Though you wouldn’t know it from talking to most industry experts, there is an EV revolution unfolding at this very moment, right under their noses. Counterintuitively, “the experts won’t see it coming” is an indicator that a technology disruption is happening. In fact, most evidence suggests that Tony Seba’s predictions from his 2014 book Clean Disruption, are happening as and when he predicted. Which means we’re within 36 months of the collapse of the price of oil and in the sale of gas vehicles. Which is probably worth talking about.
The first questions are typically “how can this be true?” and “how is it possible the experts can’t see it?” That’s the curious byproduct of a technology curve invading industries whose companies don’t think of themselves as ‘technology companies’, such as Ford, GM, Mobil and Exxon. Technology companies see the world through technology growth curves, whereas non-technology companies see the world through linear growth curves. It’s like a circle visiting the land of squares...they use the same basic ingredients but they interpret it in completely different ways. One views the world as corners and the other as curves, so there’s a fundamental gap in their understandings of how the world works. And when your livelihood is built on either corners or curves, there’s a huge mental barrier to seeing things the other way around. The transportation and energy industry experts can’t recognize what’s happening because they’ve never seen it before, and their mental models are based on rules which are no longer true.
For example the industry news source Autoline Network covers industry news with experts from across the supply chain, OEM base and analysts. They’re hearing that although EV’s will eventually take over the transportation industry, it remains at least five to ten years away before things really get serious. That is completely wrong. As I’ll show shortly, things are already very, very serious, and any company that hasn’t already put their shoulder into transforming to the new competitive model is likely going to exhaust their resources before they can make required changes.
I learned as a management consultant that when an executive sets their mind to a forecast, there is almost no information which can change their mind. Unfortunately the mechanics of disruptions are particularly susceptible to misinterpretations, since the bottom of an exponential growth curve appears to match the slow, linear growth that incumbent executives are used to seeing. Linear growth fits their mental model. But the basics of this disruption is as follows:
- This is a technology disruption fueled by the confluence of EV, AV (autonomous vehicle) and battery technology converging at price point enabling mass scale
- Technologies always follow an S-Curve of market growth.
- S-Curves are inherently exponential growth curves, not linear growth curves.
- Experts in the technology industry recognize S-Curves, experts in the transportation and energy industries, who do not view themselves as competing on technology, do not.
Based on “not-an-S-Curve” thinking, the auto industry is forecasting that all EV and PHEV vehicle sales combined will reach no more than 6% of US new vehicle sales by 2025 or 2026. That equals 1.02M US new cars sales out of 17M total. Wards Auto and the National Automobile Dealers Association both cited that forecast when discussing the future of the automotive industry. For comparison there were a total of 329,000 US electrified vehicles sales in 2019, (239,000 BEV and 90,000 PHEV), which equals just under 2% of new car sales.
Up until the 2012 introduction of the Model S, EVs had been limited to city cars and second cars, not only-cars. For an EV to be an only vehicle it needs to have over 200 miles of range (preferably 300+), allow fast charging, but to be mass market it also needs to be affordable to purchase. The first car to meet all three criteria was the Tesla Model 3, whose cumulative sales are plotted in red on the graph below. The Chevy Volt, Tesla Model S, Nissan Leaf and Toyota Prius Prime are the top four grey lines.
The reason the red line on the graph above doesn’t appear to follow the same rules as the other lines is that it doesn’t; it benefits from the technology advantages converging with cost advantages, and the impact on sales is evident. The growth curve of the Model Y and Cybertruck will be steeper than the Model 3’s since both were designed to be easier to manufacture and are facing similar or better demand.
Another diary will return to the subject of robotaxis because those will drive the steepest part of the EV adoption curve (the S-Curve), but for simplicity sake let’s ignore that for the moment. Let’s assume that EV adoption follows the replacement model rather than the TaaS model, while still acknowledging that robotaxi services will become commercially available at scale in more and more communities by 2025, even if they aren’t generally available in most states by then. I base that argument on the fact that robotaxis are already commercially available in the Phoenix and San Francisco areas, with smaller trial runs happening in Colorado, Michigan and Nevada. With the exploding investment in autonomous technologies the number of vendors offering trial and/or commercial services will continue to grow.
What isn’t clear in the graph above is that the red line is only possible because of the successful scaling of battery cell manufacturing at Giga Nevada (formerly GF1) by Tesla and Panasonic. If vehicle manufacturers don’t control their battery supply, they don’t control their vehicle production. Which is one reason it’s so astounding that although none of the incumbents outsourced their engine design or manufacturing for their gas vehicle lineup, every one of them has outsourced the design and production of the battery cells which will determine the performance of all their electric powertrains. Well, the batteries and the software combined, will determine their performance, and legacy manufacturers have historically outsourced both. Which leads us to the heart of why this disruption is going to be so dang disruptive: it changes the competitive basis of automobile offerings away from incumbent core competencies and into areas where they lack expertise.
Historically the automobile industry has competed through the emotional appeal of it’s brand as expressed through the sounds, feel and design of vehicles which performed more or less equivalently. In practices the industry calls segmentation and stratification, a mid-level economy car from one manufacturer will cost and perform similarly to the comparable model from another manufacturer. The differences will be aesthetic and emotional rather than based on measures and performance. For manufacturers that means the brand identity must combine with the ‘personality’ of the engine to drive the design decision making. Until now. How do they design a vehicle without engine noises or torque curves? That’s not what they’re experts at.
When every company in that market needs to acquire expertise in battery chemistry and cell manufacturing, and in software for vehicle operating, infotainment, climate control and navigation systems, there’s suddenly a huge push to acquire talent. But there are limited experts to go around, and the top professionals are going to take jobs with a future, not just a paycheck. They can visit a battery laboratory and testing facility and tell whether that company has a chance of competing or not, on day one. The top experts aren’t signing up with the also-rans, they’re going to organizations with technologies that are already competitive.
Batteries provide the path to being the lowest cost provider to move 100kg 100km. The platform which can achieve the lowest cost per passenger mile will win the robotaxi wars, whenever they come.
Tesla’s vision for the future was discussed in 2012, when Elon Musk announced the first gigafactory. To control EV production requires controlling battery cell production. Tesla’s Giga NV supplied the production of over 300,000 vehicles and over a GWh of stationary storage. But as of this writing, there are 97 gigafactory projects underway, and that doesn’t include the terawatt-scale cell production plans Tesla is expected to announce in April. Which means EV production could theoretically reach 100x the volume of Tesla’s 2019 sales, or over 30M vehicles/year by 2025 or 2026. Even half that would be looking at enough battery supply to manufacture over 15M cars/year by 2025. And while those numbers sound impossible at first, it’s important to know that they are consistent with Tony Seba’s predictions and model for disruption.
Sales of gas cars are going to crash over the next 36 months because buyers are going to realize that they can’t re-sell their used gas cars. This relates to robotaxis in the long run, but it’s already happening because of ride share services. Ride-hailing services accounted for 20% of passenger vehicle miles driven in San Francisco in 2018, and half of those we ride-sharing. This means that more travel is being done by using the ‘phone car’ and less rides are going by ‘real car’. As more car owners opt for Taas and either sell or don’t buy a car, the demand for used cars will drop just as the supply begins to climb. This will cause the price of all used cars to plummet, which increases the spread between the cost of a used car and the cost of a new car. That new value gap will make the purchase of new gas cars increasingly unattractive, and sales will collapse. That will accelerate the death spiral of gasoline engine vehicles in the US. And it will happen over the next 36 months.
If there are readers in the Scottsdale or San Francisco areas who have direct knowledge of used-car prices in the local area, please share your observations in the comments.