In my workflow, for motion graphic videos,
Adobe Soundbooth:
I start with the Audio. I get the music decided upon first and then I record any dialogue. Dialogue is usually edit to fit the length of the song or vice-versa depending on which is better or more important. This establishes the length of the video and really lays the foundation for the completed product.
Adobe Photoshop:
All the images to be displayed in the animation get created and saved for use.
Adobe After Effects:
The images and audio are import into after effects and laid out and animated based on the completed audio. This includes any lettering or annotations that fly out to the dialogue or images that get relieved at a specific point in the music etc. Color correction, motion blur, glows, lens flairs, -all kinds of things, get applied to the video and then it is rendered.
Adobe Premiere or Final Cut Pro (Mac only):
The after effect file(s) and any conventional footage of talking heads or locations get all put together in Premiere and any final tweaks are made to audio. Transitions between clips are added and the final video is output.
If I don't have any transitions or conventional footage that needs to be put together with the motion graphics, I can just output from After Effects as a single file and skip Premiere.
