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

Good morning, everyone, and welcome. I'm Toni Sacconaghi, Bernstein's IT, hardware and electric vehicles analyst. And I'm super happy and honored to have RJ Scaringe from Rivian join us today. Just a housekeeping item, I will be leading the fireside chat. There's an opportunity to ask questions. For those of you who aren't familiar, if you look at your program agenda, there's a barcode here. You can scan that, and you can add your questions via the pigeonhole technology, and they should show up on my iPad over here. So without further ado, why don't we jump in?

So RJ, we were here almost a year ago today. A lot's changed. Maybe you can offer your reflections on what's happened in the auto and the EV and broader auto industry over the last year and your reflections on Rivian as well.

Sure. Yes. We're just saying it is wild. It's been almost precisely 1 year since we were last talking up here. I mean a lot has changed. We've seen this time last year, the supply chain, not just for us but for across the whole industry was really challenging, a lot of surprises, a lack of predictability, and that certainly was impacting us and really continue to impact us through the second half of last year. But as we look at this year, the supply chain is much more stable in that for us, not just in terms of the broader supply chain but our specific suppliers. And that's allowed us to have a lot more predictability, how we're running our plans, how we're running our operations. So much, much more efficient operations for us this time this year than we were last year. I think the other big shift, of course, is what we've seen happening at the macro. So we're seeing a [indiscernible] continue to decline, [indiscernible] pricing sensitivities. We've seen typically the lower-priced segments pricing return back to normal as I would put it. It looks like my mic is acting up. So I was saying, at the sort of $40,000 to $50,000 price point, we saw prices peak last summer, where strangely vehicles are a lot smaller and lower content than, let's say, our R1S, R1T. For a moment, we're similarly priced. Those who have come back down on earth, and that was what we expected to happen. So it's not necessarily surprising, but I think it's noteworthy. And then for us, we're -- we've made a huge number of changes within our business. We've simplified our operations with a really heavy emphasis on driving down our OpEx. We've driven efficiency into how we're building our vehicles, and we're seeing quarter-over-quarter cost improvements on our path to profitability. We're going to see some major steps over the next few quarters with both technical changes to the vehicle but also some of the big commercial savings. And then, of course, working through the ramp on -- with 3 different vehicles, our truck, our SUV and our commercial vans. The peak pain, if you will, was about this time last year where all 3 of those launches were happening in parallel. And we've come out the other side. We're now, as I said, between predictability, the supply chain and predictability of our operations, going to see a clear line of sight to our profitability.

Great. Maybe we can zoom out a little bit and just talk about the broader auto market. So if you think about like 10 years from now or 20 years from now, is the broader auto market the same size, bigger or smaller in terms of number of units?

I think it's likely to be larger in terms of number of units. I think the other thing to keep in mind -- I mean there's -- just take it probably there's about 1.5 billion vehicles on the planet today. The vast majority of those, high 90% of those are combustion-powered vehicles. We, as an industry, have to convert that fleet over the next, call it, 15 to 20 years within our lifetimes, which is a really wild thought. And today, globally, electric vehicle penetration in terms of production is still relatively low. It's still depending on the country, sub-10% to sub-20%. A few places like Norway is very high. But for the vast majority of the world, it still represents a minority of production. So we have a lot of vehicles produced. We have to populate a lot of different segments, create a broad portfolio of options for customers to drive that shift. But our view is the market, in terms of the number of miles driven, will likely increase, and number of vehicles need to deliver those models will likely increase. But the thing to keep in mind is there's new revenue streams that are emerging around services. And a lot of these tied to creating more direct relationships with customers. So I think it's often overhyped the potential for software in a vehicle to generate revenue. So you can see some -- like it's easy and especially to make really big numbers very quickly. But there are real services that are going to start to merge. And the role of insurance, which links to self-driving, both of those offer software-based or usage-based revenue opportunities that are quite significant over time.

So RJ, just around the market size. So do you believe like EVs might actually cause an accelerated replacement cycle because the value proposition is more compelling, right? So if the auto industry today is, I don't know, between 80 million and 90 million units, closer to 80 million recently, historically, like 90 million, right? You could argue longer term that battery life will, particularly with LFP or even solid state longer term, will allow car to last longer, right? And that -- and if you have self-driving and autonomy, actually, your replacement cycle could be elongated and there could be fewer units. It sounds like maybe in the near term, though, you actually think the opposite might happen where the value proposition is pretty compelling for electric. And so rather than people, cars stay on the road for 15 years, they stay on for 12, we have an accelerated replacement cycle, and we actually overshoot that 80 million or 90 million units. Is that -- ain't they're right? I know you're not a mass market player, but I'm just trying to get a sense for how you think about market because I also want to talk about competition more broadly.

