"Okay smart-ass, you've shot down all my dreams of achieving scene stardom and world domination with my glenz phong metaballz. Any last words?"
Look around you
Keep your eyes open. Look at water, at trees, at buildings, at people, at the sky, at clouds, at cars, at reflections on the asphalt.. You'll get inspiration. Believe me.
Know your limits
Okay, you're a beginner. You're no Smash Designs, Kewlers or Future Crew, it's better just to admit it. Trying to imitate them will only lead into a disaster. Good versions of simple effects are often better than poor versions of advanced effects. If someone has made a killer effect, doing something that just about resembles it doesn't mean that it's almost a killer effect. Usually it looks like total shit.
If your 3d-engine is not as good as Smash Designs', don't do a 3d-flyby. If you only got spinning cubes, don't go for the goa-trance-psychedelic-LSD-coloured style. Think of what you can do with the stuff you've got and then do it as well as you can.
I can give you a warning example: When we were designing our would-be-debut Viping, we had ambitious plans. So ridiculously ambitious that an all-star team of Kewlers, Farbrausch and Smash Designs members couldn't probably make them work now. It's embarrassing to read the design notes and watch the demo now, because it turned into a steaming pile of junk. Don't try too much.
Don't give up
If you thought you'll be doing a celebrated megademo instantly, you thought wrong. If you get nasty comments on your production or people laugh at you, don't go away crying and decide that those evil people can shove their demos where the sun doesn't shine. Instead ask them how to improve your work. If they're still being nasty to you, they're not worth your time. Remember: the worst demo is one that doesn't receive any feedback. Even the worst feedback is better than none.