Upcoming IP Investor Dialogues: Spirax Sarco, FAST, COST, CPRT, POOL, SNOW, DDOG - Register your interest at help@inpractise.com 

Anonymous

{audio:00:01:15} I think a good place to start would be to lay out your experience and the history to the Business Innovation unit within Daimler?

{audio:00:01:29} In the late 2000s, Daimler looked into new business ideas and new business models to complement their core business of manufacturing and selling vehicles, from passenger cars to vans, heavy trucks, buses and so on. The board decided to give the authority to a group of younger executives with a budget and basically a white piece of paper. This group embarked on assembling ideas, mainly in things that could compliment the core business. That was the only guideline they had to follow, “Don’t come up with something like medical systems. Keep it somewhere in relation to the core business.” So, the group started looking into megatrends at the time. As you might recall, in 2007/2008 that was the breakthrough of the smartphone, always on and always available. The other megatrend was urbanization. The more we understood about China, the more it became clear that, particularly in Asia, people tend to move into urban centers. We looked into the consequences of growing population in these urban centers, particularly what that means in terms of mobility. Then we had the first trend in sharing, if you remember Airbnb came to the market around that time. There was a headline at the time that said, “Sharing is the new owning.” Another megatrend was a restriction on the availability of fossil fuel resources, climate change, and so on and so forth.

The group then looked into the question, “What does this mean to our business?” Out of the megatrends is there anything that can attack our main core business in manufacturing and selling vehicles? One of the areas that became particularly interesting was urban mobility, that's where this car sharing idea started, so to speak.

{audio:00:04:36} Daimler claim to have 3.6 million customers and roughly 15,000 vehicles. Could you define the concept of car sharing?

{audio:00:05:23} We have to go back a little bit in history, because the model as we know it today evolved over a period of a couple of years with different pilot projects in various cities. We started in a small environment with not a lot of supporting technology just to understand what the typical utilization and rationale of the customer is. How many customers per vehicle do we need to attract?

It became quite clear that two main things are necessary for some kind of success: that's accessibility, meaning it must be very easy to find and access a vehicle; and the other thing is availability. You need to find a sweet spot where you hopefully have enough vehicles on the road so that you can satisfy the demand that you have generated for this new concept. It became clear through these pilot phases that you need urban conurbations of a minimum 500,000 people, just to get going. You need to have enough vehicles on the road so that you can say that they are ‘usually available’ and on the other hand, for the economics to work, you get enough throughput, you get enough people who have the demand for sharing a vehicle. That's why, if you look into the history, there are a couple of cities where Car2go and I think DriveNow as well, tested and then moved out. Copenhagen because it's a typical bicycle city. Stockholm didn't work. London didn't work. London was for other reasons, because of the difficult legislation in London with 35 boroughs, which all have different rules and regulations. You couldn't make Car2go available in all 35 boroughs, so it didn't make sense because Car2go was a free flow car sharing model, and not a station-based model.

{audio:00:08:11} So that means you can use a mobile app or GPS to locate a car and then you use the car for a certain period of time or distance and when you park the car it's within a certain distance of the city?

{audio:00:08:28} Correct. You need to define a certain area of operation. So, you want to make sure that you can manage the vehicles. At certain times of the day you have to make sure that you move the vehicles back because, based on the data that you have generated in the past, you know that there are certain peaks in demand in certain parts of your area of operation. The typical thing is that in the morning you want to have as many vehicles as possible in the living quarters of a city. And then people go to work to the city center and you want to make sure that you have an equal distribution. When it comes to the end of the day, you need more vehicles in the city center because people like to use a car to move back.

{audio:00:09:32} Is there a manual element to this, in transporting the vehicles and making sure that accessibility or availability is in the right spot at the right time?

{audio:00:09:43} There is a manual element to it. You have a certain number of jockeys, as we call them, who move around the city and identify where there is a bundle of vehicles, where the number of vehicles that are parked there exceeds the projected demand at a certain time. It's a labor cost. You have to meet these labor costs to make sure you have a relevant accessible number of vehicles.