Yes. I think there may be some of that. It's going to be really -- the next 10 years will be particularly interesting. If you think about it, put yourself in the headspace of a buyer, as we approach the end of this decade, imagine buying $80,000, $90,000 ICE vehicle or historically buy it and think about having useful life of, as you said, let's say, 12-plus years, you're not buying a vehicle that's going into a world where it becomes less and less useful and becomes more and more of a strange anomaly, this thing that has limited resale, this thing that feels like it's from an industry of the past. So I do think we're going to see the shift to be somewhat nonlinear. The question is, if demand shifts in a nonlinear fashion. So let's say we go from sub-10% to 60%, 70% of new vehicle sales by 2030 being electric, can the supply chain keep up? Can we produce enough of those vehicles? Can we produce enough batteries for those vehicles? And we haven't seen supply chain constraints in that regard yet, but I do think that will end up becoming a limiting factor for how fast we electrify. I think that the broader thesis of autonomous vehicles could potentially have higher rates utilization and, therefore, you need less vehicles to deliver a certain number of miles, I think that's a fair broad view, but there's so many societal shifts that have to happen in terms of how we use our vehicles for that to have a meaningful impact on overall demand that it's hard to predict when that hits. We've spent a lot of time, and I've talked about this in the past. We've taken more bullish positions on how fast that happens in the past. But increasingly, we're seeing customers really like the idea of still having something that's theirs. They may go from 2 cars to 1. But it's -- I think there's a lot of unknowns on how we respond as society to different types of vehicles.

And you mentioned the ability to manufacture potentially being a choke point of adoption is very rapid. Do you have a base case in your mind around what you think EV penetration could be, let's say, in 2030? So if it is, let's say, 50% of new cars sold in 2030, a, do you think the industry can support that? And when you talk about constraints, is it battery manufacturing? Is it lithium refining? Is it a specific raw material? What do you ultimately think are the -- would be the gating factors for the industry if we did have a more rapid adoption, like the 60% or 70% scenario that you'll...

Yes. I mean we've already had some early samples of what that's going to look like. The battery cell manufacturing, in the very short term, we may see some constraints, but really in the longer term, that's a relatively well-understood [ processability ] of battery plants. And what we've seen with IRA has added a new layer of complexity where the need to produce cells locally has created a short-term push to create battery production -- battery cell production in the United States. But the longer-term challenge we see is the supply of lithium hydroxide for high nickel cells and lithium carbonate for lithium iron phosphate. And we saw that very clearly over the course of last year. Prices peaked around [ $80 ] a kilogram for lithium hydroxide. It's come down a lot very recently, and a lot of that peak was just driven by speculative purchasing and agreements that are going into place, and really, it still reflects at $40 or $45 a kilogram. It still reflects a significant premium over the cost of making lithium hydroxide. Depending on your source, it's $8 to $12 a kilogram, and spot market price is north of $40. So we think long, long term, lithium pricing will come down, but the mismatch between supply and demand is creating this elevation in pricing. And we think that will become a real choke point as we look at the end of this decade.

And just on that because we had a lot of questions about that. Are you typically on longer-term supply agreements? And so if we do see spot prices go from $80 to $40, what's, in reality, happening with Rivian in terms of your procurement given that I suspect the majority of your procurement is on a longer-term contract and a small minority of spot?

Yes, this is a great question. I think even a deeper version of the question is, do vehicle manufacturers need to take positions in upstream supply of lithium? And so do we buy processing facilities? Do we buy into assets, into the raw material assets themselves? And if you think about the last year, we spent a lot of time looking at different deal structures, and you can go from as far as buying a live end or buying a lithium processing company to buying specific lines dedicated to your capacity to long-term purchase agreements with price floors and price ceilings to protect both sides. But what made it really challenging last year was the spot market price was so high that frankly, there were a lot of really bad deals done. And a lot of those deals that were done over the course of the last 12 months now look really bad. They were done in the context of $80 a kilogram, not in the context of $40 a kilogram. So it was almost impossible to do a good deal over the last year because there's so much irrational behavior. I think that as we look out at the years to come, there's going to be more rational thinking around what's the deal structures. But I do think there's not a -- you can't play in the electric vehicle space at any level of volume and expect to be cost-competitive without highly intentional structures around purchasing lithium. And the lightest embodiment of that is going to be price floors and price ceilings that are agreed upon by both sides, the heaviest would be owning the asset.

Right. And the lightest now is -- so if you can indulge me on the examples of lithium has gone from $80 to $40. Help me understand what a longer-term supply agreement would mean when a price gets cut in half or doubles. Like do you -- is your beta half of that? Or generally speaking, like how conceptually do we think about that?

