Skip to main content

How Volkswagen can reinvent itself as an EV company, with CEO Herbert Diess

A new electric Microbus offers a glimpse at the future of VW

Share this story

Photo Illustration by Grayson Blackmon / The Verge

Today, I’m talking to Volkswagen Group CEO Dr. Herbert Diess. Volkswagen is one of the two biggest carmakers in the world — there’s a battle every year between VW and Toyota for the top spot; they both sell around 10 million cars a year. To put that in perspective, Ford sells around 4–6 million cars a year, and Tesla currently sells around 900,000 cars a year. So, yeah, Volkswagen Group is big.

It’s also complicated: Herbert oversees a huge range of familiar car brands: VW, of course, but also Audi, Lamborghini, Porsche, and Bentley, as well as Skoda and Seat in Europe. A recent management shuffle also put him directly in charge of Volkswagen’s software subsidiary, which is called CARIAD, and, of course, he’s in charge of VW’s aggressive efforts to electrify its lineup and shift all its cars to being EVs.

And that process has its roots in the Dieselgate scandal.

From 2009 to 2015, Volkswagen sold cars with diesel engines that were programmed to circumvent emissions tests, and most countries opened investigations and issued fines. Here in the United States, Volkswagen was ordered to pay $2.7 billion in environmental fees and, importantly, invest $2 billion into clean-emissions infrastructure.

That $2 billion investment turned into an EV charging network called Electrify America, which aims to expand to 1,800 stations and 10,000 chargers by the end of 2025. For comparison, Tesla already has just over 900 stations and 10,000 chargers. If VW can get it right, it can spin the consequences of Dieselgate into a pretty significant competitive advantage as the country transitions to EVs.

Herbert and I talked about all of that and the overall transition to EVs, including VW’s new EV microbus, which is very cool. We also talked about how a company like Volkswagen can move as quickly as smaller competitors like Tesla, whether or not Volkswagen will let others buy the software it’s developing for its own cars, how he’s thinking about self-driving cars, and more. If you listen, you’ll notice a real theme is vertical integration: Herbert thinks Volkswagen Group has to do a lot of things itself instead of depending on suppliers.

OK. Herbert Diess, the CEO of Volkswagen Group. Here we go.

This transcript has been lightly edited for clarity.

Herbert Diess, you’re the CEO of Volkswagen Group. Welcome to Decoder.

Thank you. I’m excited to be here, Nilay. Thank you for the opportunity. I’m looking forward to your questions.

There’s a lot to talk about. VW just announced that the Microbus is coming back as an EV. I saw some photos coming out of CES that look very exciting. The Microbus is obviously a legendary car. I think people are excited about it. 

VW itself has just restructured a bit. You just took over leadership of the software group called CARIAD, which is fascinating. I actually want to start with the basics: VW is a massive major car company that serves lots of businesses and markets. It’s an old company. I don’t think people are familiar with how it works. How is VW structured?

We are kind of a conglomerate of brands. Volkswagen has Skoda, SEAT, and Volkswagen as worldwide brands in the volume manufacturing segment. The premium brands Audi, Bentley, and Lamborghini belong to Volkswagen. Porsche — the top of the range, sportiest cars — belong to the brand. Then we have trucks: Scania and MANs are a totally different business, but also about mobility, it’s about transport. Then we have a couple of service entities. We’re just in the process of acquiring a car rental company — car sharing.

So, as you said, it’s quite big. It’s complex, but I think it’s a good setup for being successful in the future. We have scale, we have technology, and we have exciting brands. As far as how we are going to run the group: we have quite independent groups of brands. Volume is led by Volkswagen. All premium brands are led by Audi. Porsche is basically on its own. Toward the customer side, the brands are very, very independent, but we share technology. We share hardware and software platforms. We share mobility services and finance. That is how we capitalize on the economies of scale.

I feel like I hear this same idea a lot from car CEOs on the show — you have a house of brands. You want all the brands to be differentiated and operate in whatever segments, whether it’s volume, premium, or sports cars. Volkswagen also owns Lamborghini — I could just talk to you about Lamborghini for an hour –– but you’re going to get economy of scale from centralizing platforms and software. You’re going to have just a few basic platforms, and those platforms are going to be expressed by our brands in various ways. That idea sounds great, but the execution of that idea is obviously very hard. If Volkswagen has a different need for the platform underlying the ID.4 than Audi, how do you make that decision?

We have been quite successful in platforms now because they are really widely used. Excluding Porsche, all volume brands — Audi included — are using the MQB platform. MQB is a quite versatile platform we are using all over the world: from India, Latin America, China, and Europe. That has advantages because we can really offer the best platform technology at the best price. This is why we are achieving our goals today. For instance, we are basically winning every test comparison, and we share this platform with Audi. It works. With MEB [for electric vehicles], now we get to the next stage.

The world is changing: it is transforming into an electric automotive world where you will have less differentiation when it comes to engines. You still will have differentiation, performance-wise. You will have a rear-wheel drive, front-wheel drive, and all-wheel drive. You will have different sizes of batteries, but when it comes to the converters for the electric units or such, you get much more similarity than we have today in the industry. This is why we think hardware differentiation will probably be decreased in the future. It’s what we also see from the competition: we all share a similar battery. Now, this is a floor bottom battery that can fit in a huge range of cars; this is common. You have rear-wheel and front-wheel electric drives that are very similar in their design. We have a certain differentiation currently between 800-volt and 400-volt technology. We have Porsche and  Audi are running on 800-volt. Volkswagen and the volume brand are on the 400-volt batteries, but we see a conversion even there.

