Depends on what you metric you're using for 'failure'. I normally define it as some mixture of revenue and traffic (I don't mind low revenue if it's high traff

Depends on what you metric you're using for 'failure'. I normally define it as some mixture of revenue and traffic (I don't mind low revenue if it's high traffic and vice versa). But there are other important metrics too. People have already mentioned what you learn from the project. If you learn something, it's not a complete failure. Another metric is the effect you have on people. One of my coolest ideas (imho) was http://thingsinbooks.com http://thingsinbooks.com. It gets about 10 visits per day, so by that metric it's a failure. It never earned any revenue so it's a failure by that metric. I didn't really learn anything from doing it either. But despite the lack of traffic, people still link to it and even upload their finds from time to time (126 finds so far!). So in that way it has some elements of success, even though I've failed to capitalise on the interest and turn it into a popular site.