CarTalk # 49 (A chat with Steve D. Geskos from Rose Park Advisors)
..who happens to be a prolific automotive investor and a mobility futurist
Hi friends
I’ll say that I learnt so much in the course of this chat. There’s much to read and absorb in this issue, so let me get out of your way.
Buckle up!
Sachin Seth (SS): Steve, I want to start with thanking you for being one of my best readers. I love your engagement – instead of just saying “good issue”, you reply with anecdotes, reports, thoughtful pushbacks that enrich my worldview. It’s why I want to bring our dialogue out in the public eye!
Steve Geskos (SDG): Thank you! I find your work to be very stimulating and the way you emphasize your content and provide unique points is something that I genuinely enjoy!
SS: All those late nights were not in vain, then! Okay, zooming out, I wanted to start with the investing firm you work for and its investing philosophy, which as I know is rather unique relative to a lot of other VC shops. Mind talking about that briefly?
SDG: Sure! Rose Park Advisors (RPA), which is where I work, invests using the principles of Dr. Clayton Christensen, who was a professor of business strategy at Harvard Business School for many years. Dr. Christensen co-founded the firm with his son, Matt, in 2008. Dr. Christensen developed the frameworks of disruptive innovation several decades ago. Those frameworks have become the handbook for entrepreneurs and public companies to address the question of “ how to build and sustain a competitive enterprise”.
(Note for readers: Harvard Business Review consolidated all of Dr. Christensen’s seminal research at this link)
SDG: Dr. Christensen realized that all the impact he had made in the academic and business world could be directed towards the investing world by applying the very same frameworks. RPA searches for businesses that already exhibit the hallmarks of disruptive innovation, or can be coached into doing so!
SS: What stage does RPA invest in?
SDG: We're stage agnostic. We’re flexible in that we look for disruption in every form it could appear. At the same time, we’re highly disciplined in that we only invest in companies that fit Dr. Christensen’s disruption frameworks.
SS: So different from the VC spray-and-pray model! Got it.
SDG: Exactly. Disruption is not about the technology per se, but the business models enabled by advancements in technology. The classic example of this thesis is that Amazon and Netflix didn’t really invent anything new. Both harnessed existing technologies, repackaged and reformatted them, extended them in different ways that were disruptive to the incumbents or opened entirely new markets that didn't exist before.
SS: True. Netflix didn’t disrupt content, it disrupted distribution of said content. I do also want to capture your background, Steve! I know you're looked up to as the automotive expert within RPA - how did that come to be?
SDG: When I started my career in the early 2000s, it was right after the .com boom. Back then, it wasn't a heretical statement to say that many old-school cyclical companies might not exist. Everybody thought that “not very interesting” companies like Caterpillar or Boeing will go the way of the Dodo bird. This was also prior to China and emerging markets erupting onto the scene.
In the aftermath of the dot-com saga, some of the leftover sectors that a young analyst like myself would be responsible for covering ended up being cyclical industry. That’s how I ended up researching aerospace, defense and multi-industry companies – any “smoke stack that moves” was where I cut my teeth.
SS: Wait, did you really say “a smokestack or that moves”? Did I hear that right?
SDG: <Laughs> Yes, anything that moves or has a smokestack. This covers the entire energy industrial complex and transportation complex. All these sectors are cyclical. Cyclicality in business cycles teaches you a lot about the resiliency, about the foresight and planning needed to survive in highly competitive markets. There's far less room for error in cyclical industries. If you have a gross margin of 20-30%, versus 70-80% in Tech, and a much different cash flow profile, that creates very different types of management skillsets and business acumen to execute. I started following a lot of public companies, both on the OEM side, as well as the supplier side and the overall ecosystem as it related to transportation and logistics. Automotive is global, competitive, and capital intensive. It teaches you to think bigger picture very quickly!
SS: Well said! On cyclicality - the current upswing of automotive takes the name of “new mobility” as you know. Electrification, autonomy, connected, shared paradigms. I am curious to hear how you've been applying Dr. Christensen's frameworks to new mobility.