That is why we decided that in an entirely electric future, one platform will do it. This will be a differentiated platform: from 85 kilowatts to 800 and then 1,000 kilowatts at the top of the range. There are a lot of similarities which we can leverage in scale — even more so on the software side. If you drove an Audi or Porsche and Volkswagen today, you would probably have different hardware and software for navigation, for climate control, and even for the window lifter. That’s not necessary. … Software offers a huge opportunity for economies of scale, still allowing for brand differentiation.

Now, a Porsche can remain a Porsche, even better than today. An Audi can remain an Audi, and Volkswagen will offer a broad range of products, but the basic software stack can be very, very similar. Software is relatively expensive in automotive. Now it’s a one-time expenditure. … We think that we have a good chance to also become very competitive in software if we build a common basic software for all the brands. This is exactly why we founded CARIAD: to get from a world where we bought software with the embedded systems to our own software stack. The brands have all the tools and all the freedom to differentiate for sportiness, for comfort — for all sorts of different use cases. This is why we think we are well set up for this new world, where economies of scale will probably play a bigger role than in this old auto world. Differentiation will probably be more difficult, but through software, it will be very, very possible.

Plus, we are currently sharing the 800-volt platform between Audi and Porsche. The cars — the Taycan and the Audi GT — are so different in the perception, the range, and in the segments they are running that I think the new world will allow for a lot of differentiation and economies of scale.

I want to talk about that in depth, but I just want to ask two more of what I think of as the Decoder questions. You run a huge company — lots of different kinds of decisions to make, especially as you transition. How do you make decisions? What is your decision-making framework?

“The best knowledge is in the room.”

I think my life experience tells me that normally, the best knowledge is in the room, so I’m always looking for different opinions for a variety of insights for discussion to get to the right decision. Those sole decisions from the stomach — and I’ve experienced some of those — are often wrong. I believe in good team decisions, in a competition for the best ideas. That really works. You have the competence in the team room. 

The other question: I interview primarily US-based CEOs. In America, CEOs — especially founder CEOs — are given an enormous amount of executive authority. They tend to run the companies. Maybe they make a lot of those gut decisions. Volkswagen is a European company; it has a big supervisory board. The unions have a lot of seats on that board. Could you explain how that works? Does it constrain your ability to make change as a CEO? For our audience, that setup is very different from any sort of American company.

Yeah, it is. It is complex, and you need more alignment, and a lot of discussion. It doesn’t allow for random decisions because you have to defend your case as a shared interest in the future of the company. You have to talk to your finance audience and convince them that you want to run a business. You can’t make a lot of gut decisions — you can make a few, but then you have to prove you were right. In any business model, you have to explain your story. You have to explain why you have to take the people with you.

It’s probably more complex in the Volkswagen world, where the ownership is quite special, I would call it. It’s quite heterogeneous because of several brands. Porsche is totally different from Skoda’s truck brand. Scania’s commercial vehicle brand is totally different from cars. It’s a complex task, but it’s working. We can demonstrate results because we think we are quite fast in the transition. We decided to go for an electric-only platform five years ago. Some of our peers are coming to that conclusion now. We have strong cash flows from our existing business. We are continuously investing. We have invested several billion in software now. We have acquired a car rental company.

We are investing heavily in the transition — probably more than our competitors. That means that we came to that conclusion that we could convince our stakeholders quite early. It’s not necessarily that we get slow, but we have to admit that those founder companies bring a different speed to the game. It’s a new challenge for us, and we have to rethink our processes, rethink our organization, and we have to improve a lot to remain competitive. 

There’s one specific founder/CEO car company that I might be thinking of: Tesla, obviously. Tesla’s great advantage is very tight integration of hardware and software. Even through the chip shortage, Tesla was able to swap out chips and rewrite the software based on what chips were available in a way that virtually no other car company was able to do.

No, no — that’s not true. We also do that continuously, but you have to imagine, we are much more complex. We have a variety of supply chains, which makes us — and even the platform strategy — a little bit more vulnerable, but we also already did some redesigns. We are still probably a bit too slow. I agree.

You recently took over the software division, CARIAD. I buy what you’re saying about the future of cars. We’re basically going to have skateboard platforms: the battery and motors and various configurations will go up and down in size and performance. On top of that, you can layer industrial design, and then you can layer user experience. The user experience is not a hardware problem. It’s a software problem, and that problem is very challenging. In many ways, you could say that software is actually the car engineering problem now. How do we own all of the components so that they talk to each other? How do we integrate them? How do we build a compelling user experience? How do we deliver software updates? When you turn the car into a computer, you inherit computer problems. Why is CARIAD a division of Volkswagen instead of Volkswagen, if that is the future?