So as a manufacturer, we want to have a really low floor and a really low ceiling or -- and the supplier wants to inverse of that. So I think this is where last year, it was hard to negotiate. Those were price -- negotiated price floors you'd want to see or price ceilings you'd want to see were just high. And we've -- we're in a different world now. And I think what's happening is we are getting back to a discussion, what's the cost to make it based upon the asset and then agreeing upon reasonable long-term margins where the relationship has a long-term commitment. And so it's mirroring the long-term commitment, which gives the supplier continuity of demand and continuity of pricing with, in our case, the same. We want continuity of supply and continuity of pricing but recognizing we may overpay somewhat, we may underpaying somewhat given what's happening in the spot market.

Right. But is typically a contract like 1 or 2 years long? And are there abilities to adjust that? Or...

It really depends. It depends if it's an existing asset versus a new asset that's going into place. We've seen the likes of [ Almarp ]. They're building new assets. Sort of pricing and the structure on some of those is going to be very different than something that's fully depreciated and sort of been running for a while.

Maybe we could just go back to kind of addressable markets. So we have this market. And one of the things that has maybe concerned me has just been -- you have a traditional industry, the auto industry, which has been a tough industry from an economic perspective, that has gotten a lot more competitive, right? You have, I think, 7 new competitors with market caps at $7 billion or above, so viable global competitors. The industry has become more global. Chinese have evolved to be strong competitors, not only domestically but are now set their sight in Europe. It feels like you have a tough industry that competitively is getting a lot tougher. A, do you agree with that assessment? And do you expect a consolidation?

I don't know if it's getting tougher as much as it's -- the metrics of what it takes -- the criteria to be successful in the long term is shifting. So long-term success criteria requires an advantaged cost structure. So that means appropriate levels of scale, robust supply chain, efficient production operations, vehicle architecture design freeze, assembly, ease of build. But it also requires a product that customers want. And to build a product that customers want, it's very hard to do that in a historical way of outsourcing a lot of the -- not a lot, essentially all of the electronics and, therefore, all of the software stacks because the experience you have as a consumer tends to be fairly choppy, manifests and features that are relatively stagnant over the life of the vehicle, meaning there's very few [ over-the-air ] updates. When they do happen, they're very hard to do because the coordination cost between multiple Tier 1 suppliers is very high. So our view is to be truly competitive in the long term from a customer-facing point of view. Controlling and owning the software stack is really important. To do the software stack right, one has to control the ECU topology. To control this cost structure, it's even more important to control the ECU topology, the computer -- the topology of computers in the car because you can then consolidate the computers massively. So if you look at every car on the road today with really the exception of Tesla, the compute architectures are domain-based, meaning they're function-based. So you buy a brake control module from Bosch, you buy a body control module from Continental, or you buy a head unit control module from, I don't know, Valeo. And you mix all these different control modules together, and they communicate over CAN network and sort of works in this plug-and-play fashion. But there's no holistic integration at all, and you have way more ECUs that you'd really want. So what we're doing is we're introducing our -- next year, we have this massive simplification of our network architecture where you consolidate the number of computers in the car down to very, very few. So there's a computer in the front, there is a computer in the middle, and there is a computer in the back and does zonal-based functions. And that's really hard to do because you're not buying a computer that speaks to a function. You're building a computer that does a bunch of things in the front of the car, a bunch of things in the middle of the car. So that represents massive cost advantages, both in terms of the bill of materials but also in terms of how the car goes together, simplifies the way the harness goes into the vehicle, therefore, simplifies how the body structure is architected, therefore, simplifies how the rest of the content layers on top. Carpets, interior panel, seats, everything just gets easier to build. So the reason I say all this is I think to approach the auto industry from a competitive point of view and to not own some of these big enablers either from the customer-facing side of features and refinement of features holistically as well as the ability to really simplify the way the vehicle is built, it's hard to do that in the historical context of what we've known as a tiered supply chain that builds a lot of this content.

So maybe I could just dig a little more into detail on that. Because when I hear you on the differentiation side, I maybe don't appreciate it as much on the cost side, right? So let's say we have a $50,000 car, and the building materials is $40,000. That includes labor, depreciation, stuff that you buy, et cetera, et cetera. Like are the electronics historically a meaningful percentage of that $40,000 in such a way that you become a viable player because you have rearchitect that from a cost perspective? I get it from a differentiation perspective and a customer experience. But it's not clear to me, and so maybe you can shed some light or provide some context on numbers for that. But I just -- when I think about the bill of materials on a car, and I will talk about this hopefully a little bit later when we talk about you're improving your profitability per unit. I just don't think of electronics as really being a massive cost differentiator, and maybe I'm wrong on that.