SDG: I think a lot about how Dr. Christensen’s frameworks apply to new mobility. Everybody focuses on electric vehicles. I like to think about electrification merely as the enabling technology that will change how the powertrain will be treated.
Vehicles are becoming a completely different type of platform to produce and to interact with. On the production side, it was no secret 10- 15 years ago that most Tier-1 suppliers could make an automobile. Continental, Bosch - all the big Tier-1 suppliers had the capabilities to make a vehicle. They just chose not to because you needed the systems integration, the ability to deal with the consumers, the ability to market to sell a vehicle. The biggest reason, however, was optimizing around bespoke internal combustion engines that these players just didn’t want to get into.
What was lacking was standardization and modularity around engines. There wasn’t a plug-and-play architecture relative to the internal combustion engine.
SS: Mind expounding on this idea more?
SDG: The idea of standardization and modularity is very relevant to Dr. Christensen's frameworks - what can you buy off-the-shelf and integrate, vs having to engineer yourself. For example – when Dell came on the PC scene, the real disruption to the industry’s business model was that computer components had become good enough and there were industry standards that let you configure a PC to your liking. When ordering a PC you could choose your type of processor, your hard-drive, memory, mix and match them and even upgrade over time. Standardization of interfaces and modularization of components enables the same for electric vehicles, which will inform the strategic questions around how OEMs compete in the future, how their suppliers fit into the ecosystem and how their consumers will be best served.
SS: A Lego-like EV, then? Is modularity in EV the real disruption?
SDG: To a large extent. The concept of modular architectures hasn’t been pioneered by EV companies. It was pioneered decades ago by the Japanese and German OEMs. However, the limiting factor was always the powertrain – the beating heart at the center which was the gas engine. Today – the beating heart is the battery. Switching to electric gives you the skateboard where the drive motors and the power source can be designed to provide commonality regardless of the top hat (vehicle body) that we put on top of this skateboard. We’re coming up to a point where no one cares about engine specs like they used to – “does it have 4 or 8 cylinders” is not only anachronistic, it is meaningless. In new mobility, it’s the software-enabled experiences (infotainment, path planning) that dictates how consumers will interact with vehicles and rank them.
SS: Do you think this shift to modularity breeds a new class of OEMs?
SDG: I do. As components become good enough and become commoditized, you can control them in different ways and manufacture vehicles in different configurations, going back to my Dell example. This opens up the aperture for all sorts of new vehicles that can be created that heretofore were probably impossible. One can become an assembler of vehicles in its truest sense and be able to piece together the different building blocks for the vehicle in a much more powerful way. Modularity is going to revolutionize automotive over the next 10 - 20 years.
SS: In keeping with modularity, let’s talk battery swappability. I should let the readers know that you and I exchange many spirited messages about this tech (my view are captured here). I continue being on the fence about it. However, I listened to your podcast “The Disruptive Voice” with John de Souza from Ample – who are the US leaders in the battery swapping space. Is RPA an investor in Ample?
(Note for readers: very good read on Ample in TechCrunch)
SDG: We are an investor in Ample. I urge your readers to listen to the podcast so I don’t repeat the whole thing. What comes out clear in the podcast is that battery swapping is not for all vehicle types. It is only meant for certain types of vehicles, particularly at the low end of the market. Think high volume, low margin – fleets and liveries operating in highly dense urban environments, as well as in areas where typical charging networks won't be practical.
SS: So for fleets with high degree of utilization - medium range vehicles that are always on-the-go, right?
SDG: ..exactly! Or where people can't afford charging times nor do they have the luxury of charging overnight in a single family home.
SS: Is Ample aiming to manufacture packs that are inter-operable and swappable between different OEMs each with different form factors? Or are they just going to be developing the technology layer which allows any pack to mate with any vehicle?
