Whine whine. I have an idea. Why not make a product to sell to consumers, instead of making shares to sell to the public?

Not really the point when discussing the investor market. If companies don't have quick exits, the typical venture investor can't make their fund operate.

That said, I do believe building a company that creates value on the basis of selling a profitable product is all that matters.

Well said.

How about downscaling the size. Sites like Twitter could be done by one person and a bit for subcontracting. Sites like Facebook by a team of 10 to 15.

Too much money and too much megabuck sponsored hype. Simple micro startups are still left out even though they are more than viable.

Sorry but that statement is crap. I'm sure if Twitter could be done by one person, they could. These companies have a LOT more going on than just banging out code.

I think the whole huge employee numbers are just a way for those big companies to justify to their investors their high valuations. "Oh yeah we have 700 employees, so we are a huge company, so 500 mil in VC funding is really not a lot for a huge company like us"

There seems to be an almost pathological bias against profitability with a lot of startups.

So I heard that some think that is because for a long while the R/D effort was outsourced to other companies, away from larger ones. I dunno if that is the case but in the company I worked for got acquired and did some pretty serious R/D. It could have been profitable but acquisition was obviously the easier way out so the investors took it.

I'm sure something like Twitter is useful but only to a large company, perhaps? I dunno.

That's not really it, if we're going to speak in generalizations it would be more fitting to talk about the investors that favor high risk/high reward growth patterns.

A site like Facebook would be impossible with just 10-15 people. Designing, implementing, and managing something on that scale is an insane amount of work. It probably takes a dozen ops/monitoring people just to keep the thing running.

Only with standard corporate solutions.

a Website (httpd etc) + Web Application server (CMS/CMF) + Application Container (J2EE and all that) + RDBMS (Oracle, MySQL...)

And multiply that to development, QA, and production environments.

Yes, that way you need millions just in licenses and hosting to run even a lol-site and hundreds of employees to keep the beast working.

Good! now startups will be forced to restructure their business and actually find a way to make real revenue.