The biggest postwork change is using Photoshop to enhance the colors in the image. I can get lights close to what I want, but I've yet to get a render that I couldn't improve using Photoshop curves to beef up colors and contrast. That's fairly clear with this image where the original render looks rather drab compared to the corrected version.

Photoshop can be used to add things that cannot be readily done in Poser, such as using a bubble brush to add air bubbles to the image.  Trying to do that in Poser would take days whereas the bubble brush only took a few minutes.

After that, Photoshop is used to correct anything else that doesn't look right. I painted over some of the morphed hair strands that were starting to look like spaghetti, pushed in the jutting right elbow and shoulder, got rid of the odd right thigh highlight added radial blurs and darkening to make the far regions look farther, and a few other things. There is only so much that Poser can do efficiently and then it's best to use a paint or photo editing program to finish it.

Some Poser artists use Poser simply as a starting point and paint on hair, clothing, backgrounds, and a variety of other things in 2D programs such that the end result barely resembles the original render.

My visual indexes of Poser content are at http://www.sharecg.com/pf/rgagnon