In my experience during the dot-com bubble, the biggest predictor of failure was having taken VC money. When you can't make a VC's timetable for returns (and few companies can in a recession), they're apt to pull the plug and divert cash to those startups that can deliver returns satisfactory to the VCs investors, like Akamai or Google. While if you bootstrap and refuse outside investment, you can batten down the hatches, cut expenditures, change your business model (like HotOrNot), and come through the recession on savings.