{audio:00:10:26} How do you look at the limitations in adoption? You mentioned how there needs to be a certain density of population, 500,000 people, so that's one key criteria, plus there needs to be a certain structure of the city in terms of transportation and infrastructure. Does that limit how you can sell the service?

{audio:00:10:53} To some extent. Car sharing as we know it is based on the theory, and I think the theory is still valid, that people like individual mobility. They want to move around when they want, not based on a timetable. But I will say that the better the public transport in a given city, the more difficult it is to establish car sharing as an additional form of transportation. In some cities it might be an add-on, because you have significant restrictions on parking or enormous cost when you park a car or you have to use a garage or something like that, where people consider car sharing as a very welcome addition to public transport or bicycle sharing or what have you. I think the mix of mobility offers in a city plays very much to the attractiveness of a city. I think car sharing can do this, can increase the attractiveness of urban centers.

{audio:00:12:21} What is the typical use case for car sharing?

{audio:00:12:25} I think from a user perspective, it is a very flexible individual form of mobility in addition to other forms of transport within the city. We have tested other forms where we offer car sharing for particular user groups i.e. when you have a company with many campuses around an area, corporate car sharing. We tested it and established it in China, where you make car sharing available to your employees to move around these campuses to go to meetings, or maybe one place doesn't have a cafeteria so people move over. So there are different possible use cases, and the team are still in exploration mode. You could expand the car sharing into light commercial vehicles. Where you have a user group that has mobility demands at different times of the day so you have a pool of vehicles to go and do their business.

{audio:00:14:05} So what's the state of B2B, commercial and consumer roughly in terms of market share?

{audio:00:14:26} I would say at the moment it's 90/10. So 90% is B2C and 10% is B2B.

{audio:00:14:37} Those B2C journeys, are they to work or for leisure? What is the average distance for example?

{audio:00:14:47} I hope my knowledge is still valid, but a typical ride is somewhere between 10 and 15 minutes or in kilometers around 10 km. So it’s an urban form of transportation. Although Car2go and ReachNow, as it's called since the merger with DriveNow, offers daily rentals as well, or hourly rentals. But this is just in addition to the base use case of flexible short distance, short time car rental, if you like.

{audio:00:15:41} Do you have any idea about the growth in users?

{audio:00:15:48} The number of users you mentioned before. Car2go claims that they now have 3.8 million users across all the locations with roughly 13,000 or 15,000 vehicles on the road.

{audio:00:16:11} That hasn't grown much in the last few years?

{audio:00:16:14} Not in the last 2 years, I believe. I think there was a certain kind of cautiousness. It generates a lot of discussion with the board, as you can imagine. On the one hand, most users are very satisfied with these car sharing offers, on the other hand, it’s very capital intensive, because you have to have a number of vehicles on the road. It's still difficult to break even because it needs some time to establish itself. According to my knowledge, there are a few cities who have broken even on a variable cost basis, and maybe the board was a little hesitant to invest more aggressively, and let the team focus on consolidating those locations where they are present. But this is speculation because I'm not in the inner circle anymore.

{audio:00:17:35} Could we talk about the economics? Let's say we have a 500,000-population city, how many cars do you need to place there?

{audio:00:17:55} Typically, in cities of that size you have a pool of 400 to 600 vehicles on the road. You aim to get at least 15 rides per day. That was our base assumption at the time and that was based on a smart car only because it was the most typical urban car and the cheapest car in the portfolio. So, at the time, to break-even in a city the aim was to have the average revenue of €45 to €50 per day per car. That translates to roughly 10 rides per day at €5, at a net revenue, without value added tax and so on and so forth. At the time I believe it was 30 cents per kilometer and a little bit of time when you park and so on and so forth. So with that, you have 5000 rides per day, which is a micro market share if you compare it to rides on public transport, bus, train, underground and so on. But that was the rationale behind that. DriveNow came to market a little later, but more or less, it was a copy of the Car2go model. There was very little difference in pricing structure and very little difference in the number of vehicles. So if you combine both, you just double the numbers. If the economics then stands, I can’t judge. It’s tough.

