It's become so easy and inexpensive to launch a software or Internet-based startup that the current "leading edge" of thought is that startups should "fail fast" (make mistakes early) and then "pivot" (adopt a different strategy, or even develop a different product or service) until they land on a viable market opportunity. We've lost sight of some fundamental first principles that startup teams should think about before they write a line of code. These are first principles in the vein of "The Art of War"--how likely are you to survive engagements with customers and competitors.
For example, direct, head-on competition with entrenched competitors is likely to result in failure. If you're building a new search engine, you're likely to fail (see Cuil and endless other companies) unless you've got a parent company that makes more money than it knows what to do with (see Bing.) The best outcome in this case is that you'll develop some unique and interesting technology, and larger competitors will find it cheaper and easier to buy it from you than to build it themselves.
If you're planning to go after much bigger competitors, misdirect them and keep your mouth shut about your true intentions until you're too big to kill. Netscape was growing incredibly rapidly as a browser company, but Microsoft was largely ignoring it. However, as soon as Netscape announced that it was positioning its browser for running applications on any operating system, Microsoft saw it as an existential threat and did everything it could to destroy the company. Since Netscape still had tiny revenues, Microsoft could "cut off its air supply" by giving its browser and Internet servers away. Had Netscape kept quiet about its intention to turn its browsers into an application platform until it was big enough to withstand attacks from Microsoft, it would have survived. (AOL, which acquired Netscape, has abandoned all use of Netscape's brand name and products, and moved out of the last of Netscape's buildings in Silicon Valley in late August.)
Google learned from Netscape's demise and kept quiet about its long-term plans. It was a search engine, and search was at best a minor part of Microsoft's business. It generated advertising revenues, but Microsoft made its money through selling software, so that wasn't perceived as a threat, either. The first iteration of Google Apps was seen as a joke by Microsoft and dismissed. Google hired Andy Rubin, the founder of Danger (the developer of Sidekick mobile phones, which was subsequently acquired by Microsoft) and adopted the funky smartphone operating system (Android) that he had been working on. Again, it was under the radar and not worth Microsoft's time. By the time Microsoft fully realized how many of its businesses were under attack, Google was too big for Microsoft to kill.
If you're going into any market dominated by "old media" companies, position what you're doing as a way for them to retain market share and/or make more money, and make sure that they agree--otherwise, they'll kill you. Napster completely disrupted the business models of the big record companies, and they sued Napster out of existence, but not before they were crippled by music sharing. Apple stepped in and offered the record companies a way to make money from the growth in usage of digital media players. The record companies bought in, which ultimately resulted in Apple becoming the world's largest seller of music, and made the record companies dependent on Apple for their survival.
If your business depends on information or support from entrenched companies, you're going to have an uphill battle. Steve Huffman, one of the founders of Reddit, set off to create an airline ticket comparison service, Hipmunk, that makes it easy to find the lowest fares. This, however, threatens the airlines, which see it as decreasing their potential profits. So, Hipmunk has struggled to get access to the flight and pricing information that it needs for its service. The company has launched, but its most likely exit strategy (if it survives) is to be acquired by a search engine or a larger travel service.
When startups take on entrenched competitors, they almost always roll out the David and Goliath story. The reason that story has so much resonance is that the little guy beat the big guy, when in the real world, the big guy almost always wins. The lesson for startups is to avoid taking on the big guys until you're big enough to fight them as an equal.