It depends on what you call it. I would say this is Volkswagen. I think you’re absolutely right in your verdict. The differentiation, the competitiveness, and the customer experience will all depend 90 percent on software. Now, I’m still convinced that design, performance, brand image, and marketing will all play a role, but software is becoming really dominant in the integration, but that was always the case. Those successful brands could always demonstrate that the integrity of a product is excellent. A typical Volkswagen had to be a typical Volkswagen for the car to be a success. The same applies to BMW and Mercedes. Product integrity in a complex, open room of solutions has always been a core task. What’s really new is that it’s so decisive to be able to integrate software in the different properties.

This is a very difficult task, but it’s also something not so easy to copy because it’s so complex. A car today already is 10 times more complex than a smartphone. It has 10 times more lines of code than a smartphone. The criteria is extremely different. The real-time environments are really difficult: if you think about braking or steering, you have to build in redundancies, and then you have to make the whole thing communicate to each other. Being updateable means that part of your software runs in the cloud, and you get the continuous updates. ... The question is: who is going to succeed? Who can do it? 

“A car today already is 10 times more complex than a smartphone.”

Your assumption is that we are in a difficult position, and you’re partially right. But I would say that today, no really big company could have a really good car software-wise and hardware-wise, because it’s not so easy to acquire both of those capabilities. The problem we are facing today is that these competencies are between suppliers and sub-suppliers: Bosch, Conti, and ourselves, for example. We need a stronger vertical integration to be competitive. This is where one of our American competitors is probably a little bit ahead of us, but also this American competitor still has to learn some lessons about being trustworthy and reliable — how to deliver on all the promises. 

We have to restructure and rebuild new skills and acquire new skills there to be capable. What makes me confident that we can succeed is that any other potential competitor has the same problem. Even for a software company to really understand the necessities and the complexity of a car — it’s a huge task. I haven’t seen a software company so far succeeding. The big supplier companies have a chance. The big American or European suppliers, Bosch, Conti — they would have a chance because they own part of the software stack when it comes to braking and steering. Still, integration, customer interface, and the cloud backend are huge challenges for them. 

We are in a process. This process will still take a few years until we have companies that are capable of doing the whole stack, run a fleet of cars, continuously update, and deliver on all the customer expectations worldwide. This is a process, and I’m convinced that our setup — strong brands, a hard shift into software, putting all the resources for software together — has a good chance to succeed in that race.

Your point about reliability as well — you just don’t see a lot of Audi owners making YouTube videos about panel gaps on their cars. Actually making the cars is difficult. That is a competency that other companies have to draw up.

“Don’t believe that just any of these startups showing one car at a motor show will succeed.”

You can learn everything now. This American competitor shows that you can do stunning things if you’re really committed. What I’m saying is: don’t expect that you can do this game in one, or two, or five years. Tesla is a case where it took 15 years, and it took a lot of resources — I don’t know, $15 or $20 billion to get to profitability. Don’t believe that just any of these startups showing one car at a motor show will succeed. Some of them we take seriously, but I’m quite sure that not all will have the substance and durability to compete.

I probably agree with you there. Let me ask you about software: VW is very good at making cars, and manufacturing at scale, and making sure that its vehicles are reliable. You sell a gas car. VW’s cost for that car to maintain it over time is zero: if the customer has a problem with a traditional gas car, they’re on the hook for repairing it — buying new tires and replacing parts and such. That actually can generate some profit for VW and its dealer network over time.

With a software car, my expectation as a customer is that you are going to continuously provide me with new updates to the software: you’re going to run a cloud service that delivers new backend features, or mapping support, or whatever it is. You’re going to run a radio network and partner with a cell carrier. Those are all costs that VW has to carry just for me to drive the car that I already bought. Where is the revenue that supports those costs? I don’t think people want subscription fees just to operate their cars — or at least I haven’t seen that market develop yet.

It’s probably a bit premature to say that pay-per-use is not going to happen because that subscription framework is really taking a lot of space in other businesses. We also see that the sharing economy is getting momentum. This is basically also why we bought a car rental company: we think this business model of sharing usage will increase fast. Sharing will not substitute ownership, at least not in the next 10 years or so, but it will grow a lot, and this allows for additional revenue streams. I fully agree. We have to run the car, and we have to provide services, but you have to see that this allows us an additional source of income.

Imagine: for a long time, we did not have access to a customer as a company. The customer access was exclusively with our retail network, basically. What we experienced over time was that people walk away from our retail outlets and go to third parties to substitute some of the spare parts or buy new tires. Now, we have a new opportunity to be in direct contact with the customer, which is totally new for us. It’s a quite violent experience, because we knew that we had customers, but we didn’t know them. Now we’re getting to know them. We get to know their preferences, their anger, and what they love in the cars. We get a much better customer understanding. We also get access to the customers — now we can talk directly to them, and that allows us to deliver better service for the customer over time.

For instance, that access allows a very critical figure in our business model: residual values. We are financing a large proportion of our cars — about 60 to 70 percent in some markets; in the US, it’s 90 percent. Residual values offer us more important information than retail prices. You can imagine that with updates, upgrades, knowing about the car, the battery status, you can improve your residuals of the car quite considerably. You can know which extras you need, provide for some features, and you can keep the car fresh. My perception is that there are far more opportunities than risks in these changes. The risk is that you have to acquire new skills; you have to learn to talk to the customer, which is a big challenge for us. You have to learn to deliver good working software updates that keep the customer happy and pleasantly surprised. This learning is a very, very healthy process, and if we take the opportunities, there’s more upside than downside.