{audio:00:20:17} Those numbers are at maturity?

{audio:00:20:42} Sure. As I said, that was our base assumption for breaking even. I assume that one of the reasons why Car2go and DriveNow left certain cities is that they couldn't achieve it. So then you don't want to just stay there to have a presence and lose your share. At the end of the day, these companies offer car sharing not to better the world, they offer car sharing to make money. I think this is always the ultimate game. Now, let's accept that because it's still relatively new, almost the same age as Uber and so on, all these new mobility offers struggle to make a dollar, or a pound, or a euro in profit. They follow the theory that, at some point, you are the market leader and then you can maybe optimize your pricing structure and move slowly towards profitability.

{audio:00:22:13} How do you look at the cost structure then? What is the split between fixed and variable costs?

{audio:00:22:57} Well the fixed costs are clearly the order and payments system. There is a certain variable element to it. The more customers you have the bigger the costs, probably you have to grow with your system as well. Then clearly you have the vehicles as a fixed cost. You have certain management functions that you have to establish. You have a base fixed cost in the logistics cost because you don't know how many rides these jockeys have to make. You have to have the personnel on the ground.

{audio:00:23:43} Are they full-time employees?

{audio:00:23:47} Or contract employees. But even if you have a contract employee, if you constantly employ 20 heads from an agency that's a fixed cost to you. It only goes away if you stop the business, because we’re not in autonomous cars yet that can drive themselves back to the area of operation where there is a higher demand. There are lots of fixed costs: you have insurance, you have road tax, and in some cities you pay a flat fee per month per vehicle to the authorities for the parking element of car sharing.

{audio:00:24:37} So it's mainly all fixed costs then?

{audio:00:23:40} Mainly fixed costs, yes. The variable cost is fuel if you like. That is one of the obstacles of car sharing. And I think it’s not that different a challenge from what a traditional rental company has.

{audio:00:25:04} Does the car sharing company pay the fuel for the car?

{audio:00:25:09} Yes, because the fee that the customer pays is an all-in fee. So everything is being taken care of, the insurance, road tax, parking, fuel and so on and so forth. This is arguably the attractiveness of car sharing, if you structure it in that way. You don't need to bother going to the pump, you don’t need to bother paying parking fees somewhere, be it online or with coins and so on and so forth. The convenience element in car sharing is excellent from a user's point of view, I would say. But, for an operator, it comes with a lot of fixed costs.

{audio:00:26:03} Let's say you do 15 rides per day, at roughly 10-15 minutes per ride, then the car is only utilized for roughly 2 and a half hours, which is 15% of the time that people are awake. Is that enough utilization to make it work?

{audio:00:26:30} As I said, that is the base assumption. Clearly, we want to have 5 hours or 8 hours of operation. But, if you consider the following, on average, private registered vehicles are used less than 1 hour per day. So, if you get to 2 and a half to 3 hours, for argument’s sake, you're already tripling the usage. It might not be as many rides as a typical taxi, but I'm not familiar with the revenue structure and the utilization of a taxi that is maintained 24 hours a day in 3 or so shifts. Possibly they do more, but that’s speculation from my side.

{audio:00:27:29} What would you expect the utilization to be in Madrid or a city where it seems to be working or growing pretty quickly?

{audio:00:27:43} I can only speculate. But, just based on the announcements I've read, the new cities that Car2go entered were quite successful. So, in Madrid there are probably even more rides than our base assumption. Hamburg, my home town, is higher. I would say Berlin is good. The Italian cities, from what I'm hearing, have a high utilization, lots of users. I cannot speak on the current status in the United States. In Canada, Vancouver, from memory, was the first city that broke even after two or three years. Vancouver runs a slightly different model. There is no real free flow, it’s a mixture of free flow and the station-based model that Zipcar operates.

