In an item that John Battelle of Searchblog dubs "VCs versus The Platforms," both Google and Yahoo are reportedly setting up early stage venture arms to invest in companies that they presumably would want to acquire if all goes well. 
Paul Graham has also written an excellent post on the topic called, "The Venture Capital Squeeze" that addresses how the changing dynamics of the startup game so totally reduce the cost startups face in getting a product to market (and properly marketing the business post launch) that for large categories of companies that can run in a capital efficient manner, the traditional VC cost model is dis-advantageous for the entrepreneur.
On the flip side, I have written a bunch of posts where I argue that many, many of these new fangled Web 2.0 companies are $25M outcomes (if they execute), potentially a great outcome for entrepreneurs but horrible as venture investments.
Enter the platform players -- think Yahoo, Google, eBay, Amazon and Microsoft -- who can play the Cisco hand by agressively investing in early stage deals that may only need $1-3M to launch and be run profitably, especially if the prospect of a distribution deal is in hand.
Those who remember how Cisco successfully moved into segment after segment by using M&A as R&D should be able to see a perfect storm emerging whereby lots of product extension plays exist to build out the platform providers' proprietary-ness in a non-disruptive fashion, make entrepreneurs a lot of money, and do so in a matter that avoids the expensive acquisition after several rounds of venture investment.
This is great for the platform provider, great for entrepreneurs and great for customers since it suggests a model for traditionally slow-moving platform providers to get their running legs and continously inject entrepreneurial DNA into their culture, products and services.
I also think that this is great for VCs as a whole because it will force them to focus on their knitting; namely, building large standalone businesses focused on securing discrete billion dollar market opportunities. Too much of the business right now is focused on product line extensions NOT large standalone companies.
The only wrinkle that remains to be seen is that, more often than not, corporations prove to be lousy venture investors, as much as anything because they have nebulous goals, fuzzy underwriting criteria and arms-length synergy plans between the upstart and the business unit where alignment is greatest.
The suggestion here is that this needs to be one of those "begin with the end in mind" types of deals. When actively investing in a category, do the bake off with startups in the space and set the understanding that the deal is a likely acquisition with the multiple of valuation (upon acquisition) based upon a tiered set of milestones.
Milestones can include agreed upon product functionality, integration-levels with the mother ship, number of users of the parent service that embrace, paying customers, frequency of usage (and related emotial investment metrics) and time-based and product lifecycle related earnouts.
Pretty cool, though. The platform play now includes rich, free APIs, cool widgets to integrate with, a large user base to plug into. And cash!
