As wingerlang said, "I try to make it related to the functionality."
I do this completely. Every. Time. But on the condition that the name is catchy and the domain is available. If I cannot come up with a good name for it or if the domain is not available, I forego the side project.
https://mypost.io
I think I wanted mypost.com originally or post.io but they were taken. After thinking about it, I figured: My Post IO (Input/Output) I spend a few minutes on GoDaddy and usually stop by this website to check domain availability across the board: http://www.domainsbot.com/
I might spend days thinking about the name of it.. if its just not catchy, I really can't continue. Just to name a few that I had come up with that are projects "in the works" and not yet available ... GoMenu, MyBiz, etc.
I used to be afraid to "share the idea" of my side projects with people, or Hacker News, but honestly, I'm pretty sure if you tell two developers to develop the same application.... this application will likely have the same functionality, but the UI and the approach of how it works will almost always be different.
So for example GoMenu is just an idea for restaurants to be able to create menus and put them up online. I brainstormed it and I was thinking, "Menu To Go? Go To Menu? MenuGo? GoMenu?" And I figured: People want a menu on the go and restaurants want to get discovered.. why not? Anyways, this concept is not anything at all out of the ordinary. A search will reveal to you that plenty of these types of websites already exist. So why would I even bother? Because I tend to focus on UI, user-friendliness, and affordability when developing apps.
Got to love that capitalism and competition! At least 3 billion people are currently using the Internet.. if you could get a few thousand to use your web app and pay for it, isn't that enough motivation for you do something? I'm not saying that anyone should build the next Facebook, because you can't. Sure, you actually can, but you cannot become as popular and used as Facebook. And that's totally okay! But you can certainly be the competition! Even if less than 1%.. lets just say, 0.1% of Facebook users decided they did not like Facebook anymore and decided to use your product because it offered something unique, wouldn't that be a win for you to become the competition?
I tend to buy up domain names and most of the time, I do use them, though I have had a bunch that just fell through the cracks, or honestly, I just thought the domain names were not as catchy as I had hoped. So when it comes to my side projects and naming them...
1. the functionality must be in the domain name
2. the domain name must be easy to remember
3. The extension of the domain name must be a fairly common one that is also easily memorable. (while .com is the one I usually try to get, I've been fond of some TLDs like .xyz, .io, .cc, .biz, .blog, .online, etc.)