Yes. I mean, well, it's helpful to go through it. So let's say if we look at the vehicles, total bill of materials, when you compare an EC -- an EV to an ICE vehicle, from a drive unit perspective, the EV is simpler architecture. So a motor's [ a lot few ] moving parts in an engine. So over time, because we get scale, motors will be cheaper and -- in most cases today, they are cheaper, if you will, apples to apples on which you find an engine. The battery is much more expensive than a fuel tank, and it always will be. There's not a world where a plastic bin basically is going to be the same price as a 80- or 90-kilowatt-hour battery pack even if you were able to get sell cost on to sub-$50 a kilowatt hour. The battery is inherently going to be more expensive. So that -- and those are the 2 big areas you'd look at with an EV that are going to be different. So you have to offset the increase in the battery cost, and you can do that through the fact that a battery electric vehicle allows for a simpler architecture. You can simplify the way the body structure is set up. You can use fewer parts. You can consolidate parts. You can do that with consolidation -- consolidated stampings, consolidated castings, so large high-pressure die castings, which we're seeing a lot of companies use. Tesla's somewhat famously using them. A lot of companies are, we are. But then really, the next big opportunity is to say the harness. Now the harness is not a $100, multi-hundred dollar piece of -- a component of the bond that's in the bill of materials. This is several thousand dollars. So if you can take the harness complexity down by a factor of 2 or 3, you are talking about savings in the $1,000-plus range. And then if you take all the computers and consolidate those and take both the margin stacking that occurs from buying them through suppliers out plus the duplicate nature of so many connectors and computers themselves, you're talking about a couple of thousand dollars. Now let's say that's a $1,000 on the harness side, plus maybe $2,000 on the ECU side, $3,000 isn't huge, but it's also not small.

That is the full margin on the car.

Yes. That's still...

$3,000 on $50,000, 6%.

So we think it's really important, and it's one -- for us, one of the big enablers for our overall cost reduction curve as we go from where we are today with our R1 platform to what we launched less than a year from now with some big updates to our platform, where we really squeeze every dollar out of everything under the skin. I think things like body architecture, I talked about, it's worth noting that as you simplify the architecture, the way the vehicle goes together, you just end up with simpler sequencing. But as we look like R1 has the -- it's our launch product. It's our flagship product, has a lot of content in it that represents a flagship product. As we look at R2, our next platform, the amount of focus that's going into ease of assembly and part consolidation or part elimination even better is enormous. And every single enabler we have on that, whether it's harness, whether it's ECU, whether it's assembly sequence, whether it's the opportunities for large-scale part consolidation is being leveraged in a way that we didn't, as an industry, do in ICE vehicles because of the disaggregated nature of the supply chain and all the components that are going to it.

Right. Well, maybe that leads us into the discussion on -- and you've started to do it, sort of the bridge between -- so what -- today, let's say, your cost per car might be $150,000. I mean I know a portion of that is LCNRV. But maybe we can -- again, if I think about it like really simplistically and maybe that makes it easier for the audience, I think about, okay, well, when I'm building a car, best practice car in the world is maybe 40 hours of labor, maybe it's 40 to 80 hours of labor per car. We kind of know how much that is. Depreciation is like $2,000 or $3,000. Battery, if it's 135-kilowatt pack, it's probably $20,000 plus, and then you buy a bunch of stuff and put in it, right? And again, you have this LCNRV, which is also contributing to your COGS. But you're trying to go from roughly $150,000 per car in COGS to $80,000, $90,000 over the next few quarters. So within that context of sort of the building blocks of a car, why are you so different today? And how does volume either improve labor per car or improve depreciation per car or improve COGS per car within sort of that really simplistic framework of building blocks of a car?

We spend a lot of time on this. This is...

I can imagine, this is like [indiscernible] for you.

So the biggest driver towards healthy margin structure on R1 is volume. So if you think about the path to our target state margin profile, about half is volume. And the reason for that is the fixed cost absorption in a large-scale plant when you're running a plant well under -- at subscale is really high deal. So that's by far away...

Now when you talk about fixed cost, is that depreciation? Is that overhead labor associated with that plant? Because again, at [ this day ], depreciation is $2,000 or $3,000 per car. So if your plant is theoretically half -- running at half capacity, maybe $5,000 or $6,000 per car, it doesn't feel like a $20,000 difference. You know what I'm saying?

