I can’t remember where I was—perhaps sitting in the back of a taxi cab—when I read a tweet from Anti-Fragility author Nassim Taleb that said, “To ‘Uberize’, remove the middleman, theme of the times.”
The thought struck me again this last week as I tried to find something out of CES in Las Vegas more interesting than a selfie stick. The energy and excitement, not to mention the valuations, in the economy lie in companies that ‘uberize.’ Even though those companies no longer make anything material, what they do seemed to follow a classic formula, an investing thesis that came out of the first generation of hyper valuation that took place in the 1990s.
I don’t think of Uber as a force that dis-intermediates—as we olds used to say—transportation, but one that creates value for itself, its drivers, and its users, by developing a new layer that integrates them all with maximum utility. A very talented developer once told me that the secret to a world-beating service like Dropbox was to make something very, very complicated seem devastatingly simple. To me, uberizing meant trapping a series of innovative processes—phone-enabled geo-location, payments and driver management and distribution—into an app-accessible service.
And that brought me back to Taleb’s tweet. Because Nassim is a neighbor and acquaintance, a guy I’m used to seeing sitting over salads and espresso discussing his ideas, reading his tweet began an internal monologue, the kind we seem to have more and more these days with the ubiquity of smartphones and the possibility of the kind of asynchronous conversations that social media enable. And Taleb brought to mind something another friend once wrote.
Andy Kessler was once a fund manager in Silicon Valley. He moved there in a flush of self-confidence in the mid-1990s, thinking he had a bullet-proof investment thesis. Although we remember that era as the beginning of the dotcom bubble—carbon-dated using the Netscape IPO in August of 1995—a lot of money was being made at the time in hardware, like CD-ROM drives and the first generation of cell phones.
Kessler raised a $70m fund and bought the stocks of companies that had pulled the dropbox trick at a basic engineering level and solved a particular problem elegantly. Then they routinized it, patented it and printed it on a silicon chip that could be sold to manufacturers of consumer devices like computers, printers and flip phones. When I pinged Andy to see if my recollection of the thesis that helped him turned $70 million into $1.1 billion in 18 months was similar to what Uber was doing inside its app, he reminded me that chips were only the beginning of the story.
Andy wasn’t investing in chips. He was backing clever guys who had found a way to share their innovations cheaply but still make a handsome profit on them. Andy’s guys were figuring out how to harness lasers to read CDs or display better graphics on video games. Today, the same process is happening with drones. At CES, drone manufacturers showed how an industry collectively moves forward finding challenges and provoking solutions. Chris Anderson, CEO of 3D Robotics, neatly sums up Kessler’s worldview with the contemporary industry:
The good news from CES 2015 is that the semiconductor giants are throwing billions of dollars of research and [production] capacity at problems we, the drone industry, need solved. So between Qualcomm’s work on real-time vision built into their Snapdragon program and Intel’s work on RealSense vision, which is a standalone chip, those things are now going to be driving next year’s drones, and they are going to be available at a cost and speed that we, the drone industry, could never have done on our own.
Focusing on new and advantageous intellectual property was a winner because that IP was where the profit margins would reside, as the chip companies that Andy first invested in two decades ago morphed into design and engineering firms that contracted out their manufacturing.
“Chip makers no longer had to make chips,” Andy wrote in an email. “Instead, they could just design their architecture or function and ‘print it in silicon’ which means translating the design into a set of interconnected transistors and having them manufactured (then) in Japan, Taiwan or Singapore and (now) in China. The chip company would own the intellectual property and not have to invest millions (now billions) on a semiconductor fabrication facility that might last all of 3-5 years anyway. These foreign foundries would build and upgrade the fabs and then sell enough ‘printed’ chips to cover their costs and profits.”
The American firms that had designed the chips to solve new problems sold them to manufacturers for much more than it cost them to contract the fabs to produce. That’s one reason the American chip industry moved overseas. As the supply chain moved toward lower-margin, capital-intensive aspects of itself abroad, value continued to accrue where the more complex work was being done.
The iPhone, for example, is designed in California but built in China. The profits, even the ones that Apple has been doing complex financial engineering to repatriate, all accumulate to the American firm. The Chinese, not just the chip makers but now the fabricators of every component of the phone, are left with table scraps.
Justin Fox just made a similar observation about Samsung and its brief moment in the sun. Because Apple owns the IP of its ecosystem, Apple harvests the profits. But in the Android ecosystem that no one owns (but Google still benefits from) the profits go to, well, no one.
“It’s innovate the next thing or die,” Kessler said, acutely aware that IP innovation is fleeting. “Where die means low margins from commodity products.”
Just about the same time I was reading Andy’s email, I noticed a tweet flagging Farhad Manjoo’s CES column which talks about how the category of Consumer Electronics got thinner and thinner over the years under the assault of ever more featured phones. The common theme coming out of CES this year was, everything is tech.
But when everything is tech, nothing is really tech. And the victory of the smartphone platform has meant that problem-solving IP is now migrating from hardware into discrete software in the form of apps.
Manjoo describes this migration through Sonos, the home music system that turned out to have a far-sighted CEO.
“Today, what every customer expects is for their device to be a platform,” said John MacFarlane, the chief executive of the connected-speaker company Sonos, referring to a design practice in which the machine’s intelligence and user interface are built out of flexible software rather than baked into the hardware — thus enabling future improvements through updates.
Kessler agrees with MacFarlane. “An app maker, say Dropbox, has a similar advantage” to chip designers who print their ideas on silicon, he wrote in a later email. “It is a new paradigm that rhymes. Someone writes some clever code to make it do something unique combining storage and communications. The parlor trick is building a multi-billion business on top of commodity servers that someone else lets you rent, just like those semiconductor fabs.”
Instead of the Taiwanese building factories, we have Amazon, Google, or even IBM supplying commodity server power.
The problem is that innovation only lasts so long. “Pick any company name and you can fit it into this IP etched in servers strategy,” Andy wrote. “Uber is more of an interface to the real world. It is clever code and yes, a middleman between riders and drivers, where dispatchers used to sit.”
“Yes, the lead is precarious,” Kessler went on, reminding me that all innovation is fleeting. “It always is. Can Uber and Airbnb be displaced? Of course, easier communication with drivers, better selection of apartments/homes.”
“Their success was there for the asking,” he wrote. “Now it gets harder as others add new features (carpools, etc.) It is still IP being printed by others (Apple and Samsung do the heavy lifting; so do the limo/taxi drivers) but the value of the IP becomes questionable. It comes down to branding and that type of stuff and less about IP printing.”