NIO, China’s leading smart electric-vehicle (EV) company, has attracted attention for its unique user-centric operating model. However, the model has also triggered heated discussions about profitability and sustainability. NIO’s cofounder and president, Lihong Qin, sat down with McKinsey’s Mingyu Guan and Tony Zhou to shed light on the company’s strategy. He described the logic behind NIO’s user operations approach, discussed input–output economics, and considered the implications of the company’s innovative approach for the industry.
This interview has been lightly edited.
The underlying business logic of user operations
McKinsey: User operations is one of the areas in which NIO really stands out. Why did NIO choose user operations? What is the underlying business logic? How does it increase NIO’s competitiveness?
Lihong Qin: From a business model perspective, the success of user operations is predicated on whether a company has natural, long-term touchpoints with its users. Let me share an example in real estate and property management, which I worked in before I cofounded NIO. The property management business is naturally suitable for user operations because property owners can stay in a property for ten years. Real-estate transactions, however, are not—the buyer and seller basically stop contacting each other once the deal is done.
In the automotive industry today, most traditional car companies don’t own many user touchpoints. Let’s take a look at the typical transaction flow of a traditional OEM. The presales stage focuses on attracting potential customers through various marketing methods. In the transaction stage, the entire business flow is completed after the vehicle is delivered from the OEM to the dealer. In the after-sales stage, unless the car is recalled there is little engagement between the OEM and end user. Therefore, through the vehicle life cycle, traditional OEMs have hardly any user touchpoints, which means there is no opportunity for user operations.
At NIO, we enjoy long-lasting user touchpoints through direct sales and services, so we meet the precondition for user operations. As for why we favor user operations, there are two key considerations. First, because we have so many long-lasting touchpoints and so much in-depth engagement, we create hard-to-ignore value potential through the user life cycle. So the question really is whether we can unlock value through user operations. Second, we hope that our car brand can create premiums from user operations.
Based on these considerations, we have built a series of user operations protocols. A consumer’s willingness to pay is always driven by both rational and emotional factors. To meet the rational needs, we strive for excellence in product R&D, the supply chain, and manufacturing. However, we also aim to build long-term customer satisfaction through user operations—an important lever to add value to the brand. [See sidebar, “McKinsey commentary: A new path to long-term customer satisfaction.”]
Let’s say NIO didn’t do direct sales or user operations and simply copied the distribution and dealership model of traditional OEMs. How much would our cars sell for? My personal hunch is maybe 20 percent below today’s price. For a long time, Chinese car companies were subject to a pricing ceiling. For the same product, even where Chinese cars were better, they would be priced at $3,000 less than South Korean cars, $5,000 less than Japanese cars, and much lower than German cars. That seemed to be the only way to sell Chinese cars. Could Chinese OEMs dare to say no to these prices or even set the price higher than those of German brands? How could they justify that? Although this is a hypothetical question, it was an important topic of discussion back when we started our business.
Through concerted effort, including user operations, NIO has achieved a very straightforward result: a decent brand premium. NIO’s current transaction price is higher than that of some traditional German luxury-car brands—a lot higher, actually. If the traditional strategy means pricing at thousands of dollars lower than German luxury brands do, and our actual price is now thousands higher, the difference between the two price points can be seen as our value added. This is one of the most direct accomplishments of our efforts in user operations.
Would you like to learn more about McKinsey’s Automotive & Assembly Practice?
McKinsey: Besides the price premium, you also mentioned unlocking value throughout the vehicle life cycle. Can you share some insights on this? Also, some people say that the “user operations” simply means that the salesperson manages the relationship with the customer better. What’s your view on this?
Lihong Qin: You can’t simply equate user operations with relationship building or other superficial interactions. User operations requires the support of a complete set of system capabilities. NIO focused on building its own ecosystem from the very beginning, including the NIO app, NIO House, the Worry-Free Service, and NIO Life. These not only help us boost user satisfaction and create a brand premium but also benefit us significantly by unlocking value throughout the user life cycle. Taking NIO Life as an example, the current penetration rate is over 50 percent among existing users, and the average annual spending per user is $300. But of course some of the products are bought with redeemed NIO points.
You can’t simply equate user operations with relationship building or other superficial interactions. User operations requires the support of a complete set of system capabilities.