That opportunity in front of the customer — does that look like commercial opportunities? Are you going to sell more things in the car?

Not only. Right before Christmas, I had a chance to drive the next generation of the software, which we are going to deploy to some of our ID cars in the lineup. It’s a huge improvement in the driver assistance functionality — it’s really empowering. Even in Europe, this driver assistance is better performing than some of our American competitors. This experience of being able to improve the car and the user experience in the hand of the customer is such a new thing. Normally, we would have worked until launch. Then we’d prepare a model year update, and then update after another three years. Then we’d forget the car and think about the next.

This idea now of getting the car better in the hands of the customer is really exciting. It’s not only about additional business opportunities. It’s about additional satisfaction: getting close to the customers to understand them better. This will change our business model.

Talk to me about that. That’s a big change: you have engineers working on the ID.4’s driver assistance stack. You have to pay them — and I assume you pay them well — which is a cost. Then the engineers have to deliver the stack over a network. You have to provision and provide for that wireless network that will hit the car. That’s all just cost. I agree with you that this is a great benefit to the customer — but at some point, you have to line up revenue against that cost. That is the heart of the business model change: instead of assuming that there’s a two- or three-year upgrade cycle, you’re weighing the cost of supporting the cars for longer and comparing those costs with the revenue associated with that platform support. Are you realizing any of that revenue yet? Are those plans in place?

We have that revenue in mind for sure as well. Customers will be prepared for some features they didn’t buy at the start, probably after a few years or after a few months — even if they consider taking another option or another software feature, the customers would be prepared to pay a monthly fee or a one-time expenditure. We estimate that the big difference is that we keep the residual value of the car higher, we keep the car fresher, and we get a better customer experience. When Apple started the smartphone business, the income from their services, from the apps, was very, very low. Now, it’s—

Very high.

Reasonable. Yeah. That income is reasonable, but still, they are making a large margin on their hardware because people are prepared to pay for it because of the services they get. I think this logic will also apply to our industry. People will not buy any more cars that will fade out quickly and lose residual value fast. State-of-the-art will be a decisive customer criteria. At the end, it’s about competition: if we can deliver better service, better cars, and a better customer experience, then the customers are prepared to pay for mobility. We see that mobility has a value. Customers are probably even prepared to spend more on mobility because they love mobility. They love cars; they love individual mobility.

Let me ask you a very reductive question: do you think the future of Volkswagen is a hardware company, a software company, or a services company?

It’s combined. It’s all three. We will remain a hardware company because we need excellent manufacturing skills as well as finish quality in the future designs. We need direct customer contact, so we will remain part of that. Software will be decisive for the differentiation and for gaining the economies of scale, so we will become a software company as well — talking to the customer and delivering day in and day out. We will also become a service company because mobility will change. Pay-by-use sharing fleets, mobility services, and autonomous vehicles will change mobility quite considerably. People may not own a car anymore: they may just use cars, and we will be a part of that game. We will definitely become all three of those companies, so we will remain a bit complex, but a lot more vertically integrated. 

I’ve talked to a lot of CEOs of companies that make components of autonomous cars. We had the CEO of Luminar on the show and the CEO of Argo AI. You partner with Argo; they make very advanced autonomous systems. There’s even an ID Buzz prototype driving itself around various cities. Argo AI is a software supplier: they make a package, and then they sell it to you in the way that your traditional suppliers would sell you seats or door handles. Do you think that’s something you have to vertically integrate, or can you still have suppliers like that?

Actually, we don’t treat Argo as a supplier, we own a major stake in Argo. We have a lot of equity in Argo, and we have a very close and trustful relationship with the management team there. We provide knowledge and integration into the car. We run the fleets. In the end, it’s not about software: it’s about good integration between the right sensor setup, the right computer hardware to make all the perception, the planning, and then finally making the right decisions to run the car safely. Our aim is to be able to drive a car as Volkswagen. We have two areas: one is pushing robotaxi technology with Argo. This involves shuttle services, limited areas, relatively slow speeds — they are typically ODD, which is learned and programmed. Then it goes area by area and city by city.

The other way we are pushing is private mobility: we have the Audi team and CARIAD team working on that because we think that autonomous driving will not only cover this area of robotaxis, but also private cars. Step by step: first we tackle driving at level three or level four on open highways — German autobahns — and then we get into more complex areas. This will happen at the same time. It’s two different technologies. Some of our approaches involve sharing computing technology. Some sensors are shared — but ultimately those tasks are so different. At one stage, you think about a completely autonomous car, which is able to handle an area right from the start. On the other hand, you think about driving at higher speeds in a less complex area in a car that is also able to take over from a driver for a certain period of time.

That leads to different approaches, to different sensor concepts, to different algorithms, and to different AI learning tasks. That is why we are pushing both. In both areas, sooner or later, we clearly have the target to be able to drive the car. Those [robotaxis] have three different business layers. One is the platform — the customer interface. You have to book the car and pay for the car, and that is on a platform level. Then someone has to run the fleets, someone has to build the cars, and someone has to run the autonomous driver. This is currently Argo AI, which is why we have a stake in that company: we also want shared ownership in this area.

