It's an interesting post, because there are probably two types of 'pivots':

1) Change the solution to the problem you're solving 2) Change the problem you're solving

(1) is a lot easier than (2).

I think the key here is to love the problem you're solving - if you've chosen a big enough problem (or important enough), then there should always be a market for a solution... don't get discouraged when the particular solution you tried doesn't work... you just need to find a new one that will.