.
Cars seemed to embody every major vision people had for the future of mobility.
Electrification was the spearhead of a transportation revolution. Intelligent driving was the first major application of AI. Shared mobility, enabled by charging infrastructure and Robotaxis, was positioned as a catalyst for urban transformation.
At one point, the automobile was where almost every major technology narrative converged.
Then AI and embodied intelligence emerged, and the direction of attention changed.
Large AI models became the narrative black hole that seemed capable of “swallowing everything.”
After OpenAI, NVIDIA's market capitalization at one point surpassed US$3 trillion. Jensen Huang was surrounded by crowds at virtually every major forum.
Embodied intelligence, meanwhile, offered investors an even more intoxicating story: the “ultimate physical form of artificial general intelligence.”
A startup with barely a dozen employees, whose product had not even reached the threshold of mass production, could command a valuation of RMB 5 billion or RMB 10 billion.
Money, talent and attention—all the scarce resources—were flowing toward these two sectors.
What about cars?
Smart EVs are still selling. New vehicles are still being delivered. Technology is still evolving.
But in the eyes of investors, the industry has become “certain.”
And “certainty” is not a particularly sexy word in capital markets.
Battery energy density improves by a few percentage points each year. Intelligent driving systems are gradually moving from L2+ toward L3. New-car launches have shifted from “disruption” to a competition over specifications.
Refrigerators, TVs and giant sofas in cars have long since lost their novelty.
The industry seems to be moving from an “infinite game” to a “finite game,” and from concept-driven growth to operations-driven growth.
William explained this very clearly during the meeting.
“AI is hitting the automotive industry in three ways.”
“The first is pricing. Memory production capacity is tightening, and even PCBs may be in short supply. Some companies are willing to pay five or eight times the original price. If I were a PCB manufacturer, why would I supply an automaker instead? This is something we never expected. Copper, aluminum and batteries are also being pulled toward AIDC. The impact is very real.”
“The second is talent. People working on intelligent driving are leaving to start embodied-intelligence companies. Valuations are very high. I don't even know when some of these companies will reach our level of revenue.”
“The third is the attention of the capital markets. Right now, it feels like if you don't attract attention, you are seen as lacking imagination.”
In reality, NIO has built a very solid technological foundation.
The company has established a 12-domain full-stack technology R&D system around intelligent EVs, covering chips, in-vehicle intelligent hardware, the three-electric system, a full-vehicle operating system, intelligent driving and other areas.
Its cumulative R&D investment has exceeded RMB 67 billion.
From the perspective of its technological capabilities, NIO already has the foundation to enter many of today's hottest sectors.
William acknowledged this during the meeting:
“From the perspective of our capabilities, there are many things we can do. For example, we can do energy storage, which is highly related to our battery capabilities. We can also make various intelligent electric tools and products. And we have a lot of software capabilities. We have 12 full-stack technology domains, and they correspond to many different fields.”
So the issue is not that NIO lacks imagination. It is that many people lack imagination when it comes to what NIO could become.
Chips are a very good example.
After NIO's proprietary intelligent-driving chip, Shenji, successfully completed tape-out, the team was not simply kept inside NIO and funded as another internal department.
Instead, it was established as an independent entity and raised external capital.
Its valuation has now reached the RMB 10-billion level, attracting interest from several leading semiconductor investors and automakers.
William revealed:
“Shenji's external revenue this year could reach several hundred million yuan. There is actually quite a lot of interest.”
NIO's energy business follows a similar logic.
NIO handles tens of billions of kilowatt-hours of electricity every year. Its nationwide energy network—with 4,058 battery-swap stations and more than 25,000 charging piles—is itself a huge distributed energy asset.
To many people, it is simply “infrastructure that supports car sales.”
Look at it differently, however, and it is an existing virtual power plant, a network of distributed energy-storage nodes, and an entry point into electricity trading.
“Isn't NIO also an energy company?” William asked rhetorically.
Once this asset is repriced, its value could go far beyond the four words “energy infrastructure.”
But NIO has chosen not to personally jump into these businesses right now.
William put it very directly:
“Our resources are limited. My energy is limited. The energy of our team and Lihong is also limited. We are simply focusing on doing things properly in China. We hope to achieve market share across the country similar to what we have achieved in Shanghai and the Yangtze River Delta.”
For other areas, NIO has its own approach.
It can spin businesses out and raise external funding. It can incubate entrepreneurial teams.
But it does not necessarily have to enter those markets itself.
“Our approach to building these businesses is different. We can make investments through spin-offs and external financing, or cultivate entrepreneurial teams. There are different ways to build these capabilities. We have to allow ourselves to be willing to ‘let others go first.’”
This is a sharp contrast to NIO's style several years ago, when it seemed determined to do everything itself.
From scaling back NIO Phone to taking an indirect approach to embodied intelligence, NIO is now exploring its boundaries in a much lighter way.
“If we have no room for growth and still refuse to consider opportunities in other areas, then as CEO, I wouldn't be doing my job properly. It's actually not complicated. We are simply being pragmatic.”
Pragmatism was one of the most frequently repeated words during the meeting.
4. NIO’s Second and Third Growth Curves
NIO's growth engine is gradually shifting from being new-car-driven to system-driven.