On the other side, yes, we also have suppliers there. We’re working with Bosch in the other area, but we want to become able to really drive the car, because we think the big differentiator is — and the big step change in the industry is — are you able to take over the responsibility for driving the car? That makes a big difference.

When do you think Volkswagen will sell a car without a steering wheel?

Difficult to say. First thing will probably be the fleets we are running now together with Argo AI. We should get started in 2025 or 2026. Now, those will still have the steering wheel for getting the car back from certain locations, but towards 2030, we should be able to sell cars without a steering wheel for transport services. Now, for private use, most of those cars would probably still have a steering wheel in case of heavy weather conditions, for example — or let’s say you drive into a very remote area where the car is not able to handle it.

You’ve talked a lot about software. One of the highest margin things you can do once you develop the software is sell it or license it to someone else. This is maybe the foundational conflict in the traditional computer industry: Microsoft is very successful as a software licensing business. Apple is very successful, but they will never license anything to anyone — Apple wants to sell an integrated product. Android is very successful as a licensing business. Where do you land? Once you’ve built all this capability, will you license the software to other companies?

“If you compare us to Apple or Google, currently we are running like Google.”

It’s very difficult to predict. Right now we are very open. We are sharing the platform with Ford because we think that the economies of scale we can generate are beneficial for both companies. Ford in Europe is using our MEB platform, and that applies for hardware and software. We predict that there won’t be too many really self-sufficient software stacks in the world because they will be really expensive to maintain and update worldwide. We are currently considering sharing, and would be open to partnering, as long as that doesn’t slow us down. We are prepared. If you compare us to Apple or Google, currently we are running like Google. We are open to third parties to use our technology.

You mentioned Google and Ford. I had Ford CEO Jim Farley on the show. Ford signed a big deal with Google to do the user interface, provide mapping services, application support — all that stuff in the car. What Jim said to me was, “Look, Ford isn’t any good at this. I’m done throwing money at it when I can just buy it from Google, and we’re going to do the stuff we’re good at.” Are you thinking about similar kinds of deals where you would hand over navigation to Google Maps or something like that?

I can’t compare us to Ford, but we think we are not too bad at it. If I compare our routing capabilities, which we currently have in the car based on map data and our own routines, in some cases, our routing is better,  more accurate and leads you faster to the target than Google does. If we compare charging infrastructure data, we absolutely can compete. We believe that it’s necessary to keep control on this domain of user interaction, because now it makes a difference if the user talks to Google versus talking to Audi or Porsche. We can deliver a specific service. CARIAD will help us because bundling all the synergies in the group is 10 million cars every year. We have a chance for the right economy of scale. We are also open to third parties to deliver: we will not close the windows for Google, because we are working in some car lines with Google support. We have a good relationship, but Google knows that we will try very, very hard to stay independent. 

I want to make sure we talk about the Microbus and EVs. You just mentioned EV charging networks, and you’ve announced the new Microbus that seems like it’s coming soon. If you think about car design, that skateboard, is the Microbus the sort of thing you can just execute now faster? It has a nostalgic design, and you can just make it happen as an electric vehicle?

Yeah, definitely. The electrification gave us the chance to bring it back. The Microbus in its historic time frame was quite unique: it offered a huge amount of space on a very reduced platform size. It’s a relatively small car with a huge interior. As auto technology developed over time, cars have become much more complex. Engine technology has become much more complicated, much more bulky, and much more voluminous. Over time, the car lost its initial room efficiency. Room efficiency means you have lots of room to move, to sit, to work, to rest, to camp on a reduced size of a platform. Only electrification and the MEB could bring back this opportunity.

In this new Microbus, the interior size is much bigger than the existing T7 bus, which we are offering on the platform of a reduced size even compared to the current ICE [internal combustion engine] car. What we really bring back is that initial room efficiency rather than a huge space. You have very low angle overhangs over the front, very much like the first Microbus. You typically need a one-box design, and that is why electrification brought back this opportunity. When I joined Volkswagen, bringing back the Microbus was a dream of mine because it is probably the most emotional car the world has ever seen. It came at the right time: in the hippie era of the ‘60s, the Microbus was a cultural item. It created so much positive awareness, not just in the United States, but worldwide — though not in China because they didn’t experience it — but in the rest of the world, the Microbus is so nicely perceived. It’s so tied to positive memories of Volkswagen that it’s really a brand essence. From my first day at Volkswagen, my wish was to bring this car back to life, and only electrification gave us that opportunity.

There were probably six or seven attempts to bring back this initial icon, but the real chance came with electrification. We pushed it through and I’m really excited. I think America is waiting for this product. In Europe, many people are looking forward to it. I think the team has done a great job: the car will perform very nicely as a modern car while still awakening all those memories of freedom from the very liberal time of the 1960s.

During that time period, the Microbus and the Beetle were very accessible cars. They were not expensive. How much is the new Microbus?

I can’t say. The American team has to decide on that. It’s premature because the car will only come to America next year. It’s not yet decided as the car will have the newest features and a lot of range. It will not be as cheap as the first Microbus.