SDG: That's a great question. Their battery packs will be relatively agnostic to the vehicle. The analogy from this era of mobility is that of octane ratings – most people don’t care about an 87, or a 91 rating. All they care about is that there's gas in the vehicle. In the new era of mobility, what consumers and fleet operators might want is.. just.. electrons! If you can get the amperes and voltages right and the resulting range, the actual chemistries powering a vehicle are largely irrelevant to the consumer. Think about consumer electronics, right? As long as your device works, do you really care if your AA cell is from Energizer or from Duracell?
SS: Agree, but at some point someone decided that an AA cell has to be X centimeters in length and Y centimeters in diameter. I don't know if OEMs are shaking hands on a standard battery size and dimension. For example - Tesla Model 3 is a small vehicle, Model S is bigger. The Mercedes EQS is a 17-foot long beast. Given that there's no standardization in chassis, how's Ample thinking through addressing all shapes and sizes of.. uhh.. electron-stacks that move?
SDG: Obviously I won’t speak in great detail but Ample is partnering with the OEMs, they're partnering with fleet service providers and ride hailing companies to trial some solutions. In some cases, they are swapping out the entire battery box. In other cases, just the different chemistries of cells. I’ll leave it at that.
Think of a world where an OEM can sell a vehicle without a battery, letting you buy the battery after the fact, upgrading it as new technology rolls around. You’re effectively getting a “power by the hour” contract which is very similar to the aerospace industry. In the aerospace world, the engine can come off the wing, letting you swap out engines between aircrafts of a certain type, as long as you have the right standards and connectivity. For instance, the CFM 56 can be swapped from one Boeing 737 to another, letting you get creative about cost management and lifecycle management of the aircraft.
SS: Oh wow. I didn't know that airplane engines are swappable. What was the motivating factor that allowed different airplane manufacturers to coalesce around standard engine sizes?
SDG: It was mostly driven by what the end customer (airlines) wanted, as well as the design requirements of how an aircraft was built. Aerospace OEMs like Boeing do not have engine capabilities. There are only three major engine OEMs in the world - Pratt & Whitney, Rolls Royce and General Electric. They form an oligopoly.
Aerospace OEMs – Airbus and Boeing – form a duopoly which buys their engines from the oligopoly. The aerospace powertrain is separate and distinct from the airframe. In addition, it is a completely different sales process between the airline, the airframe and the engine OEMs. At any one of the engine oligopoly, you pay an upfront cost and then have a “power by the hour” contract. Switch it back to the Tech world - you're paying in a SaaS model for the long-term maintenance and reliability of the engine!
SS: I hadn’t really internalized this. Yeah, there could be some parallels to new mobility here.
SDG: Absolutely. In automotive, you may soon start charging depending on consumer tastes. You can have business models that innovate charging for features on a $/mile basis – whether it's insurance, whether it's certain performance features after delivery, it’s a blank canvas.
SS: Now in this new mobility world, would you say that battery companies like CATL or LG Chem could be the Pratt-Whitneys?
SDG: Within Dr. Christensen’s frameworks we ask where the value will migrate to in the value chain. So the question is – will battery makers evolve and become so consolidated that they resemble the aerospace engine manufacturer oligopoly? I personally think the difference here is in the complexity, regulatory certification and in the capital intensity of manufacturing an aircraft engine, which lends to an extremely deep moat that is difficult to breach. It's a self-reinforcing moat – the more flying time and distance that you have over a potential competitor, the safer bet you are in the eyes of the FAA and airlines vis-à-vis an upstart engine company.
I'm not sure that the same drivers (reliability foremost, capex second) apply to EV batteries. EV batteries are still a cost play and the one with the best unit economics wins. As long as an OEM is getting a “good enough” performance, they’ll pick the cheapest battery provider. That does not apply in the aircraft engine world where a cheaper “good enough” solution will not fly, pardon the pun! If they’re a performance-chasing OEM like Ferrari, or Porsche, they might insource batteries to fully integrate and optimize performance and utility.
There was a time Ford made its own tires! With time, external vendors started making systems and sub-systems that were “good enough” which birthed companies like Autoliv which live to provide airbags to every automaker under the sun. However, outsourcing ICE powertrains was a no-go for a very long time. This is what's changed with EVs.
