Facebook - as seen in The Social Network, with Justin Timberlake and Jesse Eisenberg - is seen by many as a proven model for an internet business Photograph: Rex Features
You don't just walk into the San Francisco offices of Uber, the "ride-sharing" company. Get past the security guards and into the lift, and you emerge in the company's dark-decor reception, where another guard, with telltale earpiece, is ready to check would-be visitors for signs they might want to cause trouble.
That is the problem with disrupting people's business model, in this case taxi companies the world over: some of them might want to come and disrupt you.
Uber has had its share of protesters. But as the practised eyes of the guards indicate, it's part of life in the technology fast lane. According to co-founder Travis Kalanick, Uber is valued at about $17bn, following another round of funding in June. This, for a company some feel is just trying to muscle into the taxi business by combining satnav systems, spare drivers and a smartphone app, feels to some like a bubble valuation.
But what constitutes a technology bubble? If it's gigantic valuations, you don't have to look far. When venture capital company Andreessen Horowitz put $85m last week into online publisher BuzzFeed in exchange for 10% of its equity, that implied BuzzFeed was worth $850m, more than the 137-year-old Washington Post. Five-year-old Uber raised a little over $1.2bn from venture capitalist for about 5% of its equity, to give it that $17bn valuation.
And if Snapchat, the vanishing-picture service, could turn down a $3bn offer from Facebook in November 2013 (according to the Wall Street Journal) and was in talks last month with China's Alibaba over funding that could value it at $10bn, isn't that evidence of a bubble?
Chris Dixon, a partner at Andreessen Horowitz, defended BuzzFeed's funding as an example of a new-media company being built on an emerging technology – past examples include Time Inc, CBS and Viacom. "BuzzFeed is a media company in the same sense that Tesla is a car company, Uber is a taxi company, or Netflix is a streaming movie company," he wrote – meaning that they're far bigger than that. "We believe we're in the 'deployment' phase of the internet. The foundation has been laid."
But are the foundations solid enough to stop sky-high valuations crashing down? The latest debate over whether we're in a "tech bubble" has been raging in Silicon Valley and beyond since at least late 2010, when venture capitalist John Doerr suggested we were entering a "third wave" of innovation, focused on smartphones and social networking. To the suggestion that valuations were – even then – "overheated", Doerr replied: "I prefer to think of these bubbles as booms. I think booms are good. Booms lead to overinvestment, to full employment. Booms lead to lots of innovation."
Overinvestment also implies bust, though. The question of which moment marks the peak has been raised repeatedly: at the time of corporate social network LinkedIn's flotation in May 2011, with Facebook's purchase of mobile picture network Instagram for $1bn in April 2012, Facebook's flotation in May 2012, Twitter's flotation in November 2013, and so on. Last week's cash injection into BuzzFeed was just the latest trigger for headshaking on both sides.
Writing for Business Insider last November, co-founder Jim Edwards laid out a lengthy list of reasons of why tech is overvalued: low interest rates are driving a wave of funding into companies as "virtually any other kind of investment [than cash in the bank] is likely to pay more", and there is a shortage of skilled staff. Worst of all, he suggested, was that companies with "broken" business models (such as design retailer fab.com, which laid off hundreds of staff, yet won more funding), or without meaningful revenue (such as image-pinning service Pinterest), or with no revenue at all (Snapchat) were all highly valued.
The key dispute is whether large user numbers translate into large profits. Veterans of the first dotcom bubble recall how "eyeballs" were, for a time, the most important metric: how many people were looking at your site?
Those making the investments – and many on the sidelines – are certain that this time things are different. August 1998 saw the launch ofpets.com, selling pet food and products online ("because pets can't drive!"): it went from flotation to liquidation in less than nine months. What's often overlooked is that it had more than half a million customers at its demise – remarkable in the pre-broadband era. But then those customers were getting a bargain: the company was selling goods for about a third of what it had paid for them, an unsustainable business model in any circumstances.
The difference from the 1990s, and even from the mini tech boom of the mid-2000s that halted with the credit crunch in 2007, is that there are now multiple proven internet-only business models – from Google's search-based advertising, to Facebook and Twitter's socially based advertising, to LinkedIn's "freemium" model, which charges users for better access.
One area under close focus has been the "sharing economy" – which tends not to be about sharing at all, but about making money out of underused things – be they space, time or products.
Uber and room rental company Airbnb like to be described as part of this – Uber prefers to be called a "ride-sharing company" (and insists it is not a "taxi company"). The "sharing" tends to be mediated (for money) by the company. Thus Uber doesn't employ cab drivers or (obviously) customers, but links the two, and sets the price for a journey – including its "surge pricing", when high demand can increase a fare fivefold.
But it contains huge value, according to a new study from PricewaterhouseCoopers (PwC). "In some ways, the sharing economy is a throwback to the pre-industrial age, when village communities had to share resources to survive," said John Hawksworth, PwC's chief economist. "They built up trust through repeated interactions with people they had known all their lives. Digital communications allow sharing across a global village … with trust established through electronic peer reviews."
Uber's valuation has been queried, but some see it as justified – its model of uniting buyers and sellers separated by short distances has wide applicability.
Can Snapchat make money? Certainly, by selling in-app purchases, or "stickers" (virtual elements to attach to a message or picture) – a very profitable model for the Line messaging app in south-east Asia. Pinterest is exploring "promoted pins". Twitter is ramping up its advertising. And don't forget that when Google started, its founders had no aim except to offer a really good search engine – and no clear business plan.
Dixon is certain this isn't a bubble; this is an embedding of technology companies from one sector into everything. "Tech is now spreading through every industry and every part of the world," he said. "The most interesting tech companies aren't trying to sell software to other companies. They are trying to reshape industries from top to bottom."