{audio:00:29:05} It seems very difficult to break even with this.

{audio:00:29:14} Yes, I think so. I think that was one of the rationales behind the current merger of BMW and Daimler, when they put all their new mobility offers together under one roof. I think it was clearly driven by optimizing their cost structure.

{audio:00:29:44} Did it change the costs much though?

{audio:00:30:04} First of all, what you can eliminate is a second payment and a second booking and a second ordering system. So currently, probably BMW and Daimler both have developed their own systems independent of each other and if you can put those operations on one platform you can eliminate some running costs.

{audio:00:30:39} I have no idea, a couple of million certainly. I can’t say that it’s complicated but It's a complex system.

{audio:00:31:01} I thought the synergies would come more on the revenues side i.e. you have one brand, one platform and therefore you can drive usage.

{audio:00:31:16} Probably, at some point in time, they’ll merge the brands. Currently, they're keeping them separate. But coming back to what you can save on a city level, if you merged the city offices, the back office staff that both companies have can handle double the amount of customers easily. And even simple things like rent, you only need one back office instead of two. Maybe those are a lot of small things, but small things can add up to a lot of money.

{audio:00:32:00} Do the OEMs run the car sharing business to make it profitable or is it to collect data?

{audio:00:32:16} I can only speak based on my experience, where, yes, the aim was to make it a profitable business model, knowing that this is a big challenge. It was not just to test something and stay there forever. Normally you test something, and either it works or it doesn't work. If it doesn't work, you stop it. Collecting data in today's environment is not easy, at least personalized data. Crowd data you can gather with the data you assemble with Car2Go, i.e. average distance driven, average minutes used, the concentration of mobility demand in certain cities and so on and so forth. Parallel to this, in most cities, there are ongoing discussions with the provider of public transport to make it part of the mobility mix in an urban conurbation. But, at least in my experience, you always have the struggle about who owns the data. So there was no breakthrough in saying “Let's both collectively do it and create a mobility network that is superb and maybe we can take it even further from Madrid to Seville or whatever and expand on this.”

{audio:00:33:56} What kind of data are they collecting?

{audio:00:34:03} As I said, how much time do you spend in the car, the distance driven, et cetera. But it's not personalized data.

{audio:00:34:17} So you can't leverage that data?

{audio:00:34:20} No, not really. At least, I can't think of anything where this data could be leveraged in terms of revenue generation.

{audio:00:34:32} Are they collecting any data that will help with autonomous vehicles?

{audio:00:34:43} Well that could be an area where you say, “What is the typical area of operation when you do car sharing?” and then you go autonomous, but this is far off. Getting autonomous vehicles on the road is far off.

{audio:00:35:13} If you look at Waymo, they claim to have ‘millions of miles’ in data imaging and sensors. In their presentations, the OEMs claim that connectivity and mobility is where they’re heading, but we don't see or hear about the actual quality of the data that they have.

{audio:00:35:39} Yes, and I wonder about OEMs, based on their typical thinking - this is not a criticism - if they really are creative enough to determine how they can use data as a revenue generator. Companies like Google come from a totally different background and they use technology for their core business, whereas the OEMs come from a technology background and have always struggled to find intelligent solutions, in terms of applications other than their core competency.

{audio:00:36:38} How do you see the OEMs’ positioning in autonomous vehicles?

{audio:00:36:49} I think the OEMs’ position is underestimated in terms of capabilities. In my view, they do a little less media work on their autonomous capabilities.

{audio:00:37:13} Which makes it difficult.

