Rivian Automotive, Inc. (RIVN) Earnings Call Transcript & Summary

Hello, everyone. Welcome to the next session of Redburn's 2022 CEO Conference. My name is Charlie Coldicott, and I am the Head of Global Automotive Research at Redburn. For this session, we are really pleased to be joined by RJ Scaringe, CEO and Founder of Rivian. Thanks for joining us, RJ.

Yes. Thanks for having me, Charlie. It's good to be on.

Firstly, a quick reminder for everyone. If you want to ask a question, please don't hesitate to submit one. There's a button at the bottom of your screen where you can type your question, and I will read out anonymously for you.

But RJ, I think where I'd like to kick off is asking about demand. So obviously, we've seen order intake slow materially at some other EV names, people like Lucid, even Tesla in certain regions. And a lot of customers are having a tough time right now. Are you confident that Rivian can sell everything it produces whatever the macro backdrop?

Yes. So it's a great question. It's foundational to the business to make sure that demand continues to exceed our production capacity. What we've been challenged with, really, over the last year that we've had such a large and growing demand backlog that managing customer expectations and making sure we're communicating appropriately to customers, some of which were going to be in the backlog or a line for over 2 years. That's been a really core focus. And as we sit here today, we have a very clear line of sight through 2024 in terms of demand, so we feel very confident about that, and we continue to see -- and we talked about this on our last earnings call, we continue to see our backlog grow even as we're ramping deliveries.

Great, great. And on the order backlog. So obviously, the order backlog has got quite long now, as you say, sort of reaching into 2024. Do you have, in your mind, an ideal sort of length of order backlog that you'd rather not go too far before it sort of begins to impact demand?

Yes, it's a really good question. I mean the psychology of a backlog is some backlog is actually helpful. You're in line for something you feel like others are in line for. But as it's so long, it starts to be a deterrent for demand. Analogous to as you walk to the restaurant, you want to see a crowd, but you don't want to see a line around 2 blocks. So that's where we are today. The line is getting and the backlog is getting to a point where it is a real source of frustration for someone that says, "Hey, I'm super excited about the product. I love what I see. I got a chance to drive it. It's fun. Can I get one next year?" And we're like, "Well, not easily." so that's something we're, of course, working to solve as rapidly as we can by ramping production. And it's a very high-class problem to have is peak in terms of having this level of backlog, but we would like to get to a stage where we're not carrying as much backlog, where it's not a 2-plus year wait time to get a vehicle.

I guess one way that you can control the backlog is by adjusting pricing. And I suppose, for a lot of onlookers, the pricing of the R1S and the R1T today seems potentially on the low side. I mean, if I look at some of the competitor products like, for example, a new Range Rover, which has just come out, that starts at about $100,000. It can go up to $200,000. In comparison, the R1S, arguably, looked quite underpriced. So do you anticipate -- or how much room do you think there is for adjusting the price upwards in the coming years?

We definitely see this in the numbers. We see this in data as we continue to have orders come in. And the R1S in particular, has -- we think a lot of pricing room still is there. And we've talked about this before with additional driving that's coming in, it allows for between the dual motor and a quad motor, it allows for us to start to create some more stretch on our pricing. And when we came out of the gate with the products, we clearly had underpriced them relative to the market as we now see feedback coming in and the demand being, as you said, so strong. So it's something that we're paying very close attention to. And there's tens of thousands of dollars of difference in pricing for our product versus products that are very similar in terms of size, footprint, but the performance in the Rivian and the on-road and the off-road capabilities is really quite unique and there's nothing quite like it.

Another issue, I guess, on the pricing side has been clearly the cost environment has been very different from when you originally envisaged the product and designed the product and budgeted for it. A major element of that has been battery raw materials. Can you talk a little bit about how you see the outlook for materials like lithium and nickel? And firstly, in terms of do you think that availability is going to be a limiting factor for EV production, whether Rivian specifically or for the broader industry? And if so, when? And secondly, how that will impact overall affordability.

