Turns out I was right - much of the Instagram and SnapChat usage is coming out of the pool of attention that previously belonged to Facebook - but the cycle time seems to be ~10 years, and getting longer with each generation of social software. If the history of mass culture & mass consumerism repeats itself, eventually the cycle time will be greater than a generation - forms of social interaction for the baby boomers stayed roughly constant for their whole lifetime. At that point it's meaningless to talk about shifts to the next big thing, because we'll all be dead before they happen.

http://www.wired.com/2007/10/facebook-future/

So if facebook is the benchmark, it doesn't support your valuation of $200B for snapchat with 50M in revenue.

* GM * Ford * Honda

Combined. That is ridiculous and just highlights how big a bubble is brewing in SV.

We value a couple hundred million fickle users more than hundreds of thousands of actual employees, physical factories, and entire supply chain industries.

The valuation (market cap) of a company is a reflection of the value of the assets that the company owns. That includes IP, trademarks, brand assets (somewhat perversely, this also includes user habits), business relationships & contracts, any resellable physical goods or cash that the company owns, and all future cash flows that will accrue to the company. Basically, the stuff that an employee, supplier, or customer could not take with them if they chose to do business with someone else.

Now, there's also the concept of economic value, which is the value that customers of a company (or industry) are willing to spend on its product. This can be reasonably proxied by the total revenue of the industry - for your GM + Ford + Honda example, that's about $150B each, or $450B (and leaves out the 800 lb. gorilla, Toyota). That's a lot larger than the total $60-80B market for all brand advertising.

But building cars has a long & large value chain. GM + Ford + Honda can't do it all on their own. They each employ roughly 200,000 people, many of whom are relatively highly paid thanks to union agreements. They have literally thousands of suppliers that they buy parts from. These suppliers themselves need to buy parts and supplies from labor-intensive industries like mining. As a result, the company captures comparatively little of the profit from all of this work. Facebook, for example, makes about $6B in net income off $20B in revenue, while GM makes $10B in profit off $150B in revenue. This also implies that further growth will benefit companies like Facebook or SnapChat much more than it would companies like GM: if margins stay the same, Facebook gets to $10B in profits with $30B in revenue, while GM would have to get to $300B in revenue (probably not happening) to get to $20B in profits.

It's a plus, from a shareholder value perspective, that SnapChat only employs 200 or so people, and their largest expense is paying Google for AppEngine bills. It means that whatever value SnapChat creates, the shareholders will capture almost all of it. Indeed, based on how often car commercials appear on TV, it's likely that a good portion of GM & Ford's revenue will become SnapChat's profit, just like how a good portion of SnapChat's revenue becomes Google's profit.

I completely understand that from a shareholder perspective companies like facebook and snapchat are huge wins.

However my argument is that as a society these companies have very little value. I believe that the current wave of tech startups are the equivalent of shareholder junk food. I nice relatively immediate return, and no real long term value.

Was valued at 25B in 2000, ended up getting acquired by Yahoo for 235M. In terms of stocks - from $241 down to $1.63

That easy replication of the product is exactly what I'm talking about. Search engine is actually a much more sophisticated setup and even in that space multiple companies went belly up.