"I don't know how you can reasonably value a company that has only existed for a year or two at that level."
That is a fair question.
Brad Templeton once observed that if he sent a joke to everyone on the Internet with the caveat "If you chuckled, click here to send me a penny." and only 1% of the population found it funny. That would be a $100,000 payday with a billion readers where 99% of the people didn't even think it was funny.
The secret truth that Brad captured (way before it was well known too) is that the rate at which a pure information company can earn money is essentially unlimited. Not a lot of people could imagine Brad's scenario, they got bogged down in the details like "gee processing the payment would cost more than a penny" or "how do you force them to pay a penny" or whatever. But now Google comes along and capitalizes on exactly that idea, except that rather than have the users click on something if they think its funny, they have someone click on something if they think it may be an answer to their question, and rather than processing them one penny at a time, Google gathers those clicks and just does an accounting transaction which becomes an actual payment once a month for the aggregate transaction. But they don't make anything, rather they convert an action into money, and they do that at web scale. No limit to how much money they can generate. So from an investment value, they have a very very large rate of return. The actual limit is how much money advertisers are willing to inject into the system.
So if you have a company whose revenue is only limited by the amount of money its 'customers' are willing to inject, and you can make a credible case for the fact that customers will be willing to inject $100M+ per year into the system, then you can make the argument that the company is 'worth' a billion dollars. (or multiple billions of dollars).
So while I agree with the unspoken criticism that some recent valuations seem high based on available information, if those valuations are coming from folks who have done their due diligence then I expect they have some line of reasoning that justifies that value to their own satisfaction. In the 'individual investor' bubble of the late 90's it was clear that individual investors were doing no diligence at all.
There's a flaw in the argument that most people miss, namely that since it's so easy to create and distribute information everybody does it, making it so much harder to create a viral effect.
To extend your example: If just 1% of the population sent a Brad Templeton joke to everyone on the Internet you would receive 10 million emails with a funny joke. If you chuckled at 1% of them you'd send a penny to 100.000 people. Not a very likely scenario.
Since none of us receive 10 million joke emails this means that things aren't as simple as they appear. The fact is that since so much information is being created the bar for virality is extremely high, and it's rising proportionally to how easy it is to create information. The chances of your funny email reaching millions of users are as high as your chances of winning the lottery. The resemblance to slick entrepreneurs pitching their wares with the clasical "if only 1% of the market uses our product we'll make billions". That's just not how it works.
Note that I'm not necessarily disagreeing with your thesis that once a startup has reached virality it can be worth a substantial amount of money, I'm simply stating that reaching that point is as hard as ever.
"There's a flaw in the argument that most people miss, namely that since it's so easy to create and distribute information everybody does it, making it so much harder to create a viral effect."
An excellent observation. And you have to consider that not all information has the same value and that its value varies over time.
Brad's scheme does have the flaw you describe (and it was another of the rat holes people would fall into, "Well Gee, once he sends out a joke what is to stop anyone who got from sending the same joke out to everyone but making them the recipient of the penny?")
The mechanism is valid, but, as you observe, won't work with easily reproduced information.
Google sells information that is both 'rare' and temporally relevant. They sell 'a person is looking for X right now' and advertisers buy access to that information with their AdWords bids. The nice thing about this information is that its very valuable to certain people (people who think they have a product for people looking for X) when it gets created and a couple of seconds after it gets created it is worth 100x less (maybe even 100,000x less, basically only as trending data).
That's actually very interesting - what you're proposing is that to create a viral effect you don't send the same joke to everyone but tailor it algorithmically to suit each recipient based on his history or some other parameter.
The trick, of course, is being able to gather and mine the data needed to do so.
It seems that facebook, Google, and a lot of other successful companies are following this model.
I was at a conference last month and was talking to a guy from Elsevier (a major scientific journal publisher.) I was telling him how getting a piece of content for 5 cents that makes 5 cents of revenue was "a pretty good business" and he was shaking his head as if I was insane.
Elsevier's business model is to take content for free -- indeed, to persuade people to beg you to accept it -- and then to sell it back to the same people for thousands of dollars.
Your extension changes the analogy. One person clicking on something relevant to their search is different than one person clicking on X relevant searches.
The "chances" of an e-mail reaching X number of people entirely depends on the e-mail list you've compiled. The company I work for sends an e-mail every week and it reaches over a million people every single time.
Even if you make a 1% conversion on that list each time, you are showing continual and repeatable return, which some call a "business model". A lot of us get caught up in the fact that this model has nothing to do with the actual product.
That is a fair question.
Brad Templeton once observed that if he sent a joke to everyone on the Internet with the caveat "If you chuckled, click here to send me a penny." and only 1% of the population found it funny. That would be a $100,000 payday with a billion readers where 99% of the people didn't even think it was funny.
The secret truth that Brad captured (way before it was well known too) is that the rate at which a pure information company can earn money is essentially unlimited. Not a lot of people could imagine Brad's scenario, they got bogged down in the details like "gee processing the payment would cost more than a penny" or "how do you force them to pay a penny" or whatever. But now Google comes along and capitalizes on exactly that idea, except that rather than have the users click on something if they think its funny, they have someone click on something if they think it may be an answer to their question, and rather than processing them one penny at a time, Google gathers those clicks and just does an accounting transaction which becomes an actual payment once a month for the aggregate transaction. But they don't make anything, rather they convert an action into money, and they do that at web scale. No limit to how much money they can generate. So from an investment value, they have a very very large rate of return. The actual limit is how much money advertisers are willing to inject into the system.
So if you have a company whose revenue is only limited by the amount of money its 'customers' are willing to inject, and you can make a credible case for the fact that customers will be willing to inject $100M+ per year into the system, then you can make the argument that the company is 'worth' a billion dollars. (or multiple billions of dollars).
So while I agree with the unspoken criticism that some recent valuations seem high based on available information, if those valuations are coming from folks who have done their due diligence then I expect they have some line of reasoning that justifies that value to their own satisfaction. In the 'individual investor' bubble of the late 90's it was clear that individual investors were doing no diligence at all.