One of my pet peeves across the entire car industry — everyone is doing this — is the announcement of an EV that is not going to arrive for a year or more. Why do you think this happens? You’re announcing the Microbus in March, but it’s arriving in 2023. Will we wait a year, or even longer?

In America.

In America?

Yeah, in America, but we announced it because it’s hitting the ground here in Europe. I’ll give you the background. When I pushed very hard for this project — I have to say that I really fought for this project because it was very important for me. There was quite a lot of skepticism when we made the decision in 2015 as to whether EVs would work at all, and whether this would be a good investment or a bad investment. You have to imagine that this was still during the Trump administration: our American colleagues had been very conservative on their volume perspectives at that time, which led to a situation where the decision to bring the car to America was made very late. It required some modifications for the local American specs, which is why the car is coming much later in the US. I feel sorry about that, and I apologize for it, because I think the car is really important for America.

I can tell that you love this car. When is it coming out in Europe?

We have our market introduction in the first quarter, and then it should come midyear to the customers.

By the way, I want this car, so the faster you can get it, the better. But I want to talk about charging infrastructure. We just had United States Secretary of Transportation Pete Buttigieg on the show. He wants to expand infrastructure. He wants to subsidize chargers in places where they wouldn’t necessarily be profitable now. You run Electrify America — that’s a Volkswagen project that came out of the Dieselgate settlement with California. Do you want to run Electrify America? When you have to do that because of a settlement related to a scandal, there’s a part of me that suspects that you’re doing it begrudgingly. There’s another part of me that thinks of this as a huge opportunity to own the major charging network in America.

From the beginning, of all the penalties we received in America, I most liked this idea of a charging network because the idea of paving the way for our electric cars — or even making electric cars viable in America — would ultimately be an asset for us. I really liked it, but still, this was a big investment. It’s not going to pay off too soon, but over time, we are noticing that charging will be a business model. There’s not a lot of investment going into charging now. Of all the petrol companies, Shell is currently the biggest charging network. There’s a lot of business potential in charging. We also see that our incomes are gradually improving because the usage is getting higher. With every electric car we’re selling, it’s getting better.

We are not doing this only in America: we have been founding charging networks in joint ventures in Italy and Spain; we are invested in IONITY, which is a fast charging network across Europe. We have also invested in a fast charging network in China because we think it’s necessary for electrification. By the way, our American competitor is showing that it can be done. What we see is that it’s a huge opportunity in charging. Just a few days before Christmas, I had the opportunity to open a charging hub for Audi, which is a premium charging center in a location close to Nuremberg. This will be a wonderful experience: you have fast charging, you get there with your Audi app, you plug in your car. Then you enjoy a lounge atmosphere where you can wash your hands, you can have a drink, you can answer your emails. Charging will be part of the mobility experience and there will be a premium experience covering these areas.

That is why we think charging will play a major role. Supplying the energy, dealing with energy — it gets much more complex because the prices for electric energy are fluctuating. …. You have to be able to buffer, and cars will provide a huge opportunity to be able to contribute to net stability [in the electric grid], because you basically can load the car when there’s a surplus [of energy]. We will experience times where you get your refueling for free, because there’s just so much energy surplus because of so many renewables in the network. This is a new business model and we want to participate. Yes, in America, this program was not entirely voluntary at the beginning, but ultimately I think it was the baseline for one of our strategic approaches into the EV business.

How do you think about the pace of investment in building the charging network there? You’ve described perhaps premium charging experiences. You’ve described partnering with power companies — maybe changing how you meter or pay for that. However, you still need lots of chargers and lots of places to make all this work. I’m looking at the numbers here. You expect to have 1,800 fast charging stations and 10,000 chargers in America by the end of 2025. There are already 10,000 Tesla Superchargers. Is that fast enough?

In some countries, we are already ahead of Tesla. That’s fast enough because it has to match the growth of the fleet. But you’re absolutely right: we can’t do all the investment out of our cash flows. We are partnering and we are also looking at some areas, for instance, batteries. Batteries are a much bigger concern for me because the investment there is still bigger than in charging. We will even consider IPOs, or allowing certain parties to participate in the growth, because there will be a lot of growth in the area of EV charging and batteries, and I’m sure that we will find investors to participate, but we can’t finance all of the necessary investment out of our cash flows.

One of the ways to reduce that investment is to standardize: either standardize batteries across the industry — which doesn’t appear to be happening — or standardizing the plugs and the charging infrastructure, which is slow going. There is some interest. Every now and again, Elon Musk talks about opening Superchargers. We had the CEO of Polestar on, he says, “Yeah, they talk about it, but it’s never going to happen.” Do you think we’re going to see an actual push from the industry to standardize charging, especially fast charging?

Aside from this proprietary network, which is quite established already worldwide, the rest is really usable industry-wide. When it comes to billing or charging, there are different standards and different price tags. Here in Europe, you can use every pole that is not Tesla. It’s a positive move from Tesla to consider opening up their network; Tesla has to decide on that. My perception is that the charging network actually is growing very, very fast. At least in Europe, I drove electric cars basically the whole year during 2021. I’ve driven about 15,000 kilometers. I was in Italy yesterday, I came back through Austria. It was only once that I had to wait on a charging pole. In the US, you can go from the West Coast to the East Coast without stopping. I think we overstate the issue — it’s growing really fast, and there’s so much investment going in. It’s not only auto: now everyone is investing in charging networks.