Correct. So when I say a fixed, I'm bundling here for simplicity's sake. But if you think of when we're running at half capacity, the staffing -- the land layout is not set up for half the capacity. So the number of people on the line, so it's -- your variable cost is the same. So the number of people, the cost to run the plant is the same. All the nonpeople cost to run the plant remains the same. The depreciation, obviously, the overall not still the same, so it's divided by less vehicles. So when you put all those together, about half the reduction comes from volume. Now the other half is made up of really 2 parts. One is we have a step-up in our ASP that -- we're already beginning to see, and you can see this -- you'll see this quarter-over-quarter happen, but we had pricing that we'd originally announced and locked to -- or linked to, I should say, our early configured vehicles. And that early pricing had a big step-up that represents roughly another 25% of the step down -- or step down -- or step towards profitability. And we're already seeing that. So we're now starting at the very beginning of blending in new orders, which are much higher pricing than the older orders. And then the third component for us is the bill of materials. So it's -- that's achieved in really 2 ways. One is the commercial negotiations that we're having today with suppliers. And the other are some of the things we've spent some time talking about today, which are just technical changes we're making to the vehicle simplification, consolidation of componentry, new drive units. We just launched a new dual-motor drive unit, which is a considerable cost-down relative to what we launched with. But all those things together get us to a healthy margin structure.

Right. Yes. Just on the fixed costs, and again, like I was in sharing nature and can't help -- can only think about numbers. So pardon me, if I'm like making this too simplistic. But if your facility is running at, I don't know, roughly 25% capacity...

That's roughly where it's been, yes.

Yes. And so your fixed costs are maybe 4x higher than they typically are. If I think of fixed cost on a vehicle, again, like maybe labor is, at best practice, $3,000 per car, and depreciation is at $3,000. So that's $6,000, so you're 4x higher than that, your 18 -- you are $24,000. So it'd be an $18,000 savings. So that's probably about what we're talking about sort of directionally.

Yes, it's -- you're doing the math. I get it's about -- you can see the numbers...

[indiscernible] pretty hard.

Yes, low volume. I mean it's one of the things on auto business why running under capacity is really challenging. It's very, very...

Especially running significantly under capacity.

Running 20%, 30% of the capacity is okay. Running 1/4 of the capacity plan is hard.

Yes. Okay. Maybe we can just talk about your markets for R1 now and then how Gen 1 and then how kind of Gen 2 will evolve. When I look at the global market for SUVs above $65,000, it's about 900,000 units, with less than half in North America. Maybe you see it differently. When I look at the pickup market above $65,000, it's like 300,000, all in North America. And so I think you've talked about, I don't know, maybe 10% share, but I think about those markets, particularly since you're really North America-focused right now, having less than 1 million units, you have a facility that I think is supposed to build 150,000 units, am I right on the market sizing? And what is realistic for your truck and your pickup given the addressable markets?

Yes. Yes. Well, there's a couple of things worth noting for the flagship product for R1. The first is the way we designed it, and the intent was that it would pull in new customers, that it would be what we think it was like a white-space product that's not just pulling in customers from an existing segment but creating new types of customers. And so the tour is a focus we had, where, first and foremost, we want to create new EV customers, meaning we didn't want to just move a customer from one EV to another EV but create net new EV customers, and we've done that. So north of 80% of our customers have never owned an EV before, which is great. But the other is we didn't want to be pulling customers simply from within segment. So if we take the R1T as a case in point, over 60% of our R1T customers have never owned a pick up before. So we're not only creating new EV customers. We're also creating new truck customers. And so that has the advantage of having a really broad spectrum of different types of vehicles that our customers are coming out of and, therefore, expands the market beyond, as you pointed out, the 300,000 units that exist above $65,000 in the truck space. And so what we're seeing is the combination of features and performance and refinement, and the desirability of the vehicle is just very effective at pulling in customers across so many different segments. And we're not seeing it in the data. So we're, by a significant degree, the highest-selling vehicle at a price point of over $70,000. So -- and we're the -- in Q1, we're the fifth best-selling EV in the United States. So that's more than much lower-priced products. We're able to demand -- create demand around a high ASP on a form factor that has broad appeal outside of EV within some segment. And so we see that opportunities to continue, particularly as we make improvements to the product and we launch different variants. We have a, what we call our standard pack, which is like a base pack, which we're launching fairly soon. We have our largest pack, which we call our max pack. We're just now launching a dual motor in addition to our quad motor. Quad motor's sort of [ nut ] sits 4 motors, 835-horsepower. Our dual motor is 700-horsepower, 2-motor setup. So we're broadening the number of portfolio variance we have on R1 to more fully populate the demand curve across a range of price points. And we think that, that supports full utilization of the plant capacity and hopefully and then some but to continue to carry a backlog or a bit of a wait list, which is actually helpful for production.

Right. And so the normal plan at 150,000, you think can be just for the pickup and the SUV?

And we also have the commercial...

No, that's what I'm saying. So like what are your target volumes excluding commercial vehicles?

Yes. We've talked about this publicly. The R1 platform, we're targeting 85,000 units, so between R1T and R1S, with the remaining capacity dedicated to commercial.