Is it hard to implement user operations? Well, it is hard but also not hard. Why not hard? Think about it. A user who spends $60,000 to $80,000 on a car will stay with NIO for at least five or six years. As long as the user experience is good, it is very likely that some additional spending will be generated. Let’s take another example in the property management business. Say the owner of a luxurious house is very satisfied with the property service. Then the property operator launches some other value-added services, such as home appliance repair or child pickup. Assuming that the price is fair and the owner strongly trusts the operator, it is obvious what the customer will choose.
Then why are user operations hard? First, it is not practical for OEMs operating a dealership model. How can they bypass dealers to sell value-added services? Second, the traditional operating logic and corporate structure of the automotive industry does not support user operations. The departmental KPIs of a traditional OEM may not fully take into account the long-term interests of the company.
For example, for marketing to launch any initiatives, it must first prove their ROI within a year. But at NIO, the financial returns on what we are doing right now may not be realized for several years. Therefore, user operations can only be rolled out by those who have absolute control over a company and, at the same time, are fully invested in this way of thinking. The same group of people must make decisions and execute.
The input–output economics of user operations
McKinsey: User operations requires real investment. From the user perspective, I would be happy to see NIO spend more. But from the perspective of investors, isn’t it natural to be concerned about the profitability of the user operations model?
Lihong Qin: Regarding the financial returns, we have our own model to evaluate the profit and loss of a vehicle through its life cycle. That includes two important items: first, marketing and sales cost per car and, second, service cost per car for its entire life cycle.
Marketing and sales cost per car includes things such as advertising and new-car launch expenses. If we followed the traditional model, NIO’s costs in this area would never be lower than those of traditional luxury-car brands. In fact, they would be much higher. For example, we would pay the same as established brands for a 30-second TV commercial. But a traditional brand’s advertisement can generate one million sales a year, so the spending per car is significantly lower than that of NIO. It would be impossible for us to cover such a huge gap, even with extreme efficiency and precise placement. To compete with traditional OEMs with scale advantages, our play has to be different.
Why the automotive future is electric
How to reflect the differentiation? We think there is only one variable to focus on: user referrals. The only viable strategy for NIO is to systematically increase the proportion of new users referred by existing users. That comes back to user satisfaction, as a high referral rate comes from high satisfaction. However, satisfaction alone doesn’t necessarily lead to a high referral rate. There has to be a conversion mechanism as well. This requires us to increase the frequency of our events—especially offline events—and to make the scenarios for referral as spontaneous as possible, so that users’ friends can have a pleasant experience without any “promo awkwardness.” This type of play is hugely different from common promotional events in the industry.
Apart from user referrals, another important strategy is to host big events with a massive impact. Take the first NIO Day, at the end of 2017, as an example. The number of orders received within 24 hours of the event, when calculated by subsequent deliveries, was equivalent to a year and a half’s delivery volume. Of course, the production capacity of the new factory was limited at that time, and the long delivery time led to some order losses. In the end, the actual orders amounted to a full year’s deliveries.
The fact that orders generated from one event were equivalent to one year’s delivery capacity means we didn’t have to invest in big commercials during the next year, and that contributed to substantial and systematic cost savings. The resources saved could then be invested in services for interested users, with costs much lower than those of acquiring new leads through big commercials.
Besides the marketing and sales costs per car, another important aspect of our financial model is the service cost per car. In the first few years, when our sales were limited, the unit cost could be horrendous, but as sales went up the unit cost gradually reached normal levels. Take our road service team as an example. When sales were still relatively low, such as in 2018, a road service specialist might only serve one order per day. But by 2021, each of them was serving nearly six orders per day on average. That’s a huge increase, actually reaching the limit of an eight-hour workday.
Another example is Power Swap Station, NIO’s battery-swapping station, which costs a lot in initial investment. However, this depends on how you look at it. I told the head of NIO Power that if NIO didn’t directly operate the battery-swap service on its own but purchased the service externally and still wanted to achieve the same satisfaction levels, the purchasing cost would definitely be much higher than under today’s in-house model. User service may show a loss in financial statements, but it can also be interpreted as an investment in customer satisfaction, which I think is the more appropriate perspective.
To sum up, NIO leverages differentiation to optimize sales costs. In terms of service costs, we seek excellence with up-front investment and then optimize the service cost per car by scaling up, until ultimately achieving breakeven.
