The behavior of any big company is largely inexplicable when viewed from the outside. Odds are, nobody inside Microsoft, Google, or Yahoo knows what Microsoft, Google, or Yahoo is going to do in any given circumstance on any given issue.

When I was at IBM in the early 90's, they had a formal decision making process called 'concurrence'—a written list of the 50 or so executives affected by any decision would be assembled, and any one of those executives could 'nonconcur' and veto the decision. Even non-extreme versions of this process are mind-bendingly complex to understand from the outside.

Don't do startups that require deals with big companies to make them successful. The risk of never getting those deals is way too high, no matter how hard you work at it—and even if you get them, they probably won't work out the way you hoped.

Never assume a deal with a big company is closed until the ink hits the paper and/or the cash hits the company bank account. There is always something that can cause a deal that looks closed to suddenly get blown to smithereens—or vanish without a trace.

Big companies care a lot more about what other big companies are doing than what any startup is doing. Hell, they often care more about what other big companies are doing than what their customers are doing.

Beware bad deals. One high-profile startup did two big deals early with high-profile big company partners and became completely hamstrung in its ability to execute on its core business as a result.

If doing deals with big companies is going to be a key part of your strategy, hire a real pro who has done it before. This is why senior sales and business development people get paid a lot of money. They're worth it.

Don't get obsessed. Don't turn into Captain Ahab. By all means, talk to big companies about all kinds of things, but always be ready to have the conversation just drop and to return to your core business.