Okay. And do you -- I mean is Tesla informative at all because they had 2 high-end products, did some variants with Plaid, et cetera, et cetera. And they've been largely global in nature. And they did a couple of years at close to 100,000, but it's been tough. It's been well below that. And I think on paper, those markets are -- well, certainly, the Model X is analogous to your SUV. And I think the Model S is addressable market on paper is extensively a little bit bigger than it is for the pickup. So is there -- how does that inform you? Because I feel like Tesla has gotten to the point in its market segments where it's done a very good job of attracting customers. But there's still a pretty large portion of the population, maybe particularly off the coast in the U.S. that just don't have EVs on their mind and may not get there for a little while, right? So you're increasingly -- kind of it becomes increasingly difficult to attract the customer. And I just worry about why you have the confidence that 85,000 is the right number, kind of the -- an achievable target?

Right. You just said something which I think, in part, is a key element here, which is we're not competing with just EVs. We're just competing with Tesla. The vast major of our customers are coming out of ICE vehicles. And in the ICE space, Tesla does not really answer the need for like larger high-functionality vehicles. So Model X is 3-row SUV, but it's not really used as an SUV. It's -- by the architectural decisions they've taken with the closures on the vehicle, you can't put anything on the roof. So you can't put a surfboard on the roof of Model X. It's not something you put a lot of gear into. So it's a vehicle that doesn't really answer the large SUV need. And so that's one of the things we see is that customers that have either minivans or 3-row SUVs, big ICE vehicles like suburbans, those kinds of things need something that can carry their family and their gear. And the size of that market and the willingness to pay in that market is healthy, it's high.

Okay. And you've talked about like orders extending into 2024. Can you talk at all about sort of the rate of change of orders or backlogs in a way that makes you confident about volumes going forward?

Yes. One of the things that's really fun and helpful for us where we are right now is when we sat here last year, I started the conversation this way. The number of Rivians that were on the road was still small. So the awareness of the brand was not as high. Seeing Rivian was sort of a big deal on the road. We now have tens of thousands of Rivians on the road. And so awareness continues to grow. And our biggest advocates by far, are our customers. And that's -- we see that in the many different customer groups that are forming. So there's a multitude of Rivian groups that have come together in different cities or different regions that are going out and exploring. They're doing things together. They're creating this strong sense of community, but they're also advocating on behalf of the brand. And this is really wonderfully encouraging for us. And as we see more vehicles on the road and that -- and the role that customers are playing and telling the story, we're seeing that translate to more excitement, better awareness, which translates to more demand. And the thing that really is unlocked is initially most of our production center on the R1T as we now shift to producing more R1Ss, so this quarter, we'll produce more R1Ss than R1Ts. And long term, we will stabilize at just north of 2/3 R1S, 1/3 -- a little less than 1/3 R1T. But that's a profile that even some customers may know of Rivian, but they think Rivian in the context of R1T. The other thing is we're starting to see a lot of measures of this beyond what we can say here. So J.D. Power does an annual survey of customer satisfaction, and we came out as the #1 brand in that in our first year of production, which was awesome. So that's like above Audi, above Porsche, above Tesla that have -- and that's from customers saying, this is the highest level of customer satisfaction in the automotive brand available in the United States. So things like that translate to confidence. They also translate to actions on behalf of those customers that they are excited, they are enjoying the product and they're making it well known to their friends and family. And I think there's some sales from customers happening in this room this morning, but that kind of activity is really important for us.

Right. Right. Yes, because in following Tesla closely, I mean, the rate of change of backlog was ultimately sort of the highly, highly effective in predicting where unit volumes were going or necessitating prices and change, so -- changes in price. So I'm sure you look at daily orders and ready to change your backlog. And is there anything on that, that I mean, qualitatively, I presume that's what provides some of your confidence? Is there anything on that dimension that you can share that would be helpful for investors?

No. Just -- I mean, as you said, I think we -- we've watched the demand. I mean it is a daily thing that we watch. We have to be careful not to overinterpret discrete events that are happening and look at enough of a time horizon to pick the true trend. But last summer, I would say it was the most extreme where you just had -- demand was almost -- it was off the charts. It's just the level of demand for everything was so high. We then saw the beginning of this year and end of last year quite differently, largely with just a lot of concerns around what's happening from a macroeconomic point of view. I'd say we're now in a state of seeing confidence, not necessarily fully recover from a consumer point of view but just the stabilization of behavior, where end of last year, beginning of this year, we were sort of -- just sort of sitting tight. And you can see this in lots of metrics just in terms of how behaviors are shifting. So we're watching really closely what's going to happen at the macro if that's going to introduce some major shifts, but we continue to see demand for the products, and the functionality of the products helps us here. It's not like this is a 2-seat sports car that's a purely discretionary purchase. These are vehicles that get used as family vehicles. So they're the more resilient than, I'd say, some of the discretionary brands or discretionary premium purchases.