SS: I buy that. We've talked innovation and disruption in the electrification. Where do you see disruption happening in connected vehicles?
SDG: Ah! You can really talk out of school if you want to imagine a where things can go with connected cars.
SS: Talk out of school. Please!
SDG: We’re carrying around supercomputers in our pockets that lets us sense and process info on the fly and share with the broader world. Connected vehicles are just “bigger supercomputers on wheels” and unlock much more perception and sensing, whether it is vehicle-to-infrastructure, vehicle-to-vehicle, vehicle-to-human etc.
The limitation of phones is their energy density and form factor - you can pack in only so many sensors before you burn your pocket or it weighs too much. A car doesn't have those limitations. You can integrate sensors into it and create an immersive experience for the user. Aided by a supercomputer, you can network vehicles, do computing at the edge and create a distributed compute mesh network. This network can deliver significant services to the driver and occupants. The vehicle will be a powerful IoT device in the environment that it is operating in, while being put to use as a communications node.
SS: I get it. The vision is that the car can sense surroundings, communicate them to the internet, to each other, find ways to monetize this, all while mining bitcoin for you. I’m serious.
SDG: You think that’s wild? I remember once a French OEM telling me that if X% of their fleet in Paris was connected with the sensors one finds in autonomous vehicles, they would have enough data to tell me the size of the raindrops falling on the steps of Notre Dame, and at the exact same time could tell me the temperature of bread baking in a boulangerie half-way across the city. Anecdotes aside, the value of such data could be revolutionary for many different applications, many not even dreamt of today.
SS: Do you know of players in this space who are working on a “IoT device on wheels”?
SDG: It's very obvious to me that the Auto OEM themselves and technology companies are thinking about the connected vehicle as a platform. The most obvious one is Waymo. There are rumors of what Apple will do in the automotive world. Take your phone, plug it into a car’s electrical and sensing capabilities and you’ve basically put an Iron Man suit around your phone.
Just like the PC changed our environments – offices, homes – the same thing will happen as compute proliferates and permeates into the vehicle market and spills out on to the roads. Look, people will still want to go from A to B. But >90% of the time, a vehicle is idling. What it does while idling will be disruptive. A connected vehicle can take the form of networking infrastructure and unleash capabilities that are the domain of the Cisco’s of the world today.
SS: Vehicle-as-an-infrastructure. That’s a new thought to chew on.
SDG: A connected vehicle with cameras - everybody will have to throw out the idea of any privacy!
SS: ..and that is an Orwellian thought! Steve, are there other aspects of the automotive space, which we might not have touched on, that you consider ripe for disruption?
SDG: Disruptive often means that sometimes a nascent “not a good enough technology” can open a new market. As is often said in venture capital – most innovations start out as toys. In a classic example Dr. Christensen spoke about – when Honda entered the U S market, they made dirt bikes. The established OEMs like Harley Davidson completely ignored that market. “It’s just a niche”. Over time, Honda moved up market to more advanced dirt bikes and eventually high-performance motorcycles. And now they own a large slice of several off-road and on-road categories that incumbents like Harley Davidson have had a difficult time mastering or responding to.
SS: I hadn’t heard that “innovations start out as toys” but it resonates. The Honda example is a very good one because they have matured enough to provide the Formula One racing engine to the highest rated team (Red Bull Racing) this year. Quite the leap!
SDG: If we were to say in the 1970s that Honda was going to be providing engines to Formula One racing, you’d have been the laughed out of the room. My broader point is that learnings in this space will continue coming from the lower end of the market and that's where the real innovation will start. My message for your readers is - keep a close eye on what is happening in China and India. Whatever it is will start out as a toy and disrupt the entire landscape within a decade.
SS: Steve, you’ve been very generous with your time today and have given me much to think about! I can’t thank you enough!
SDG: It has been my pleasure! And this is one issue that won’t generate any pushback from me. <Laughs>
That’s all from me folks. Have a great week!
👇 Please hit the ❤️ button below if you enjoyed this post. Or leave a comment!