{audio:00:37:17} I think they are more realistic. If you promise too much and you don’t deliver, it can damage your brand. My last discussions around autonomous vehicles was more about their technical feasibility rather than autonomous driving as a realistic new business model for anybody. There are so many obstacles at the moment that I always say, “Autonomous is great, but let's start with handleable use cases in a defined environment,” before we say, “Can this car overtake at 120 miles per hour?” It's completely irrelevant.

{audio:00:38:11} That's why I was curious about the data collected, because the perception or the behavior of the car is obviously driven by the data collected. So the perception level and the technology stack in AV require the car to understand certain situations. Waymo and Tesla certainly claim that they have proprietary data where they can analyze that. I haven't heard anything or seen anything from the OEMs that would claim that they have anywhere near that type of data.

{audio:00:38:47} I would put a question mark to it. If you add all the intelligence, the capacity, the engineering of the OEMs based in Silicon Valley, from Tier 1 suppliers to OEMs, if you put this all together, it will outnumber the efforts that Waymo and others put in there.

{audio:00:39:15} Don’t you need kilometers travelled?

{audio:00:39:23} I think they are testing it as well, as we speak. But maybe they make a little less of a show about it. I’ve been to Silicon Valley many times and you see vehicles from all kinds of manufacturers and Tier 1 suppliers running around and collecting geo-data and what have you.

{audio:00:39:51} So they’re just not as vocal. Circling back to car rentals, how do you view the advantage of this merged entity of BMW and Daimler JV in carsharing versus the likes of the car rental companies that own their own car sharing facilities? What’s the advantage of being part of a car rental company versus an OEM in car sharing?

{audio:00:40:26} Good question. In the beginning, those companies partnered with traditional car sharing companies just to make sure that they covered the logistics side of carsharing because these companies, be it SIXT, Europcar, Avis and so on, they know how to handle large car pools. That’s their bread and butter. Are they now in clear competition with them? As long as the car rental companies stay station based, I think there is room for both without them competing with each other.

My view is that car rental companies are putting more emphasis on the automation of the rental process. I mean, even today, it’s still cumbersome. You have to go to a counter, present your driver’s license and passport and a credit card. And they hand you the key, probably not the one for the car you ordered and so on. I think their focus is more on the digitalization of their core business, rather than expanding into new areas. A big element of their revenue stream is the remarketing side of their vehicles. So, I think it would be beneficial to spend more time comparing these two business models.

Mind you, SIXT has announced that they are now entering the flexible car sharing market. If it’s just because the partnership with BMW came to an end and they want to be a little bit more aggressive then I don’t know. But I think one advantage OEMs have is that they have their own inventory available. They avoid the hassle of price negotiations that typically comes with doing business with rental car companies. But, currently, I think there is no real competition between the two entities or between the two groups of companies.

{audio:00:43:08} What do you think the future of the car rental business is in this new mobility world?

{audio:00:43:17} As I said, they have to up their game in terms of accessibility and convenience. I think if they can provide better access, maybe then people will start thinking about whether they should own a vehicle or not. When car sharing fulfils your mobility demand in urban centers, but if you want to go with a group of friends for the weekend, it's still a little bit cumbersome. You have to go to a place, which typically is remote, and then you have to go through this very cumbersome rental process. I think if they up their game in terms of convenience, accessibility, by introducing a delivered to your home service or something like that. That’s where I believe the car rental business has to go, and not this typical airport-based, stand-in-line business that you currently have.

{audio:00:44:30} But doesn't that compete more directly with the OEMs then?

{audio:00:44:34} Why would they compete with OEMs? Only if they replace car purchase. That could be the case.  I read a McKinsey report from 3 or 4 years ago, where they tried to understand what the volume challenge of car sharing and the Uber business is to the OEMs. If there is a significant change in terms of consumer behavior, i.e. more ride hailing and more car sharing, they said we would see a dip of maybe 5 years. But, because these vehicles are more intensively used than a typical private vehicle, the replacements cycle would kick in, and that replacement cycle is shorter than the replacement cycle for a privately owned and used vehicle.