Yes. We talked about this before. I think this is one of the least-covered and at least-talked about challenges we're going to see, not just for Rivian, but as an industry, as we go from low single-digit penetration of electric vehicles to a not-too-distant future 100% penetration of electric vehicles. And at the core of that, we feel, is going to be scarce around certain raw materials, and we're already beginning to see that. And perhaps the one to focus on, a specific example of that, the one I'd point us to would be lithium. So if you look at the price of lithium hydroxide today, it's in the mid-$70s per kilogram. What it was about a year ago was just over $10, and the cost is less than $10. So the cost to make a kilogram of hydroxide is depending on the source, $6 to maybe $9, $9.5. And the price today is extraordinarily high. And it's really a market-driven price where just the scarcity of supply has driven a bunch of folks, ourselves included, to go source that material to make sure we can produce. So where we see this playing out as we think lithium pricing will continue to stay high. We're going to see the same in nickel. But at the end of the day, it's going to ultimately result in a limitation for how fast the whole market can electrify. And the high pricing, of course, is helpful for bringing on new supply, but that doesn't come on instantly. It certainly takes time. These are large-scale projects, multiyear, multibillion-dollar projects to create the kind of capacity we need to get to 50%, 60%, 70% EV penetration.

For Rivian, specifically, so do you contract directly for the supply battery materials? Or -- and if so, how long into the future have you secured the supply that you require?

Yes. One of the things that we're going to see and that we're focused on is the way in which we source the materials and the batteries that go into our vehicles is going to evolve, we believe, quite dramatically where, historically, the customer and OEM would buy batteries from a battery supplier, battery supplier to invest in capital to build those batteries and the OEM would be the buyer of them. What we see increasingly and certainly the focus of our business is to make sure that the capacity that's being built at the cell level, so cell manufacturing level, is dedicated for us, in which case, we lock in that capacity for the cell. And then we go upstream the cell and we say, "What are the materials for which we need to lock in capacity?" And then each of those agreements is varied, but you'll have -- as -- and by design, we have a basket of different types of agreements. But agreements that include more upfront payment where you control pricing a little bit more effectively to structures that just provide for supply. And it's, ultimately, for us, a huge focus area in terms of how do we allocate capital, allocate capital between different types of raw materials to ensure the right price caps are put in place and supply agreements to then the right type of supply agreements with battery cell manufacturers who we're co-investing in production capacity that, as I said a moment ago, is dedicated to us.

The -- and with the next steps in your battery cell manufacturing and your battery cell sourcing, obviously, as you mentioned, you want to have partnered production with suppliers and also your own battery cell manufacturing ultimately, can you talk a little bit about what you see as the opportunities and maybe also the challenges of choosing to do battery cell manufacturing in-house? And maybe specifically, do you have a long-term goal for capacity of in-house manufacturing?

Yes, the thing to keep in mind as we talk about batteries, there's not a single solution that will solve for the level of scale we plan to operate in, meaning we couldn't scale with just one battery cell supplier or just one lithium hydroxide supplier. So it will be a portfolio of different battery cell relationships that we have with different types of structures. So today, we leverage and utilize Samsung SDI for the cells that are going to the R1 products. We have other battery cell suppliers that are only coming online which we haven't announced yet, which again will fill out a broader portfolio of cell suppliers. And we bring on different chemistries. And we've talked a lot about this in the past. But today, all our vehicles are high-nickel chemistries, where we're going to be adding lithium iron phosphate first with our commercial vehicles later this year. And then into next year, we'll be pulling LFP, lithium ion phosphate, into our consumer vehicles. So that portfolio of different approaches is really important. And within the portfolio, we also have different ways of producing the cells. So purely sourced cell, cell produced by a supplier, 100% supplier-owned equipment and facilities, too. On the other end of spectrum, self-produced and Rivian-owned facility with Rivian-owned equipment. And over time, we see, by design and by intention, a mix across those, which, of course, balances risk, but also reflects the fact that the constraints that exist in different countries and different regions are very different. And that's become even more true in the last couple of months with the IRA bill. And what that really drives is domestic investment for the U.S. market with U.S.-owned cell production. And it's something that our in-house cell efforts have aligned with really well to support growth into the latter part of this decade.

Fantastic. [Operator Instructions] Just a reminder to everyone as well, keep entering your , but before I go to those, maybe a couple more, and I'm just going to change tact actually as well. So I want to ask about software as well. Obviously, your strategy is to develop your own in-house software and you have stacked it to your own in-house electronic architecture, which means that you're effectively doing everything from top to bottom. But most carmakers are taking a very different approach. They think that they can outsource most the software and just handle the sort of customer-facing bit. And you're even seeing some -- you've outsourced the whole operating system to the likes of Google, who are obviously very good at software. Why do you think your strategy is the optimal one?