Apart from expenses, services can also generate income for NIO throughout the vehicle life cycle. Our business scope is wider than that of a traditional car company. On top of conventional business, NIO also deals in peripheral products, used cars, etcetera. Under the traditional model, the value chain is serviced by more than a dozen companies. At NIO, it is different. Services like car wrap and caliper change are undertaken by ourselves. Moreover, due to the simplification of distribution steps, pricing can be lower than that of traditional dealers—benefiting consumers while maintaining reasonable margins.
NIO also offers options, such as NIO Pilot and NOMI Mate, exclusive to smart EVs. Thanks to a solid product experience and healthy user relationships, the take-up rate is very high—between 50 and 60 percent. At the same time, more and more users are becoming accustomed to accessing paid-for services at NIO House—for example, drinks and venue rentals. Of course, part of the payment is via NIO Points.
We are currently in the stage of reputation building, so we are happy to reinvest part of our service revenues in user services. To ensure that all of these initiatives best serve the interests of our users, while keeping in mind financial performance, we established a fully fledged User Relations Department. Traditional OEMs do not have such a function, but user relationships are very important in NIO’s system. We need this department to define rules for our user community. Its role is comparable to that of the Roman Senate. In some cases, even when William Li1 and I make a decision, the department can stand up and say, “No, this will hurt users’ interests.”
McKinsey: NIO Life is a distinctive piece of NIO’s business. But some people say that for a car company, investing in NIO Life seems like messing around with an irrelevant business. What’s NIO’s view on this?
Lihong Qin: NIO Life is mainly responsible for the development and sale of our peripheral lifestyle products. Starting with three people, it has grown to a 100-strong team today. We define and design the products ourselves and let the supply chain implement and deliver. We seek excellence in staff recruitment, supply chain building, and quality standards, and it is no exaggeration to say that the department is as competent as leading companies in the corresponding industries.
For example, we have recently been working on red wine and have developed a direct partnership with French wineries to jointly develop NIO’s exclusive vineyards. Another example is our top-selling thermos bottle. We designed the bottle, defined the quality standard, selected the manufacturer, and our quality staff oversaw the delivery inspection. New products, colors, and concepts are developed twice a year. We hope that by providing users with quality services and products, they will feel they are getting great value for money. At the same time, NIO can ensure reasonable margins by optimizing the system’s efficiency.
We understand the external puzzlement or even criticism, which you mentioned. But we feel that capability building within NIO Life should not be seen as being in conflict with our main automotive business. Today, the penetration rate of NIO Life has exceeded 50 percent among NIO users, and will rise further as capabilities in multiple areas are put in place. With the scale-up of vehicle delivery, the investment here is not putting additional pressure on us, but will instead become a new source of value growth.
To take the idea a step further, it is actually more cost effective in the long term to invest in capabilities and reputation today. Let’s look at a pure hypothesis. If a total of 10,000 cars were delivered in the first year and the average loss per car was $1,000, the total loss would be $10 million. However, when you have one million users, you can easily make $10 million, even when the average profit per car is only $10. If the capabilities are all in place and each car makes $100 in profit on average, the payback will be $100 million.
Therefore, everything we do today is intended to support reasonable and sustainable profits when the user base grows large enough. And it is necessary to calculate, based on that, what level of investment is required at present.
McKinsey: NIO Fellows are a very special group. You also mentioned earlier that NIO needs to optimize costs per vehicle by expanding its user base. Does the Fellow model help NIO achieve this optimization? There is always a limit to the number of user interactions one Fellow can maintain. What’s your view on the scaling of the Fellow model?
Lihong Qin: The authenticity of the Fellow model and the upper limit of users served per Fellow will be extremely important topics for us going forward. Regarding authenticity, in our early days, given a smaller user base, we adopted both user referral and other more traditional marketing models. But the proportion of user referrals has been growing higher and higher. Therefore, we need to improve the authenticity of the Fellow model and to adjust our salary mechanism so that Fellows have more incentive to better serve users.
Regarding the upper limit in the number of users served per Fellow, there is no standard answer. You need to constantly experiment along the way—for example, by raising the upper limit through stronger digital empowerment and systematic division of work. When we started the business, I had a discussion with an executive in high-end private wealth management. They created an operating model with very strong customer stickiness, which is quite remarkable. For example, when this company’s customers move to another place, the immediate reaction that comes to their minds is not to contact a moving company but to seek advice from their financial adviser.