And is the more difficult challenge in sort of gauging trajectory in orders or in cancellations, right? Because I imagine a year ago, everyone's like, "Holy c***, I need a car. I'll just put a deposit down." And so what is the bigger variable? Is it actually the input funnel? Or do you actually watch your realization rate more?

Yes. Well, it's actually -- we're talking about what do we watch. A metric that we actually pay even closer attention to is cancellation rate. And within that, we say, what's the reason for cancellation. And by far, the largest contributor to our cancellations are wait time. I need their car in the next few months. You can't give me a car until next year. I need to get something. So that's something we're very cognizant. It comes back to the need to ramp. And part of a challenge like on R1S of having you to put an order in today, and depending on configuration, it could be late '24 or early '25. It's just a hard position to be in as a buyer to say like you have to defer the purchase by 1.5 years or more. So getting wait times into more reasonable levels is going to be, I think, helpful as well.

Right. So you've kind of teased the R2 platform. Maybe you can just provide an overview of what that vision is? I mean is this something that you think can expand your market by multiples? And maybe you could just talk a little bit about how you think about that.

Yes. So we have teased a little bit about R2. It's tough because we have...

Is there something behind this curtain? Yes, that would be fun. That would be the way to do it.

It's one of these situations where we have highly asymmetric information. So we know everything there is to know about R2. We've seen it. We're like working on it. We talk about it, and it's a bit esoteric, and like, what is this thing? We're very excited about this product. It's perhaps the most excited we've been about any product because it takes all the brand ethos and excitement that we built into R1 in terms of functionality and broad usability across a variety of different driving conditions and applies into a much smaller form factor at a much lower price point. And from a product or a technical point of view, it's a chance to take all the learnings/battle scars of launching and ramping R1 and embed those into a vehicle from the early stages of the architecture definition. And so on every metric, we're excited about it. We think that the price point and size have it hitting in a really large addressable market. So price points starting at around [ $40,000 ], growing up to around [ $60,000 ], but right in the [ mean ] of the market. It's leveraging everything that I've talked about that some of which is going into R1 in terms of network architecture, ECU consolidation, body architecture simplification. So it's just really elegantly done in terms of an architecture. And it does all that in a way that could deliver performance and attributes that are, we think, position -- will position it extremely well. So we will be showing it in less than a year, and that will help a lot just for folks to see what this represents. But it's -- but for us, it's a very important program. It really represents the scaling -- platform on the scaling vehicle for us as a business. And you alluded to Tesla. Tesla had S and X as their flagship products and then launched 3 and then eventually Y. For us, R2 is really that volume platform that takes the brand and the ethos of the brand and applies into much, much hire volumes.

And how many models will the R2 platform ultimately generate? And how do we think about [ your own ] time frames for this?

To start, it will be just one vehicle. It's another learning that has been borne out of the R1 or the first launches. We launched 3 vehicles at the same time, which created immense pain and challenge, which we've talked about. A vehicle launch of thousands of parts coming together from hundreds of suppliers, any single one of those parts can stop production. We had to go through that in the middle of a pandemic, and with the supply chain crisis, we had to do it 3x. So all that was multiplied by 3. And any mistake that happened in one of those vehicles from a supply chain or a process point of view, again, multiplied by 3. So the compounding challenges we experienced over the last 12 months were quite extreme. With R2, we're going to launch a single variant to start, simplify the launch, simplify production and really remove the challenges associated with like issue resolution. When you have this many products you're launching in parallel, just there's so many issues that they're resolved in parallel. So this is a big enabler with R2 of how we thought about it.

And do we think of that being a basis for many more? So should we be thinking kind of '25, '26 for the initial R1 product? And then subsequently a product cadence that might be every year or 2 thereafter?

Yes. There's -- so the R2 platform will have a few variants for sure. We haven't announced what those variants are, but I'd say there's at least 2 core variants on the R2 platform that are distinctly unique and different but address different parts of the market.

So we started this conversation around the auto industry, and you were talking about kind of keys for success in kind of getting scale and having sort of a new electronics space architecture, et cetera. I guess, part of me wonders, Rivian is really small. Like is it, a, should it be doing all of this vertical integration, autonomous driving, your own charging network? Is that really realistic to do, question 1, and potentially raw materials deals or ownership? A, is that -- why are you doing each of those things from a vertical integration perspective? And then b, like is Rivian better off alone or with a well-capitalized owner?