Well, ultimately, we really believe that the digital interaction with the vehicle and, for that matter, the way that the vehicle interacts with the rest of the ecosystem, so through the cloud to all the services that we support the vehicles with will be greatly enhanced when we can control, not just the features and the attributes, but control how we deliver those features and attributes. And so to do that fully, on our first products which we've launched with the flagship platform, with the R1 platform, we developed the network architecture in-house, as you said. And in turn, we then developed all the compute platforms across that. So all the ECUs, the control units across the vehicle. And of course, on top of those, we built the software stack. And so from the base level OS all the way up to the application layer. And what that allows, particularly over time, is a very rapid and very holistic feature growth, meaning we can leverage our cameras for any feature in the vehicle. We can leverage our accelerometers for any feature in the vehicle. They're not sort of fenced off to the features that were ascribed to the supplier that provided the ECU. So for example, the idea of, like, a powertrain control module that's only controlling the powertrain is sort of doesn't -- it makes it very hard to say, let's say, use a specific Tier 1 for that to have that control module then working across, very seamlessly, sensors that are on a completely different Tier 1 supplier platform. So that ability to control everything as if it's an actual computer where data is moving freely, features can leverage sensors set and compute across multiple platforms. And then importantly, that we can move very quickly to add things, and we've started to demonstrate that with some of the over-the-air updates that we've done. And I'd really characterize these as being tip of the iceberg. There's huge opportunities that continue to exist for us to continue to grow the capabilities of the vehicle and continue to grow the types of features that are on a vehicle that you've already purchased.

And the types of features that you might be able to add, I mean, one thing that we've seen with Tesla is that they can even increase the range of the vehicle with an over-the-air update. Is that something Rivian can and will do?

Yes, certainly. So when you think about range, it's a lot of things. It's aerodynamics, it's the type of tires, it's the weight of the vehicle. Of course, it's the motors, but it's also the software of how you're using energy in the vehicle. So what's the power moding strategy for all the different compute platforms, what's the thermal strategy for the cabin cooling, all these things play together to drive vehicle-level efficiency. This is certainly something we're going to see in our vehicles as we add range over time. We'll see that coming into the early part of next year.

Is that just something you need to have a large sort of size as you collect more data and then you can make those updates. Is that kind of the hurdle?

It's really dependent on how we do it. So some of it may be that we increase the usable capacity of the battery because we have the data that tells us and gives us the confidence to do that, but it may also just be we found ways to better optimize energy management within the vehicle. Then even further than that, it may be ways to more appropriately leverage the torque application across the different motors in the vehicle.

Got it. And then sticking on those potential future features, I think on the software business model, on the commercial side, with FleetOS, I think it's very clear for investors about what the potential is. But maybe on the consumer side, it's a little bit harder because, I guess, we have to imagine what these features might be. Can you talk a little bit about what you see as the main monetizable services on the consumer side and how big that market opportunity is?

Yes, as you said, the commercial side is very straightforward. You can very easily connect the dots to how to do that. And so to us, it's the embodiment of that. On consumers, it's going to be harder. We have to learn new -- sort of train our muscles, so to speak, to think about purchase behavior and the psychology of subscription in a different way. And the way we see it, the big sort of movable levers that will drive a lot of value for which we see customers willing to pay incrementally for is around autonomous features and growing that feature set and capability over time and then integration with the rest of our ecosystem, and what does that mean? Specifically, with charging. And we see opportunities to really leverage our charging network that we're building as a way to create an immediate draw into our membership platform and the idea of being part of some of these subscription platforms because of the access we can provide to our charging infrastructure.

So on the autonomous side, I think one interesting thing we've seen with Tesla's FSD has been actually the take rate has not been huge. I think, globally, it's less than 10% that people choose to take FSD. Obviously, it's an expensive option. But maybe that shows that right now it's not central to customers' decision-making. Obviously, the feature set hasn't been fully materialized yet either. But at what stage do you think the carmakers are going to start to lose out on a significant number of cells if they don't have the technology for high levels of autonomy? And by that, I mean, maybe Level 3 plus or, obviously, 4/5.