As long as the policy makers and the authorities do not start to govern mobility from a central point of view i.e. limiting the number of vehicles on the road or limiting the number of miles or kilometers allowed for an individual to drive for a year and so on and so forth, I don't see a very big challenge from a global perspective, because, as I said, the theory is still valid today that individuals love individual mobility. Arguably, there’s an irrational aspect to it, but at some point, you will look at individual mobility like a convenience good or even a luxury good. You don't buy a Rolex so that you always know the time.

If you only pay for the time and distance travelled, it's cheaper than buying and maintaining a car, so there is quite clearly an irrational element to mobility and an irrational element, even more so, to individual car ownership. But it's there.

{audio:00:47:36} The OEMs all report and communicate in similar ways about how they're concentrating on connectivity, autonomous, mobility and electric vehicles. How do you look at the position of each of the OEMs in mobility compared to the tech guys that we just spoke about?

{audio:00:48:13} When you look at statements from the big OEMs right now, they sound quite similar. They all claim that they are not car manufacturers any more, they are mobility providers. And that's true to some extent. Every OEM has their own finance house, providing financing and leasing, and I think that's where it started. That was the first move from OEMs out of the typical ‘build a car and sell it to a dealer’ model. So that is the very old model.

Now, to some extent, they control a huge amount of the finance and leasing market through their in-house operations. The next step is wanting to get involved in how vehicles are being used, and arguably, the commercial vehicle sector is way ahead of the passenger vehicle sector. Most OEMs already provide rental operations. I’ll use Tesco as an example in the UK. Tesco runs a huge fleet of vehicles that they own. Then, when they experience certain peaks in demand, they go out and rent vehicles, either in the open market, or from the OEMs.

The OEMs have their own rental fleets to provide mobility to their customers or anybody else who pays, and they have maintenance contracts. Charter Way is what this business is called with Mercedes. You can have a contract based on utilization, meaning you'll pay for the mile and not for the month. So, it's already there, maybe not as prominent as in the passenger car market, but OEMs will constantly look into possible threats to their core business. And, if they have identified a “threat,” they will start to investigate whether they could command a share of the pie.

{audio:00:50:51} Do you think that they have the technical expertise, in terms of software, to make this shift to being more of a mobility provider rather than a car seller?

{audio:00:51:12} Well, if they buy the expertise, as they already do today, then they do. I think if you look at a vehicle that was launched last week in Geneva, I would argue that maybe 50% of the intelligence behind this vehicle comes from the OEM directly and the other 50%, if not even more, comes from engineering providers, Tier 1 suppliers and so on and so forth. So I don't see this as a major obstacle. As long as OEMs are willing to partner, and they are now proving that they are learning from the past, then it doesn't provide an obstacle if you don’t have in-house expertise.

{audio:00:52:05} A lot of the tech guys that we speak to disregard the OEMs in this, whether it's mobility, autonomous, anything highly technological. This is based on technical expertise, on culture, the asset base that they have, and their lack of flexibility. How would you argue against those points - specifically around technical expertise, fixed cost base and the culture? Are these really disadvantages for OEMs?

{audio:00:52:59} No. I think that payback is a bitch. The OEMs didn't take the tech guys seriously 10 or 15 years ago, saying, “They will never make it” and “We are the masters of the engineering universe in all aspects.”  So that was the position that OEMs took. And I would argue that I know one or two companies that still take this position. I think, it's become a little game, a back and forth, where tech guys say, “Hey, we don't need anything that requires a lot of assets. We can do it much more intelligently, much more flexibly.”

I would argue that you can't do this alone. At the end of the day, it comes to a partnership. You need to have an asset, because without an asset you don't move, full stop. And if you don't get access to the asset then you can have the most intelligent technology, but it doesn't matter.

