I fell for the same thing myself a while ago; the rules actually are different when you're selling something with zero marginal cost.
As long as you believe LinkedIn can increase their revenue without adding to costs, their current P/E is kind of irrelevant.
This is why investors basically hate consulting companies (since revenue/earnings scale linearly -- in some ways, sub-linearly, since you have to hire more layers of manager), and love products selling zero unit cost products or services directly online.
Something like LinkedIn probably does have very little sales cost related to selling a higher end account, and little marginal cost to actually providing it.
Plus, LinkedIn is a network, so the more users use it, the greater value they each get from using it, and so the higher the revenue per user possible.
Things to watch out for are huge revenue/low earnings when you're near max revenue (so, a product which can only sell to a market of maybe 10000 people worldwide, 9500 of whom already use it, and who are paying a price which is only break-even for you -- adding the extra 1k users won't really get you to good margins), or services, like Groupon, which have huge sales costs to produce incremental dollars of revenue (even if the actual product is basically free to provide).
the rules actually are different when you're selling something with zero marginal cost.
I don't think there is any business that sells something with zero marginal cost. Even if LinkedIn grows significantly without any marketing and sales activities, they still must deal with:
More users == more servers
More users == more support staff
and on and on.
Use Google and Facebook as classic examples.
And if you aren't getting more users, then you are ramping up marketing efforts to do so.
While LinkedIn is a definite success story, I'm not buying it that they are running a zero marginal cost business. I don't believe that anyone is.
Yeah, but it's a step function. You can service, say, 100k users on a couple of servers. User # 100,101 means you have to buy a new server. Technically that user costs more, but the next 99,999 cost 0. (Leaving out rules about depreciation, amortization etc - I'm just talking pure cash flow.)
As long as you believe LinkedIn can increase their revenue without adding to costs, their current P/E is kind of irrelevant.
This is why investors basically hate consulting companies (since revenue/earnings scale linearly -- in some ways, sub-linearly, since you have to hire more layers of manager), and love products selling zero unit cost products or services directly online.
Something like LinkedIn probably does have very little sales cost related to selling a higher end account, and little marginal cost to actually providing it.
Plus, LinkedIn is a network, so the more users use it, the greater value they each get from using it, and so the higher the revenue per user possible.
Things to watch out for are huge revenue/low earnings when you're near max revenue (so, a product which can only sell to a market of maybe 10000 people worldwide, 9500 of whom already use it, and who are paying a price which is only break-even for you -- adding the extra 1k users won't really get you to good margins), or services, like Groupon, which have huge sales costs to produce incremental dollars of revenue (even if the actual product is basically free to provide).