Yes, we're going to see and we're seeing, I should say, it is really interesting coexistence of Level 4/5 platforms that are very hardware-heavy, that there's $100,000 worth of compute sensor in a vehicle. They're achieving Level 5 capabilities and very specific highly geofenced domains. And the applications are, as we would expect, that they tend to be more robo-taxi type applications. And we see that technology topology in contrast to hardware-constrained systems where the bill of materials constraints means that you can spend $2,000, $3,000, maybe $4,000 on sensor and compute and, ultimately, you deliver features that are going to be more limited, but on a much broader set of roads and across a far more open environment. And the benefit of the latter system, of a more hardware-constrained system is you can put that on a lot of vehicles and you can accumulate a lot of miles. And that mileage accumulation and the presence on -- in our case, every vehicle allows us to identify the challenging corner cases that can then be used to train the system. So that's training our planning algorithms on how to respond to unique situations. It's using the image streams that come in to refine our neural nets and you have the heavy annotation work. And so mileage accumulation starts to become a really valuable element of this. And -- with that said, our approach is to focus on the latter systems. So we are not developing a hardware-heavy system. It's the cameras, radar, high-precision GPS, of course, ultrasonics, but that's allowing us to deliver a set of features to start, and it's key to start. That's really, like, what we call a level 2. So you can get on the highway. You can take your hands off the wheel for some time. The vehicle can drive itself. You can have a little bit of time back, so to speak. But where we see really significant value creation for customers is going to level 3 where your hands can come off the wheel, eyes can come off the road for extended periods of time, not 30 seconds, a minute, but 30 minutes, a full commute. And as that starts to happen, the ability to charge for that, the ability for customers who are willing to pay for that becomes quite powerful. And we believe that, over the next decade, that will go from being something that is a unique feature to something that, not only are customers willing to pay for, but becomes a really core sort of table stakes feature or set of features in the overall sort of product strategy. And so for this reason, we've taken the approach of leaning into building this stack as well and leveraging the fact that we have significant mileage accumulation already starting and that will continue to go, to continue to grow. But how we think about annotation, how we think about the camera topology, how we think about the integration with different sensing modalities, in our case, in particular RADAR, to make sure that we move into, you said it, about this level 3 plus. And then it starts to get fuzzy, what's different between level 3 plus versus level 4. But effectively, we see it is allowing you to get your time back on your commute and to do it confidently and to know that you're in a very safe operating vehicle while that's happening.

Okay. Great. So I want to feed in some questions from the floor now. One that a number of people have raised is kind of on the near-term production outlook, the production ramp-up. Obviously, your guidance implies still quite a significant step-up in production in Q4 relative to what you've done year-to-date. Can you maybe talk about -- give us the latest on the main bottlenecks and how you're going about solving those?

It's been -- and we've certainly talked about this very openly, very publicly. And it's been a challenging year for us. We launched 3 different products, really 4 products, 2 different versions of our van, the R1T, R1S across 2 different platforms and 2 different production lines. Launching a single vehicle is a complex task. Launching these 2 different platforms and 4 different vehicles has been incredibly challenging. And doing that with the backdrop of some of the supply chain challenges that we've all witnessed across many businesses, but certainly in automotive to a very sort of high degree, that's created challenges that were unforeseen. So we saw, really, in the first half of the year, as we spooled up or started to rotate the flywheel of our supply chain to get, in each vehicle's case, roughly 400 suppliers across 2,000 different parts, get each of those supply chains ramped, to get them ramped consistently so that were limited by the slowest part, so to speak. So you can't build a vehicle with 99.9% of the parts. You need all of them. And as we came through the middle of the year, a lot of those challenges we're seeing, where is almost a daily basis of a line being stopped or down started to stabilize. And then in Q3, we talked about this. We really got it down to a very tight set of challenging areas from a supply chain point of view and allowed us to the very sort of almost start to really see the plant run as it was intended. We're still certainly not there. It's a plant that's designed to produce a lot more capacity than what it's done so far this year. And that's a painful way to operate a large manufacturing facility. It's not capital-efficient. It's not utilizing the resources in the facility the way they're intended. So the big step for us was saying, let's go to -- let's add a second shift. And to do that, we needed to ensure we had supply chain, meaning it does us no good to bring on a whole staff of people to run a second shift and then tell the team to stand around and wait for parts. So we had to make sure the supply chain was stable enough and we ultimately had the second shift come online end of Q3 and is now in that ramp period of going through -- getting up to 100% or is close to 100% OE year, operating efficiency of a line, which we'll see play out through Q1 of next year. Now as that's happening, it's not to say we're completely out of the woods on supply chain. There's still real challenges that exist. Fortunately, they're well understood and less surprising than they were in the first half of this year. But we've -- as I said in our earnings call, a couple of weeks ago, we lost 5 days of production already this quarter to some unforeseen -- or I shouldn't say unforeseen but unfortunate challenges around [ jets ]. And now we really do believe that, that's going to start to subside. And part of the challenge for us is as we ramp. We're producing more vehicles every day and the supply chain needs to be more robust, and that's getting added. But we're very focused on this. We're focused on making sure the supply chain is stable. We continue to improve how we operate the facility, our production facilities. And then, of course, and I don't want to talk about production without mentioning our focus on cost. So driving cost out of the bill of materials and driving cost out of the operation of our plant.

