In my daily life I see a lot of fledgling companies and startups, some in existence for many years, delayed adolescence. Several observers have noted that a number of startups looking to make their mark in the Webosphere are becoming roadkill. That should be no surprise. Given you can start a company for next to nothing and there is no shortage of ideas, venture capital or angel money chasing the next big thing, companies will flame out early and often. Many startups are mining the same area, amplifying a feature rather than breaking any significant new ground. Only a few will survive, and luck, timing, connections, focus and a bit of technology will determine their fate.
Part of the problem for fledging companies is on the customer side, especially B2B and B2C. A company may have a neat technology that we could use at CNET/ZDNet, but there is a serious risk in growing dependent on a company that may be gone tomorrow, for whatever reasons, or come under the control of another company. In some cases, being the guinea pig is worth the risk and investment on the engineering side if it uniquely satisfies a strategic user need. And, that is the key for any startup--figure what strategic user need the company solves for what set of customers and dig deep. That is the only way to avoid ending up in the abyss, enjoying the benefit of the doubt when trying to get to the major customer references stage.