Thinking of startups as building blocks of a larger corporation

I was wondering has anyone encountered posts that discuss the topic of using many startups to create larger entities. Instead of adding new departments or divisions within a startup, use other startups to solve the problem.

For example, Microsoft would be made up of 1,000's of smaller companies that approach their markets using a "bottom-up" approach (from computer science). That way you can keep rapid innovation but have the scale of massive corporations.

This blog post sums it up pretty well: http://forceforthefuture.com/2009/12/the-lego-model/

Larger corporations have economy of scale, have more clout with investors, vendors, partners, etc. They are not after innovations either. Their goal is to milk money off their market. So it doesn't make sense for any corp honcho to create startups instead of growing his corp.

One of the facts about companies of any size is that most of them fail. They're hard to get off the ground and it's hard to actually get more money than you spend, especially between having no employees and having them.

So with 1000 startups all working together -- assuming the logistics of that communication isn't already insurmountable -- you can assume that hundreds of them are dying. Not good for those depending on them.

And to keep up the quantity you'd need new startups constantly joining the fold, and of course new products are unsteady and new company directions unsure.

Finally, it's not clear that there are 1000 different things that need doing in a company. And for those things where it makes sense to get other folks to do it, why can't you just do it now anyway?

So the real question is: What's the advantage of having 1000 separate moving parts?

How it is different from the fact that there is 1000 small companies in one place.
1000 warriors didn't make army if they are not ruled by one.

I Would say that this is what Apple does with it's App Store. Of course the individual developers are not part of Apple, but they do in some way share a brand (Apps Store) and infrastructure (distribution through Apps Store and Software Development Kits). I strongly believe in creating a platform for other companies to flourish. Microsoft is essentially doing this as well. Stock exchanges are another example.

I guess in the older days, such building block entities were called "research". Research was supposed to come up with the ideas to forge into new markets, expand product offerings, and make breakthroughs that could be carried forward with a company that has resources. Many died, some survived, and even fewer prospered. But there were successes.

Now, with today's 10-k 10-q focus, no one wants to invest in research & M&A has taken over the role of innovation. If companies did a better job of commercializing their research (xerox parc comes to mind) perhaps we'd be in a different world.