It's why people are saying that everyone can make, manufacturer and assemble a good car. That's right. There’s a lot of competency around the world. But you need the brains behind it. Look at Tesla, they shook up the automotive market, but they’ve had enormous trouble building up a very stable manufacturing base. And they have been doing this for 10 years. So, it's not kindergarten. That would be my argument. We don't get anywhere if we take the position of you and I. We have to understand that if we want to move something forward, we have to look at how we can work together.

{audio:00:55:14} That’s clear. My question then would be: if you are purely an asset or commodity provider to Uber to Lyft in a car sharing model, where does the value actually accrue? Is the OEM’s profit going to be lower going forward in this mobility world?

{audio:00:55:46} I think it's very simple. At the end of the day, I would look at the margin that I can achieve in selling vehicles or leasing vehicles to Uber and Lyft compared to the margins I get from private customers. So that's not really a valid threat.

{audio:00:56:06} It would be lower right?

{audio:00:56:10} Well, it depends. As in every business, it depends on how powerful the customer is. At the end of the day, if the mobility market consisted of just of ride-hailing and there were two companies in the world: Uber and Lyft, then you're right. But you tell me when we get to this situation.

{audio:00:56:39} But there's not many more though? There’s Uber, there’s Lyft, and there's a couple more in India and China, obviously Daimler have Mytaxi and there’s some in Europe. But on the whole, there is not as many as there are OEM brands. There's not many choices. I think people are questioning the value creation of being an asset provider versus a mobility provider.

{audio:00:57:13} Maybe they’ll fix it at some point. I'm a little bit familiar with the Uber model, granted it's 3 to 4-year-old knowledge and they might have changed their business model. The Uber model, at least that’s how they started, went to the extreme: “We don't own the asset, we don't own the employee, we own nothing basically.”

{audio:00:57:40} “We have the intelligence for the system and we have the customer.” Now, there was a point where the biggest challenge for Uber was to get drivers. It's a pretty rubbish job, where they make very little money. And the biggest challenge for the drivers was where do I get the car from. There were drivers who drove for Uber in a rented car. You know that this is basically breaking the law in every Western country. When you rent a car, you sign in the rental contract that you are renting it for private use. So we had discussions with Uber and they were very interested in partnerships. They probably didn't only speak with Mercedes. They probably spoke with every OEM who opened the door to say, “For heaven's sake, how do we get the cars onto our platform?” As I said, it depends how far they want to take their model.

We did some work in expanding the sharing idea, “Sharing is the new owning,” over and above just vehicles. You can do it with every asset that there is. I challenged my team and said, “How can we established ourselves in a sharing world as the master of the universe because we have the asset?” Before you share something, you have to know someone who has an asset.

{audio:00:59:24} But you don't own the customer and that's the question.

{audio:00:59:30} It comes back to the fact that it only makes sense when you have both. If you have a partner who has the asset and another partner who has the customer, the only solution is that your partner.

{audio:00:59:54} Then the question becomes is that a fair partnership? Arguably, there could be many different asset providers for Uber or Lyft, but there's only one Uber and one Lyft.

{audio:01:00:14} Yes, but that's where we were five minutes ago. As I said, let me look at the margin. If the margin is the same, I sell to Uber as happily as I sell to Mr Barnes.

{audio:01:00:34} In that sense you're not particularly worried about the margin profile of the OEM in a mobility world being different to what it has been in the past?

{audio:01:00:45} Well, the margin with OEMs is much more under threat with new entrants from China. The European industry barely survived when the Japanese came. They took another dip when the Koreans came. And now the Chinese, particularly with electric vehicle technology, are getting serious, not just in manufacturing for the home market, but they are starting to export more aggressively. At the moment and for the foreseeable future, this is where the real challenge is. I'm not so concerned with Uber for the traditional OEMs. I'm more concerned that they make the right decision in terms of portfolio structure and flexibility on the cost side, because I don't see a lot of price elasticity.

{audio:01:02:00} Are you referring to electric vehicle entrants specifically or to all entrants?