Well, to answer the first question around the things we're vertically integrating. So electronics, software stack, our propulsion platforms, so drive units, inverters, gearboxes, those things. We think that's a necessary component of success to control that stack. And we've capitalized the business in a very healthy way to allow us to do that, and we're deploying capital in those R&D domains to ensure we're best in world, best in class in those categories or in those areas. And we're going to see the volume products that will leverage those come soon with R2. And as I said, within the electrified space, while we're small as an overall company with the electrified space, we're -- in the markets we operate in, we're fairly sizable on a relative basis to others. So I think the question then becomes is how quickly do others electrify in terms of what's the percentage market penetration in the EV space versus the overall space now. That links to your point on us remaining independent. That's how we've designed the business. That's how we've architected. It's the reason we're building all these core technologies in a vertical fashion. I do think that the -- this is -- we could spend 10 minutes on this at least. But the benefits of a clean sheet in terms of distribution, technology stack, supply chain, brand are quite significant. There's also challenges with that of having to build things yourself, so building production capacity and building production know-how. But there's implicit advantages of being able to design an auto business really at its core around what the future state looks like. So for example, we're not going to go build 4,000 physical retail locations to deliver our volume. I think that's an antiquated way of having to deliver a couple of million cars a year in the market. But that's a burdened legacy cost that traditional OEMs have to deal with. And it's a really big challenge in the long term if this like overheavy, overexpensive go-to-market channel. This is well documented. But we just don't have that. So we get to design a digital-first, digitally native transaction platform. And that extends then into thinking about the customer model in a very different way, where there's software features, there's ability to have the vehicles get better over time and charge for those types of things.

Right. I'll ask you one more, and we need a short response because we're out of time. But against that framework, why your own charging network, which is capital-intensive and there are partnership opportunities? And then secondly, autonomous driving, there are like 40 people doing trials, almost all of them are better capitalized than Rivian is. AI is kind of a scale game. Is it really realistic to believe that Rivian can be on the leadership front given its size? So I'll leave you 30 seconds, 2 really tough questions.

Those are tough to answer in 15 seconds. On the charging network side, there's one good charging network in the United States today. I'd say that's Tesla's network. And the rest of the networks represent -- to varying degrees, they're suboptimal in various ways, either the charging uptime isn't there or the charging speeds aren't there. The overall interface of the platform, so it's choppy at best. So we took the decision several years ago to develop in-house our own charging platforms. We just develop the hardware, the software, so it's a DC fast charger up to 300 kilowatts. It's a high-speed charger. And we're now deploying that we have a small number of sites we've launched today. We call it our RAN network, Rivian Adventure Network. So about 35 sites today, a couple of hundred that are in process. And for us, it's -- you could think of it as a cost to facilitate charging. You could also look at it as a very powerful marketing tool. But I think the thing to point out is that while nothing -- well, we're not in a position to talk about it, but it's not as if this network is being done in a vacuum without thinking about the partnership opportunities that are out there. So there's -- we have a unique asset, and we're one of a very small lever of companies that has a domestically built, domestically designed charger with now a mountain of capital in terms of dollars available from U.S. government to go build it. So we think we're well positioned to continue to scale that network and to build lots of partnerships and coalitions around it. On the self-driving front, you referenced there's 40 different companies that work on this. I think we have to be really careful in self-driving. It's easy to homogenize it all and say that all self-driving is the same. There are platforms and companies that are working on like Level 5 architectures that are very hardware-heavy and sort of are designed at the get-go from a steering wheel-less vehicle. So I think Waymo is an example of this. Cruise as an example of this. That from a topology point of view -- technology and topology point of view is entirely different than what, let's say, us or Tesla are doing with more like Level 2 or Level 3 systems. I'd call them hardware-constrained systems, systems that you can't spend $100,000 in sensors set and compute, but you can spend a few thousand dollars in sensors set and compute. And those hardware-constrained systems are getting better and better and better. And we will eventually see these 2 converge, but the benefit of a hardware-constrained system that goes into deployed fleet like Rivians, is you get the scale, the massive scale of a learning loop that feeds back into the platform. And I think a massive enabler that's changed the whole way we think about autonomy is the role of ML and that the system benefits from mileage accumulation and the system benefits from things like auto labeling. And so we're building that platform. And to do that, you have to kind of own the system. You have to own the cameras, you have to in the compute, you have to own the training [ loops ], you have to own the training models. And it takes a lot of time and effort to go do that. We've done that in -- what customers will see is our features will get better and better and better over time through software. And then there's some key hardware updates that come that allow that to further accelerate. And I think this is an area where controlling the stack does create significant operational efficiencies in terms of the rate of improvement.

Great. Sorry, we ran over. Thank you for your time, RJ. Really appreciate it.