So on that cost point, actually, so another question that we've had is that -- Sandy Munro, as most people will know, he has done a teardown of the R1T and he's talked about there being buckets of cost-reduction opportunity. Is that something you're working on already? And if we fast forward a few years, might you have been able to pull out hundreds of dollars of variable costs.

Yes, certainly. And as we look at the R1 platform, even before we launched production, we are already working on revisions on certain aspects of the vehicle. Some of those are to drive enhanced performance. Some of those are to drive enhanced efficiency, but also a lot of them are to drive cost improvements. So revising the designs, working with suppliers, working with our teams to take cost out of the vehicle. And some of the design decisions that we took early on had been a result of what was available from a sourcing point of view or even trade-offs that we made from a technical point of view to minimize risk for launch. But as we now are ramping, there's -- as you said, there's certainly exciting opportunities for us to continue to improve the cost position of the vehicle.

And going back to the ramp into next year, how are you seeing supply bottle chain -- sorry, bottlenecks easing next year? I mean do you think that the supply chain is going to be the limiting factor well into next year? What -- given that you -- as you said, you're down to now a handful of specific items, can you -- can those be fixed early next year? How should we think about production next year?

Well, I do think the -- there are 2 things playing out here. I think one, the supply chain is becoming more stable. That's not -- that's -- I'd say, at an industry level, but it's also specific to us. And I want to make sure this is fully appreciated. But if you think about our vehicle, there's, again, 400 -- roughly 400 suppliers per vehicle, 2,000 parts per vehicle. And each one of those parts from each of those suppliers has to go through a ramp curve. So whether it's a plastic panel on a door or a seat foam component or a structural element in the chassis, like, these are elements that each has a ramp curve and each supplier has to go through their own associated challenges with ramping and for the first time with us as a company. So this, we've seen those ramps occur. We've seen the -- our confidence grow with those suppliers and vice versa. And what we now -- as we look out into the next year, we continue to have a good line of sight to how those suppliers are operating a facility. So our supplier quality team spends a lot of time on site with any at-risk suppliers or suppliers we think that may not be ramping as fast. In some extreme cases, we may offer to help and be on site to help them. So we have, I'd say, a very clear line of sight there with a much more robust and a ramping flywheel-like supply chain. So the other point, the second point I want to make in addition to just the overall stability that's getting into our relationship with our suppliers and the supply chain in general is that we are seeing the positive benefits from a supply chain point of view of some of the demand softening. And it's a bit of a strange situation in our case where because we have such large demand backlog, the supply chain being less over-pressurized is actually helpful, particularly with semiconductors. And we've seen that space and across the various stages of the supply chain start to loosen up in powerful ways that we think are going to leave a lot of the challenges we had throughout this year.

Okay. Another question that's come up. Looking a little bit further ahead is, how is the R2 line design progressing? And maybe I'd add to that, for -- what are the major milestones that investors should be watching out for between now and when R2 starts production in 2026?

