I read an interesting pair of posts about Web startups competing with big companies like Google and Microsoft over the weekend. The first was a post by Bill Burnham entitled Search Applications: Search Startups Are Dead, Long Live Search Startups where he writes 

In response to a question about the prospects for the myriad of search start-ups looking for funding Peter basically said, and I am paraphrasing somewhat, that search start-ups, in the vein of Google, Yahoo Ask, etc. are dead.  Not because search isn’t a great place to be or because they can’t create innovative technologies, but because the investment required to build and operate an Internet-scale, high performance crawling, indexing, and query serving farm were now so great that only the largest Internet companies had a chance of competing.

Priced Out of the Market
While the comment might strike some as self-serving, the fact of the matter is that it is true.  Any start-up trying to displace Google, Yahoo, or even MSN or Ask (or for that matter any VC trying to fund them) should just get in their car (or hop on a plane) and go look at Google’s new server farms at The Dales in Oregon.  And if that doesn’t convince them they should head up the Columbia river a bit and check out Microsoft and Yahoo’s digs.   The costs to compete in core search , are now simply to high.

Bill then goes on to argue that the opportunity may lay in building applications on top of the APIs provided by the big software companies such as the Alexa Web Search Platform instead of trying to compete head to head with companies that have already invested hundreds of millions of dollars in building out their search infrastructure. In a post in response to Bill Burnham, Tim O'Reilly agrees that "we're entering the platform phase of Web 2.0, in which first generation applications are going to turn into platforms".

Dave Winer comes to the same realization but with a different conclusion in his post Someday search will be old too where he writes

Many years ago, when the Internet was still the domain of geeks, researchers and college students, the smart folks often said that the opportunities for new software companies were over, it simply required too much scale to compete in an industry dominated by Lotus, Microsoft and Ashton-Tate. Now it's clear how ridiculous that was, even though it was correct. The next layer comes on not by building on the old layer (a trick, the guy you're building on will eat your lunch), or re-doing what they did (what the naysayers correctly say you can't do), but by starting from a different place and building something new, and so different that the old guys don't understand it and don't feel threatened by it.

Dave seems to be disagreeing with Bill Burnham and Tim O'Reilly that building on the platforms provided by the big software companies is a good idea because the companies can turn around and compete with you. Although I can see Dave's point, simply having the same base platform doesn't mean you can build the same application. We all have access to Linux, Apache, MySQL and PHP but how many people have or can build something like Flickr? You should definitely examine the risks when building on any platform but eventually you have to be ready to face off against a competitor that may have access to the same or a similar platform as you. You're development platform cannot be your only differentiator.

I do agree more with the spirit of what Dave Winer is recommending than what Tim O'Reilly  & Bill Burnham are. If I ever founded a startup, it would take advantage of the knowledge I have of what big companies like Microsoft are good at and what they aren't. For example, YouTube would never have come out of Microsoft because the company would have been too scared of lawsuits. Of course, now that YouTube has money they are getting hit up from every angle so that fear does make sense for a big company but not a small startup especially if the exit strategy is to flip it.