I mean, imagine some big company opening a BigCompanyName-Seed division, where they do something like Y combinator is doing, but instead of giving a small amount of money for 6% they get 50% of the company for a bigger investment (at least in perspective, so if future investments are needed, there is no dilution up to a given sum).
But in the process of course the company doing this stuff should be smart enough to don't force any strict rule like technology to use, type of authentication, domain name, and so forth. Must be a startup with its own life. Just they can select things that are interesting in general, or are interesting for the general goals of the big company.
I'm not sure why this model could not work.
In some cases, the acquirer can win simply by waiting for a company to prove itself in the marketplace. They pay a small premium to instantly become the market leader relative to the risk-adjusted cost they'd have to pay to ensure an internal effort resulted in the same market position.
In other cases, the acquirer knows it's going to dominate the market once it enters with an acquisition. It wants to know whether a product is going to succeed, and what features customers are going to value most highly. So it can sit back, watch the market shake out, and then buy the best also-ran at a major discount to the valuation of the market leader.
These are decision factors that aren't open to companies that innovate internally. That doesn't make internal innovation bad; it just argues in some cases for M&A instead.
This failed utterly.
Why? Once a company is beyond a certain size, employees maximize their expected value by spending time seeking internal resources. Budget, headcount, etc. If the lab is still part of the parent company, these games are still played.
You also have companies like Hulu, that exist at the pleasure of their "investors." This might work a little better, but Hulu can't act fully in self-interest.
Perhaps the best thing I can think of is a spinoff with initial funding, control of its own board, and an irrevocable IP grant. This would potentially be a huge head start. But then what happens if a competitor acquires your spinoff (for the IP grant) right away?
I guess all this seems more fraught with risk than the simple buy a random startup model.
Couple good books come to mind. The best of which is probably The Innovator's Dilemma, for how new ideas get killed by companies.
There are a lot of reasons what seems obvious won't actually work -- even though many companies still keep kicking that can down the road. This whole area is like a siren's song both in the corporate world and in government. Everybody is building "centers of excellence" and "knowledge incubators", and whatever else buzzwordy thing they can come up with to spend money. But the record of such efforts is appalling. Just for one example, take a look at Microsoft -- tens of billions in cash, tens of thousands of brilliant people, and they're lucky to get a release of windows out in time. Too many reasons to go into in this short of a space, and of course there are a lot of true believers that would argue with my conclusion. Suffice it to say that smart companies have figured out it's easier just to write a check.
Would be interesting if anyone has any more insider info about it to share.
I think one reason is that most seed investments don't need to be that large, and large investments can actually be detrimental to the progress of a startup.
Combine that with raw talent acquisition, and a lot of startups begin to look like highly-rewarded spec R&D work for big companies. Picking up that kind of talent, audience, and potential? It's easily worth it to BigCorp, even if there's no income at all.
(I hope the tone sounds okay on that. There's nothing wrong with gaining a million users and flipping your company to Google -- assuming that's what you want to do, of course)
Of course, there's a ton of startups that don't fit that model. But many do -- at least many in the valley.