It doesn't have to be a hard problem that you solve. Sometimes solving easy, but tedious, problems is a worthwhile goal, too. Really, most industry pains are solved with CRUD apps attached to some workflow. So you might be right that the "best startups" come from your own pains... but some completely adequate lifestyle companies come from just asking around and solving small problems. And there is nothing wrong with that.
I find that for those problems sometimes you don't even need to write a single line of code. Changing the process or the way people think about the problem can be sufficient.
Although I imagine if you drew a venn diagram with easy & important & unsolved then the intersection will be relatively tiny. And from a business defensibility standpoint the barrier of entry may be too low.
