First off....
First of all, try mixing the song WITHOUT the lead vocal for a bit.
Listen carefully.
Listen to make sure that the drums, bass, guitar misc instruments aren't all taking up too much frequency spectrum.
Carve out a spot for the vocals. Vocals don't "compete" well in mixes. They need to be given room.
This often means cutting mids out of drums, and instruments and such to "carve" out a hole.
Next, after you have a frequency band opened up for the vocals, ensure that the surrounding dynamics aren't going to bury the vocals. This means making sure that guitars, bass, drums etc are compressed as necessary to keep things from being so overly dynamic that it dynamically will "mask" the vocal.
I usually mix without the lead vocals for the first two or three passes until I have 1) Frequency availability, and 2) Smoothed dynamics. Only then will I start mixing in the lead vocal.
Once you have most of the bed tracks in place and mixed properly, you can then spend your time on the dynamics and EQ on the lead vocal. You will still have to go back and tweak the bed tracks, but at least you will have a starting point.
On the lead vocal, carefully pay attention to the compression applied. Play around with attack time, release time, threshold, and ratio to make it "fit" the best. This is usually different for every song and singer.
If someone asks to have a sound that is something similar to their favourite band, I usually listen to the recording and the band when the singer ISN'T singing. Listen to the way that the "hole" is created for the vocal dynamically and spectrally.
This of course, assumes that the singer being recorded can sing well enough to "soar"
SH