The first growth curve, of course, is selling cars.
In July this year, NIO's sales in 42 Chinese cities simultaneously surpassed those of Mercedes-Benz, BMW and Audi.
The NIO brand's average transaction price reached RMB 434,600 in July, surpassing all three German premium brands. Its lead over Audi alone reached approximately RMB 180,000.
“This shows that users recognize our brand,” William said.
Individual model performance reinforces the point.
One year after launch, the ES8 has approached 150,000 cumulative deliveries. In the vehicle satisfaction report released by J.D. Power, it ranked first across all brands with a score of 85.3.
NIO's brand NPS also ranked first in the first half of the year.
William revealed:
“In consumers' purchase decisions for the NIO brand, the brand factor now accounts for more than 30%, making it the second most important factor.”
The ONVO brand has also established a solid position.
Its average transaction price during the first half of the year was RMB 248,000.
“That’s not much lower than Audi, and it is higher than Cadillac and Volvo. Only eight brands managed to achieve both volume and price growth in the first half of this year. ONVO was one of them.”
The L60 achieved a one-year residual value of 74.92%, ranking first among pure-electric midsize SUVs.
Firefly, meanwhile, has ranked first in the premium small-car segment for 15 consecutive months, “completely breaking through the valley of death that typically follows the initial new-model effect.”
The second growth curve comes from services and community businesses.
William provided specific figures:
“Last year, our services and community business exceeded RMB 10 billion. It continued to grow in both Q1 and Q2 this year, and both quarters were profitable. It will also grow for the full year.”
“This is revenue generated from the installed base, and it represents a commercial growth engine that other companies currently don't have.”
The business had already achieved a gross margin of more than 20% in Q1.
“As the installed base grows, could service and community revenue support the company? Even if the number of new vehicles sold each year doesn't increase, services and community businesses are based on the installed base, so there should still be significant room for growth.”
Within this segment, there are multiple individual growth engines.
“We have more than a dozen business lines that are growth points. For example, subscription services. Our intelligent driving subscription service generated tens of millions of yuan in revenue this year, and it is specifically targeting used-car users.”
“There are only tens of thousands of used-car users, and we've already generated tens of millions of yuan in subscription revenue from them.”
William revealed that approximately 40,000 owners of second- and third-generation vehicles are currently eligible for subscriptions, with around 8,000 subscribing on a regular basis.
“There is still a lot of room because there will be more and more used cars.”
The third growth curve is energy.
NIO handles tens of billions of kilowatt-hours of electricity each year and is actively exploring value-added businesses such as electricity trading through its charging and battery-swapping network.
If the battery-swap alliance model succeeds, the value of this energy infrastructure could be repriced accordingly.
“We are very happy to see companies such as CATL investing in battery-swapping networks, and other automakers choosing the battery-swapping route. We believe this is absolutely the right direction,” William said.
NIO remains open to cooperation, but its priority is clear:
“Our priority is still to serve our own users. If others want to come, we welcome them. We won't refuse.”
These three curves together form the bigger picture of NIO's third growth cycle.
The company is no longer relying on breakthroughs in a single area. Instead, it is relying on its broader system capabilities to continuously unlock value.
William offered a more macro view:
“In another five to ten years, the industry will reach a certain level of equilibrium. There will be consolidation, but it won't be as concentrated as many people imagine. Geopolitics is a major reason for that.”
5. Conclusion
The marathon has entered its second half.
At this stage, what matters is no longer sprinting ability, but pacing, rhythm and whether you can put one foot firmly in front of the other.
NIO has chosen a slower and steadier path.
It is not trying to launch a blockbuster every quarter. It is not trying to jump onto every new trend.
Instead, it is focusing on running the business properly—one transaction, one business unit and one calculation at a time.
One comment from William during the meeting perhaps best captures the philosophy:
“The hardest thing about running a company is knowing what to stick with and what to change.”
“We've made many changes over the past few years, but our direction has never changed. Our pure-electric and battery-swapping technology route will not change. Our commitment to continuously building battery-swap stations will not change. Our focus on user service and building a high-quality brand will not change. And our commitment to building an organization centered on creating user value and encouraging everyone to think like an operator will not change.”
Qin Lihong added another key word:
“Strategic discipline sounds easy to talk about, but it is actually very difficult to practice.”
“Today, some companies may need to find new concepts to support their market capitalization. When it comes to embracing emerging industries and emerging technologies, our decision is to embrace them actively—but not necessarily in a way that grabs everyone's attention.”
“In fact, we are embracing them very deeply.”
“Perhaps when people look at NIO today, they need to peel back one layer and look deeper.”
At a time when capital markets are hungry for new narratives and competitors are chasing attention, choosing to return to the fundamentals of running a business requires enormous discipline.
Near the end of the meeting, William was asked how he viewed the market's apparent undervaluation of NIO.
His answer was calm:
“We can't make decisions on behalf of investors. We can only focus on doing our own job well—continuing to innovate technologically, continuing to build the brand, and maintaining strategic discipline.”
“In the end, the market will come back to looking at a company's operating capabilities and profitability.”
The road to building cars may not always be glamorous.
But NIO is walking it with conviction.