The utilities companies, the petrol companies — now they also are changing swiftly. Currently, for instance, we agreed with BP. They’re running the biggest refueling network in Germany. We will convert all the fuel stations into charging within a two-year time period. There’s a lot of investment going in, and I think it will be in line with the growth of EVs. The bigger constraint might be batteries.

You actually just shifted battery suppliers, right? You went from LG Chem to CATL in China. Is that —

No, we didn’t shift. We have to use all of the suppliers because we’re growing. It’s just so much that no sole company would have been able to do all the investment. Just recently, we decided to build our own battery group. Thomas Schmall is leading that, a colleague of mine. We are also looking, for later, at some external companies — we will finance the first plants ourselves, but we will invest in batteries because so much growth is necessary, and we couldn’t find enough suppliers to do the job.

This is a kind of consistent theme of our conversation — needing to vertically integrate. Every area we talk about, your answer is, “We’ve got to do it ourselves.” 

I mean, not 100 percent, no — but whenever it’s necessary. The new business model requires it. You can’t run electric cars without networks and without batteries.

We see all this demand. Is there a reason that the industry has not created suppliers to meet that demand, and all the companies are kind of doing it themselves right now?

A good question. I have to tell you a story there, because I used to work for another company in 2012 or 2013, when I first visited a small company in China, which was called CATL. CATL would then have a turnover of, I think, 200 to 300 million. They would do pouch cells for smartphones for a big American customer. I visited them and said, “I need battery cells for cars — first for plug-in hybrids, and then for electric cars.” The guys looked at me and said, “We can’t do cars. It’s too big. We don’t have the facilities.” They were really not prepared, but I could convince them. Robin Zeng is the CEO and founder. Once in a while, I follow up with him. He made an exceptional transition. He’s now the second biggest battery manufacturer. Market capitalization is about $200 billion, something like that. Extremely fast growth and innovation at CATL.

I tried the same with the German first-tier suppliers. They would have been well-prepared. Bosch, Conti, you name them, I tried the same approach. I knew that we would need a huge amount of battery cells. The turnover of our battery cell business perspective to 2030 — when we would build 50 percent EVs — we’ll have a turnover only for the internal demand of 20 billion, which is half the size of Bosch. It’s a huge business, but I have to say, I could not convince any of our first-tiers besides CATL to invest in batteries, which I regret. Now we have to do it ourselves. We have invested in Northvolt, which is a European startup. We are working with the entire industry — LG, Samsung, CATL — because we need the quantity of batteries and the quality of batteries.

This all tracks with what you’ve said: VW will stop selling gas cars in Europe by 2035. Is that data driven by the availability of batteries and infrastructure? When do you think you’ll have enough batteries? When will there be charging infrastructure?

Not necessarily. I think this transition into EVs has certain constraints. I think the plan to get to 50 percent EV by 2030 is extremely ambitious. If we own, here in Europe, about 20 percent market share, for that 20 percent market share to maintain 50 percent EVs, we need six gigafactories. Those factories would have to be up and running by 2027, 2028 to be able to deliver on our 2030 goal. It’s close to impossible to do that. I have high respect for our team who is facing the challenge because you have to buy all the machine tools. You have to build the plants. You have to find the locations. You have to train the people. You have to make sure that the supply of the raw materials is safeguarded and is good. This is huge. We are only 20 percent of the market, so six plants. Europe needs 30 of those plants. Each plant is two by one kilometers. Huge quantities of raw materials have to be moved. This will be challenging. Then to let’s say, get from 50 percent to 100 percent, it still will be a tremendous challenge. It’s not just saying, “Let’s switch off ICE cars.” It’s just impossible.

“Electric cars only make sense if the energy is renewable — only if the energy is really green energy that comes from wind or solar or nuclear.”

The second thing is that electric cars only make sense if the energy is renewable — only if the energy is really green energy that comes from wind or solar or nuclear. As long as we have nations based on producing electric energy from coal, it doesn’t make sense to sell electric vehicles there. Think about Poland: I believe it’s 100 percent coal currently. Before we sell electric cars, we have to convert the primary sector into 100 percent renewables. It’s already fine in France, in Norway, in Austria, in Europe, in some parts of America, and in some parts of Canada — but this has to go hand in hand with the conversion of primary energy production. This requires time, and this is why two ambitious plans will not work. That will be counterproductive, even, because running EVs on coal-fired car plants is even worse than running gasoline cars.

America has a very complicated power story. You have not announced the date for the discontinuation of gas cars in North America. When do you think that will be?

It’s hard to say. Some of our peers are making statements, and I’m not sure whether they should do that because ultimately, this is a decision a car manufacturer cannot make alone. You have to ask yourself: do I have the primary energy? Do I have the networks? Are the people prepared to switch to electric cars? Do they have the right incentives? If you look in Europe, diesel now has been at 90 cents or one euro for a few months. It’s very difficult to compete with an EV car against such low prices, so this has to go hand in hand. The rollout of EVs will depend on the legislation and on the buildup of renewables, and this will be dependent on state policies and on a comprehensive policy — not individual decisions of car manufacturers.

