All around, technology companies are no longer content to provide the machinery for consumer companies — they are serving the customers themselves.
Tesla, the Californian car company started with venture capital, gained the right to sell its electric cars directly to consumers in New Jersey after a battle in which Elon Musk, its founder, compared auto dealers to the Mafia. Spotify, the music streaming service, faces a protest from Universal Music for allowing 45 million of its 60 million customers to listen to music free.
Facebook is launching a peer-to-peer payments system on its Messenger application, while negotiating with news publishers to host their stories directly, rather than providing links through its news feed. The New York Times may become a supplier of news to a technology-led distributor.
The shift could be as significant for the technology industry as that of 1981, when IBM allowed Microsoft to provide the MS-DOS operating system for its new personal computers. It turned out to be a profitable power grab by Microsoft, placing software at heart of the computing industry, while hardware became a commodity.
The new companies, such as Tesla, Netflix, and Warby Parker, the eyewear retailer, are dubbed “full-stack start-ups” by Andreessen Horowitz, the venture capital firm. Instead of building the plumbing for others, like Oracle and SAP in corporate software, they sell products and services directly, exploiting their efficiency and scale.
There is a historical symmetry in this. Microsoft’s oldest rival, the company that refused to confine itself to software, is Apple. Apple, now the world’s most valuable company, is hailed by Chris Dixon, a partner of Andreessen Horowitz, as “the canonical full-stack technology company”.
It designs its own software and hardware, including computer chips, and operates its own network of retail stores.
For the most part, though, “full stack” is a misleading term, taken from the manner in which modern software is built in modules that mesh with each other. This makes it sound like vertical integration of the kind pioneered by Henry Ford, whose company once not only built cars but owned rubber plantations, coal mines and railways — much of its network of supply.
A few industries are still vertically integrated: some oil companies, for example, span upstream exploration and drilling, through downstream pipelines and refineries, to retail petrol stations. The trend, however, has been away from corporate monoliths and towards outsourcing to networks of suppliers.
Even Apple contracts out the manufacturing of iPhones. Full stack often means “top of the stack” (or, confusingly, downstream in traditional industries), handling the customer relationship.
Uber, one of Dixon’s prime examples, is not a very integrated company. It does not directly employ drivers, although it is being challenged over the issue in US courts — it is a technology platform.
But Dixon is right that something notable is under way. Microsoft’s success came from combining technology with business innovation; it took over the high-margin segment of the computer industry, which was then in the middle. Neither hardware manufacture nor retailing was as profitable as software, largely because Microsoft held a near monopoly while the other sectors both remained fragmented.
The lure for full-stack start-ups is that the structure of many industries, not just the technology industry, can be reversed by the internet. The middle is often divided among suppliers, whether taxi drivers, news publishers or music labels, while a single network can reach millions of customers.
The technology need not be very sophisticated. Although Uber has clever dispatch software, its largest advantage is that it works through an app on mobile phones, rather than needing people to sit in smoke-filled offices, fielding phone calls. It is bigger, faster and more efficient than thousands of little cab companies.
Once it occupies the high ground of customer relationships, a company has two options. It can exploit its power to push suppliers into offering goods and services through its network on the terms it wants, as Spotify, Facebook and Uber are doing.
Alternatively, it can turn into a true full-stack start-up, acquiring or contracting with suppliers to design, make and sell products. Two neat examples although both are small, are Warby Parker and Harry’s, a US start-up that sells men’s razors and shaving cream by monthly subscription, like a Spotify for chins. It makes the razor blades in its own factory in Germany.
The striking thing about such start-ups is that although they are backed by venture capital and use software and the internet to their advantage, they are not really technology companies. Venture funds such as Andreessen Horowitz are investing in new business models, often with hundreds of millions of dollars to allow them to scale up rapidly, rather than new science.
This contrasts with Andreessen Horowitz’s $20 million (Dh investment in Improbable, a virtual reality start-up. It also evokes the complaint of Peter Thiel, the entrepreneur, that Silicon Valley prefers “incrementalist investment” to solving hard scientific problems. But if it works, why fight it?
Financial Times