Well, I was in the studio yesterday and -- with our engineering team. We're incredibly excited about the future products in the R2 platform. One of the things that's perhaps most exciting is it takes all the goodness of what we've tried to build into the Rivian brand DNA in terms of capability, on-road dynamics, refinement, functionality, and we shrink the package, we shrink it to a smaller form factor, reduce complexity and cost of the architecture, but still embed it with that very much a Rivian feel in terms of the dynamics and how the vehicle -- and what the vehicle's capable of. And the products that can come out of that in terms of enabling and inspiring people to go do the kinds of things they want to take photographs of, carrying their kids, their gear, their pets, their friends, that brand promise existing in a smaller form factor, I think, is just -- I mean it's goosebump inducing. We're really excited about those products. Now in addition to that, and you mentioned the R2 line, there are so many learnings that we've generated as a company going through the launch of R1 platform and our commercial vehicle platform, in terms of the line design and, importantly, in terms of how the vehicle is designed to facilitate simplification of production and simplification of supply chain. So there's an enormous focus on manufacturing efficiency and an enormous focus on costs as we look at this next platform, which is appropriate given that it's a much higher volume platform than what was planned for with our flagship product, with the R1 platform. So this vehicle needs to enable a step change in overall volume. Its price point will enable, of course, a step change in addressable market. And as you know, this is something that we plan to have across multiple markets around the world. So that, like, hyper focus on efficiency of the architecture has been top of mind. And it's really a great exercise of taking all the engineering teams that have been working on ramping into the challenges associated with our first set of products to now see all the things we've learned, let's make sure those are really captured and embodied in the R2 platform and its -- in the vehicle that is based off that.

So a couple of questions that feed nicely into that, I think. So someone asking any -- have you had any thoughts on how Rivian might respond to Tesla's structural chassis components, i.e., the rear-end castings, et cetera? Is that something that you're thinking about with the R2?

Well, maybe not just slightly. If you think about how do we drive cost savings holistically into a vehicle. At the end of the day, if you can consolidate parts together, so make 2 parts 1 or make 1 part 0, that's really the most effective way of driving costs out now. Of course, if you were to make the vehicle body, let's say, out of 1 part, that'd be a very complex tool. It'll have very challenging repair practices. So the asymptotic end state of, like, it's one giant part, whether it's cast or stamped, it's not really realistic from a pragmatic point of view, but it's the right idea to keep in mind is looking for the opportunity for part elimination or part consolidation. And some of the cost on efforts that you're -- that we spoke about a few minutes ago on R1 are doing that, where we take an assemblage of parts and make those into 1 part. And R2 benefits from all those learnings in that we will be very heavily looking at part count reduction and part consolidation in how we design the architecture.

On a similar topic, people asking about batteries as well. We touched on raw material prices earlier. But -- and we also talked about lithium ion phosphate, but people are asking about long-term battery technologies, which ones that you see as the most promising, for example, sodium ion or also the state? Interested to hear that.

Well, in the short term, let me say, short to medium term. I do think the chemistries that we're going to continue to see really be utilized at scale will remain really high-nickel, lithium ion and LFP. And they have chemistries that work very well for different use cases. So in the case of high-nickel, it's great for fast charging. It's great for higher power applications, the volumetric energy densities. The amount of energy you can fit into a fixed volumetric shape is going to be high. Whereas LFPs, great cycle life performance, so 3, 4x better cycle life performance, lower volumetric energy density, but lower cost as well and certainly reduce or minimize exposure to materials outside of lithium. So there's -- you're not dealing with nickel, you're not dealing with cobalt. So our strategy on all the product lines is to have cells that leverage both of those chemistries. And the high-nickel chemistries really speak to the more performance-oriented versions of our vehicle. LFP and consumer will really represent the base configuration. And then in commercial, to lean very heavily into LFP, given that the size of the vehicles doesn't require the same level of volumetric energy density, and it certainly doesn't require the same level of peak power or even fast charging, so it works really well for these larger fleets that have very defined use cycles and for which cost is of paramount importance.

Great. So changing tact slightly. A question here on liquidity. Now obviously, you have publicly stated that you think your current cash requirements are sufficiently covered by what you've got today to until 2025. Can you -- but this question is can you explain how you plan to manage your capital until 2025 and also how you see Rivian's capital intensity changing after 2025?

