This is the difficult part, in my opinion.
Sure, "Ideas are all around us" but finding one that people will genuinely pay for, in a market big enough to be profitable yet not already saturated with high quality competitors, is very difficult.
However, if you know anyone in an industry other than software development, this can be a great source of ideas. Just ask a few questions about what their day to day job responsibilities entail. Then sit back and let them talk. I find that most people pretty quickly start talking about the pain points. Some of those pain points are likely to be something that people would pay to solve. Plus you give your friend/family member/acquaintance a chance to vent which they probably appreciate as well.
Like many things in starting a business, what "sounds" simple is in fact pretty difficult.
Learning to ask the right questions is also difficult in just figuring out the needs of your "acquaintances/friends."
If you can find just one small annoyance factor, and ask a secondary question just a little bit deeper, then a little bit deeper, then a little bit deeper, you may hit a wall of "I don't know," and that could be a potential solution.
Like regular conversation, questions tend to tree from one another.
So in SV a lot of people abandon these kind of ideas because majority of VCs and incubators will not invest into these ideas. So young entrepreneurs ends up chasing non-profitable "get eyeballs" first think about profit latter ideas.
And, no - market is not saturated with high quality competitors - for example, you can make Basecamp for "industry X" (i.e. teachers?, doctors?) you will make money to live quite comfortable.
For certain definitions of startup. NB if you ask a VC what they'd call a startup, it's in their interests to funnel you into thinking startups only ever go through VCs, go for growth rather than profit, and must end in an exit event (IPO or sale to another SV company). If you choose to accept that definition, you are limiting your choices dramatically.
If not being a startup can be a very good business, why not try it out? What have you got to lose? Not much. What have you got to gain? Valuable experience, and maybe some profits. Most startups actually began as just very good businesses (very small in scale), and take years to grow to something more substantial, changing along the way.
So in SV a lot of people abandon these kind of ideas because majority of VCs and incubators will not invest into these ideas.
37 signals (for example) is a successful company which many would call a startup (at least in their early days), but was not funded by VCs or incubators. That's not the only way to start or grow a company.
> you can make Basecamp for "industry X"