It strikes me as a bad plan to depend on 400% growth that's sustainable over a decade. You're right, people are horrible at taking growth variables into account. They usually overestimate them.

Facebook may well achieve the numbers outlined above. But let's not kid ourselves. People were doing the same back of the envelope calculations for MySpace when it first started to get big. I see nothing wrong with a judicious use of common sense and caution in this case.

Things are changing very fast in this industry. So even 10 years and 2 billion users are optimistic numbers. And if you take 2 billion users, most of these users are from very poor countries, it is extremely hard to monetize these users (almost orders of magnitudes harder than users from rich countries.). That said, I don't think that Facebook is overvalued by orders of magnitudes, and the 37signals article did not make much sense to me.

Which means that if Facebook's valuation rises over the next decade there will be an even bigger gap between their user base * revenue per user than there is today. Either that or their valuation will have to remain flat for the next decade.

Does anybody else see a problem in that?

A company's valuation can certainly take into account growth potential, but when it assumes massive growth and leaves no margin bad things happen. Risk is much higher since any faltering of the company would need to be adjusted for in stock price almost instantly. If a normal company sees slower than expected growth the stock may fall a little. If an overenthusiatically future valued company sees slower than expected growth the stock may tumble by huge margins (because it's tied to the company's size 10 years from now, not today). Worse yet, if there's no margin remaining in stock price it'll be forced to remain stagnant for a very long time. This is bad for investors and bad for the company (because the use of stock as employee compensation is less potent).

2 billion active users is almost 1/3 of the worlds population. That seems like a strech.

Also you can't link the 33billion to expected. 33billion is their "valuation" now with 500M active users. There's a certain expectation that your valuation should go up as users increase.

Personally I think the per-user deal is kind of a red-herring. Revenues and Profits are the only things that really matter. Users are just a one (good) dataset for estimating growth - but hardly the only one.

"rough estimates of 2 billion facebook users"

Right. Because the Chinese and Indian Peasantry is definitely going to be on FB.

Hey, lay off man! If Chicago doesn't have to teach multiplication, New York doesn't have to teach grammar, alright?

Read the original post, Joel didn't make that mistake.

Also a dime per user/year is more than it seems when you can only collect it through indirect means (ads, charging other companies for using your platform). Remember their ~500mio user figure also includes the overwhelming portion of users who generate no more than a handful of page impressions per month.

It's not like facebook could invent a yearly membership fee or start selling t-shirts or such...

Well, they could, but I doubt more than ~20% of their audience would actually pay.

You don't think so, but they probably have.

I always used to think I was a 'leech' user of TripAdvisor, until I worked there.

Have you viewed more than 300 pages on the facebook.com domain in your lifetime?

If so, they've gotten a dime from you.

3 ads per page, each ad selling at a $0.10 CPM.

Never use tech-savvy people as representative examples when evaluating advertising.

Otherwise, even google would be out of business.

you must be kidding, right? If their CPM were $30-40 they'd be making many billions per year, not (maybe) 1 billion.