Yes. This is an important question as we think about it coming out of the end of this year. And for us, 2022 will be one of the particularly challenging years launching not just our products but launching 4 different products, launching our first plant, not having the benefit of other plants that are -- have ramped and are now into the profitability stage of a product life cycle. So to have 1 launch and 1 plant is -- consumes a lot of capital. It certainly starts in a negative margin state. But doing that across 4 vehicles, we see the stacking effect of all that at once. So as we go into next year, of course, huge focus as we've talked about a few times already today on removing costs, driving volume up to get the leverage across all the fixed assets and the facility and all the CapEx we put into the facility, which will drive us towards strong positive unit economics and strong positive gross margins. And while we're doing that at the vehicle level, at the "cost of goods sold" level, we're also really looking at the OpEx of the business and some of that manifested in headcount reduction we made this past summer, and we talked about this in the last earnings call. We've also made other changes to really drive operational efficiency to how we're thinking about the business. And that's embodied in a number of different things. But to give an example, we've simplified some of the products that we've taken certain packages and features out even the R1 product portfolio as a way to streamline and simplify production. That has the benefit of removing engineering resources from those trends. It also has the benefit of removing some of the complexity in the upstream supply of those combinations. But beyond just a trim package or vehicle variant simplification, we've also removed some of the additional vehicle models that are planned on platforms, really informed by the strength of demand that we have around the product portfolio we've set up. So even as we thought now about R2, the number of models that we believe are necessary to capture the scale of demand that we're talking about is lower than what we once thought. And we've used that to simplify and continue to simplify how we run and build the engineering and product development organization.

Great. So we're approaching the final 10 minutes guys. So any final questions, please do submit them now, and I'll try and squeeze them in. Another one here, so this question points out that U.S. West Coast and East Coast drivers have shown clearly a willingness to migrate to fully electric vehicles. The question is, "How do you read the willingness of truck drivers in the middle of the country to make the change to EV? And I suppose, on that topic, maybe you could discuss a little bit about the demographics and the types of customers who are buying an R1 today."

Yes, when we set out to define and develop the R1 vehicles and the R1 platform, we were -- our aspirations in terms of target state were to end up with a product that really served as the brand opener, the products that would sort of open up the brand umbrella, define who we are and really connect with, as I said, this idea of enabling and inspiring folks to do have the kinds of experiences you want to take pictures of. And so as we did that, we also knew the type of customer we wanted to have. By definition, want it to be diverse, meaning we wanted customers from a wide range of different vehicle types and vehicle topologies, meaning it's not just Rivian customers all previously had these 1 or 2 vehicles. And that was sort of goal #1, if you think about in terms of how we set the targets. And then the second major objective was to be creating new electric vehicle customers, meaning a Rivian customer is not just someone that's switching off from another EV, but actually coming out of something that's combustion. And we now have the benefit of being able to say, well, how do we do relative to that target state? So on the first, there's no single vehicle that -- or even single brand that represents any significant presence in terms of what we're pulling customers from. So we're talking, like, the highest percent presence of our [indiscernible] from it is just a few percent. So a really nice mix from across SUV to hatchback and wagon to even minivans, especially for the R1S to coupes and sedans. So it's a really nice broad spectrum of products and form factors that people are coming out of. But the second, and I think this is perhaps even more powerful is the number of Rivian customers that have never before owned an electric vehicle is very high, on the order of 90%, meaning we're taking a lot of people who haven't previously had an electric vehicle, either because it wasn't something that fit their demand or it wasn't something that got them excited. And we're exposing them to this world for the first time. And it's really great for us as a brand to be that introduction point. And we're seeing that across all markets. That's Southern California to up to the Pacific Northwest. That's New England. And as you asked some questions, it's also in the center of the country. And we do believe we're at this really exciting inflection point where, "Is my next vehicle going to be electric," is becoming a more and more common question. And it's being informed by the excitement around features and capabilities that previously weren't possible with combustion powertrain. I mean our R1S, we talked about before, this is as fast as a supercar in terms of acceleration, is one of the most capable off-road vehicles in the world. And you can load it up with groceries and car seats in the back. And it's equally happy in all of those environments. And it's incredibly efficient. So you have all those things in 1 vehicle at a very attractive price. It's just pulling people into the segment that maybe weren't even considering it before. So we hope to do that in a continued way as we look at R2 and beyond, where we're creating new types of customers and creating types of experiences. But ultimately, that combination of attributes is what the brand really needs to be, really sharp.

So this next question is along the lines of something that we've been thinking about at Redburn quite a lot recently, which is, Elon, when he talks about Tesla's long-term differentiating factors, he actually talks about manufacturing being that long-term advantage. So this question is asking about how automated Rivian's plants are today. I don't know if there's a number that you can put to that at all, and what it can become. I guess the topic really is how efficient can Rivian's factories be? Because clearly, that is something that is differentiating for Tesla.

