The Future of the Auto
Industry is Data. Which Means Tesla Has Already Won.
The car business used to be about engine
displacement, styling, accessories and performance. But Elon Musk's company has
changed the paradigm as well as the power plant.
General managing partner, G2T3V and Chicago
High Tech Investors
For the last several decades, marketing in the U.S. automotive
industry has been mainly about distinctive style, speed and performance, and,
to a lesser extent, safety. From time to time, any given manufacturer will run
ads that stress one differentiating factor or another and thereby try to
temporarily set themselves apart from the pack. This approach has rarely been
successful in large part because the claimed distinctions - even if potential
buyers understood and valued a duel overhead cam engine and the like - were
mostly smoke and mirrors. As we said in the early days, the only difference
between a car salesman and a computer salesman was that the car salesman knew
he was lying. In addition, the sheer glut of car ads basically turned
everything gray.
And in
the last 10 years, cars have gone from being a badge of status, conspicuous
consumption and patriotic pride ("See the USA in your Chevrolet") to
expensive, carbon-spewing sources of pollution. Millennials aren't even rushing
to get their drivers licenses. They see a car as a costly burden of fuel,
finance, and insurance that spends more than 90% of its useful life parked and
depreciating. Much like cable, which these days is a grudge buy, everything
about owning a car is a drag or worse. We are well on our way to viewing
our vehicles primarily as nuisances and incidental tools for occasional
transportation -- which could be leased, rented by the hour or the mile,
shared, summoned on demand or done away with entirely. Any and all emotional
connection is largely long gone. There's a simple reason that no one has ever
washed a rental car - we simply don't care about it.
And
then there's Tesla. Delivering the first car to Elon in 2008, an amazingly
short 12 years ago, the $109,000 Tesla Roadster was the bomb. Clean and
California cool - it brought sexy to an industry desperately resistant to and
in need of new thinking and new blood. But more importantly, it elevated the
public's perception of what an electric car could be in exactly the same way
that Apple changed the game with computers. Apple magically transformed
personal computers (and later portable music, cellphones and digital watches)
into items of desire. Steve Jobs restyled the boxy, gray, gear sold in
hobby shops into offerings of pure white wonder sold in stylish and beautiful
retail temples. He reduced IBM to the GM of personal computers. Tesla
was built on the same conceptual marchitecture.
Tesla made it exciting and
aspirational to think about owning a car again even if you couldn't remotely afford
to do so. Interestingly, and much like the early Camaros and Mustangs, the cars
themselves were surprisingly cheaply and modestly appointed on the inside, but
no one really cared. They represented innovative and radical change, a triumph
of technology, and the idea of a new kind of luxury (and exclusivity) which
their owners could share in. Detroit, in the meantime, is the "same old,
same old" surviving on pickups and SUVs which are dying breeds and simply
digging itself deeper into a dark hole every day. Trying to catch up with the
past is a fool's errand.
All the
major manufacturers will soon roll out luxurious electric cars, which will be
solid, beautifully styled, quick and expensive. But the really significant guts
of the cars, the foundational AI and autonomous features, will be yesterday's
news. Think of a decked-out golf cart tied to a two-year-old Android phone. The
race and the game have already shifted from steel and sensors to smarts, and
the smarts depend entirely on the size of the data pool your machine learning
algorithms can draw from, so that they can keep getting smarter and smarter.
Detroit will be sniffing around and stirring the ashes of Tesla's old campfire
while Elon's army is already over the next hill. In racing, you learn early on
that the other guys never wait for you. The race to build the smartest car may
already be entering its final stages. Good gets better, bad gets worse.
It's
all about scale and numbers and, as of the end of 2019, close to 900,000 Teslas
have been sold, and these vehicles are feeding boatloads of experience data,
driver behaviors, performance reports, etc. every single day into the smart
systems in Palo Alto. It's a machine learning flywheel that just continues to
accelerate. No one in Detroit has even a tiny fraction of this
number of cars on the road. GM may be spending a couple of billion to build a new
EV plant, but it'll take years to get the production and sales
numbers up to any material level. The volume of new vehicles that any of the
big guys will be fielding over the next several years is just another drop in
the bucket while Tesla is building an immense wall of data and an
insurmountable moat. There's no shortcut in the process and there's no
compression algorithm for actual roadside experience and data.
In
fact, it wouldn't be much of a surprise to see Tesla developing an entirely new
revenue stream of selling its data to the laggards in Detroit much like the
early mapping companies tried to do before Google ate their lunch. Data never
dies.