We are second or third position in the Latin American markets — in Brazil and in Argentina. They run their cars on a base of ethanol, which is basically CO2-free, though not entirely. It’s bioethanol. For them, it just doesn’t make sense to switch, so far, to electric cars. They would still need combustion engine technology. There’s a lot of buses now in India becoming electric, but 90 percent of their energy production is on coal. We have to change. We still had a coal-fired power plant here in Wolfsburg, in our headquarters. We decided some three years ago to dismantle and substitute it with a gas engine, which allows for CO2 savings comparable to 870,000 cars per year. That is huge.

I probably received a call every week from countries like India: they wanted to buy this coal-fired power plant, and just to build it up somewhere else. We could avoid it in the end, it was not easy, but think about it: it doesn’t make sense to electrify the mobility world if we don’t first make the primary sector CO2-neutral. The world is not the same. In France, they have seven grams of CO2 per kilowatt hour, because it’s all nuclear. In Poland, they have 1,000 grams because it’s all coal-based. It’s the same in South Africa. The rollout has to be staggered. It has to be primary energy, cars, networks, then primary energy, then cars.

One thing that’s facing the entire car industry up and down is the chip shortage. We’ve seen various ways of dealing with it. Some automakers are holding their cars back. Some like BMW are shipping cars without features and offering $500 credits because the cars don’t have touchscreens. My father-in-law just bought a pickup truck, and the heated seats don’t work. He has to go back when the chip is available so they can install it. There are a variety of solutions to the chip shortage. How is VW dealing with it? When do you see it ending?

Good question. It’s always the time between the years, as we call it — between year end and start of the year — to rethink and reconsider. With the achievements of 2021, I was really, really happy. I think we made a lot of progress. The teams performed extremely well. We gained market share in some areas. My outlook for 2022 is very, very positive, except for the chip supply. We suffered a lot. We lost market share in China because of the chip supply, and in Latin America because we prioritized the higher spec cars and the premium brands. That was quite a difficult situation, and we have to see that this is a structural problem which will continue. We will still have constraints. We will manage better than last year. We had quite a good start the first two weeks, but there are constraints visible. Those constraints are structural because the chip demand is just growing so fast, and we are very dependent on second and third tier supply from the chip manufacturers.

We are trying to do our best to sort, to help, to improve, to get higher margins for the chip industry, because ultimately it makes a lot of difference whether the same chip ends up in a refrigerator or in a car. In a car, it gets a much higher contribution, so we’re trying to really push and squeeze. We have a task force running. I’m optimistic that we can do better than last year. Will we be able to deliver all the cars we wanted? No, because we have a really, really strong demand. We will have some constraints still, which is my biggest — I wouldn’t call it a headache, but I have a lot of focus on this area, I have to admit.

When do you think it’s going to even out?

I think we will continuously improve over time because capacities will shift into higher margins, which are cars, but it’s a structural problem, and there’s a lot of investment getting into chip manufacturing. It’s a three-year investment period until we see that. I’m quite sure we will see constraints through 2022. Hopefully, by 2023 we will get back to satisfying demand. Then I should have a new setup because we are digging much deeper into chip design and chip supply. We are influencing much more and getting into direct sourcing. We have dialogue across our first tiers into direct chip production. We’re doing a lot. We will improve.

Let me bring this all the way back around, then I promise I’ll ask you an easy one. You started by explaining how the company is structured and where the growth is. You said that VW is not very strong in America right now, but that you want to get stronger. You think you have a product portfolio, and you’ll go from 4 percent, to 8 percent, to 10 percent market share. Will that growth come in EVs, or will it be a mixture?

It will mostly come from EVs. We are trying to push the market share with EVs. I think we have a good chance with products like the Microbus and the ID models, which are well perceived, and are selling very nicely. There are a few more models to come that would suit the Americas. We are considering some more Audi models for America. It’s not yet all decided.

And those are all EVs.

Yeah. It’s not all decided yet, but I think this electrification allows us for a redefinition of our market position, of our brand perception. I’m 100 percent sure that it gives us a new chance for getting back to where we should be: a volume manufacturer in the United States with a proud manufacturing footprint in Chattanooga, becoming much more of an American company once again. We are too Chinese and too European for the moment.

Here’s the easy one: what is your daily driver? If I had your job, I would only drive Lamborghinis.

That’s at least what I’m driving. I drive Urus once in a while — people would already look at me. No, actually, this year I mostly drove electric cars, usually an ID.3. I have two: one here in Wolfsburg and one in Munich. It’s also so I can share the customer experience — how is the car working? How are our networks working? I like to drive electric cars: one or the other Audi models. I have the privilege that I can swap cars, which is really good. I have to say, I have kind of a fallback: I will get a Golf R as my next car for the next half year or so, which I’m also looking forward to.

Very good. If it was me, it would just be Lamborghinis. Herbert, it was great to have you on Decoder. Thank you so much for making the time.

Yeah. Thank you, Nilay.

Correction January 18th, 1:30PM ET: The introduction of this post originally stated Tesla sells “around 500,000 cars a year,” which was the 2020 figure. It has been updated to reflect the 2021 figure of around 900,000 cars. We regret the error.

Decoder with Nilay Patel /

A podcast from The Verge about big ideas and other problems.

Subscribe now!