Yes. So ultimately, the degree to which we build manufacturing facilities that can build vehicles efficiently and scale rapidly is a measure of how rapidly we can scale the company. And I talked about this in the context of the R2 platform. This platform, more than anything else, is being defined by how it's manufactured. So everything from what structural role does the pack play to what's the role of the rear of the vehicle, the front of the vehicle in terms of structural contribution to things like torsional rigidity, to then how do all those parts nest together in sequence. So this is a huge focus. And it's not just how is the plant running, but it's actually how do we design the product to facilitate and enable our plants going forward to really run incredibly smoothly. And we talked about some of the vertical integration that we've taken on as a business. We talked about it in the context of network architecture and ECUs. One of the challenges in designing a vehicle is that you have a lot of different compute platforms around the vehicle and those have to be connected with wires. And one of the real exciting parts as we look at R2 and beyond is because we control all the compute platforms within the vehicle, we can really optimize for the wire harness design, which is a big enabler for how the vehicle goes together. And that same mentality applies to the structure. We can optimize the structure to coexist really nicely with these ECU and coexist really nicely with the way the harness is designed. And then, of course, the battery in the drive unit, which are big monuments, if you will, physical monuments inside the vehicle, with those being designed in-house and developed and produced in-house, those are things that we can really optimize as well. So you get this really optimal package that can drive manufacturing efficiencies, particularly as you get to scale.

Because I think there are a lot of numbers get thrown around, right? So Tesla, 10 hours to produce a vehicle or a cycle time for a vehicle versus VW 30 hours. And it's a little bit difficult from the outside to know what are the important numbers to focus on. So what are the KPIs that you look at for normal? And who are you benchmarking against? Are you benchmarking against Tesla or Ford? Yes.

Yes. The hours per unit is always a really challenging metric because it depends on what's being captured. So you can make it really low if you weren't building motors or if you weren't building…

Exactly, yes.

I believe it's in-house. So it's a hard one to do apples-to-apples comparisons with. What we try to do if we're ever doing comparisons is we look at products that are similar with similar levels of vertical integration and we break it down between different parts of the plant. So general assembly, body shop, paint shop, what does the propulsion platform look like. And the R1 platform, given the amount of content that's in it and the implicit complexity that comes with that. So electro-hydraulic roll control, air suspension, it's a quad-motor setup, so 4 motors in the vehicle. There's -- because of its presence as this really flag-bearing -- flagship product, it has a level of complexity that's implicit with the architecture. As we think about R2, there's opportunities to take some of that complexity out, still achieve the types of attributes and features we talked about that are aligned with and consistent with the brand, but without so much content. And as we look at that, certainly, hours per unit is something we're tracking and so the other thing we spend a lot of time talking about is what's our CapEx efficiency. So how much capacity can we build per dollar. And for us as, an organization, as we think about growing to many millions of units a year, that's a really critical metric that gets a tremendous amount of attention internally as we design the vehicles and the plans to build those vehicles.

Would you be willing to share what are sort of targets for CapEx per unit of capacity is, cumulative CapEx per unit capacity?

We haven't put out what our internal targets are for those types of things.

Fair enough. So we're right at the end. I've got one final question, actually, just kind of a big-picture one. In Europe, we just had the Porsche IPO. It sells 300,000 cars at an average price point of $100,000 and it trades on 20x earnings. Meanwhile, you've got Ford selling 5 million cars, but it's got a market cap that's 30% smaller than Porsche. Given a, the quality of your cars; b, the price points; c, the demographics of the typical customer, why don't you strive to be the Porsche of EVs rather than selling millions of vehicles like Ford?

Well, I would say we don't think they're mutually exclusive. We think a really strong margin profile can exist on millions of units a year, but it requires a lot of the things we just talked about. It requires a brand that deeply connects with customers. It requires a level of integration across platforms in the vehicle, and to do that, a level of control of some of those platforms, so electronics, software, propulsion platform. And it requires an emphasis and focus on manufacturability, particularly as we move into platforms. They're going to be doing hundreds of -- many hundreds of thousands or millions of units per year. So our objective is to grow and scale into multiple millions of units per year, but to do it with very strong unit economics and very strong gross margin profiles.

Fantastic. Well, look, thank you, RJ, for joining us. It's been a really interesting discussion. And thank you to everyone who's been able to dial in and great questions that were submitted. And we look forward to speaking again soon.

All right. Thanks, Charlie.

Thanks, RJ. Bye-bye.