How did the bank achieve such strong customer stickiness? The answer is optimized division of work so that everyone plays to their best. In their team, there is one role dedicated to everyday affairs, like a big sister. Such a role may not need special financial knowledge but requires a really high EQ [emotional intelligence]. Another role could be dedicated to helping customers with complaints, and so on. Through the appropriate division of work, the combat effectiveness of an infantry squad made up of seven well-coordinated soldiers will exceed that of seven times as many individual soldiers—an obvious analogy. With the dual focus on digital and organizational capabilities, the potential of an NIO Fellow is massive. Of course, we are still far from there, but the idea gets us excited and shows that we still have a lot of room for growth.
McKinsey: NIO is expanding to European countries, such as Norway, and intends to introduce the user operations system to the European market. How is that progressing?
Lihong Qin: We have just started in Norway. Qualitatively, the user satisfaction and enthusiasm we receive are the same as in China. We often assume that Norwegians are very reserved. That’s wrong. Social needs are common across humankind. When we recruited a local user advisory group before selling cars in Norway, more than 800 people signed up. We picked more than 200 to provide opinions covering our app, NIO House, and the test drive process. We have already started vehicle deliveries in Norway, but in the short term we see a lot of pressure from logistics.
The implications of NIO’s user operations for the industry
McKinsey: NIO’s user operations model has clearly become a benchmark for the industry. Many OEMs are testing models similar to NIO’s or using the agency model. Do you have any comments or advice on these attempts by your peers?
Lihong Qin: The agency model was adopted by different players many times in the past. It all needs to stand the test of time in the end. To meld theoretical beauty with practical business demands could be a huge challenge. After all, the frontline sales and service staff are not direct employees of the OEM.
But, of course, every OEM has a different situation. The average price of NIO cars has reached roughly $70,000. We have planned for a vehicle’s full life cycle value based on that pricing and use this long-term value realization to calculate the investment levels required today. It is true that many companies are now looking at NIO’s way of working. Most of them have been in the industry longer than us and have stronger capabilities.
My opinion is that it may not be a good thing to be very similar to NIO at the execution level. Instead, it could be a good thing if the thinking process and underlying logic are similar to ours, while the actual execution plan looks different. Likewise, Chinese people, no matter how much we make in a year, always reunite with the whole family during the Chinese New Year. One family may choose to have dumplings together and another may go on an Antarctic expedition. The forms of reunion may be different, but the nature is actually the same. We should grasp this nature rather than the exact form.
McKinsey: If everyone’s underlying business logic is the same, wouldn’t NIO lose one of its biggest areas of distinctiveness?
Lihong Qin: I don’t think so. I don’t believe that NIO’s business will be undermined if everyone adopts the same underlying logic. On the contrary, this will accelerate market share gains by smart EVs over traditional ICE2 cars, and that may benefit NIO even more. The automotive industry is not winner takes all. Even with extremely good performance, one player can take only a 10 to 15 percent market share at most.
I never think that NIO could flourish only when others suffer—rather, NIO may also suffer when everyone else is having a bad time. I think the more important thing in the automotive industry is to have consistent logic and strong capability building. We feel, from the bottom of our hearts, that emerging EV companies should all do better. ICE cars still have a 70 to 80 percent market share. Why not go for the bigger market instead of vying for this tiny 20 to 30 percent?
The automotive industry is not winner takes all. Even with extremely good performance, one player can take only a 10 to 15 percent market share at most.
McKinsey: One last question. Cars are very complicated industrial products. Some people might say that as long as the product is good enough, the so-called user operations idea is actually dispensable. What’s your view on this?
Lihong Qin: Good products are very important, but good products alone are not enough. There are plenty of examples in history of companies with decent products that failed in the end. And what is a good product, after all? Different times have different definitions of good products. Within the same generation, the definition could undergo constant quantitative changes.
In the context of smart EVs replacing ICE cars, the definition of a good product may see qualitative changes. Some say that a good car simply means crash safety or a five-star rating. Our products have those, for sure, but we do not focus our promotions on them, as they are entry-level requirements. Multiple car publications have done teardowns of our cars and found that NIO’s quality and materials are as good as those of traditional big-name OEMs—or even better. For us, the pursuit of good products should be forward looking rather than looking at outdated targets.
“Hit a moving target” is a phrase William Li often uses, and it is also what we expect of ourselves. It means thinking ahead about goals and standards—just as in clay pigeon shooting, you must aim ahead along the projectile instead of the target’s current position. Therefore, the big topic for us now is to define the next generation of good products and arrive at what we need to do and how much to invest today.
McKinsey: Thank you, Lihong, for your time and insights.
