There really isn't a truly typical work-flow. How I go about things changes depending upon what I am beginning with.

When making my fonts, most often I am resurrecting designs as drawn and/or printed in the late 1700s, early 1800s. So I would begin with existing scans of pages from manuscripts, though I have also scanned original drawings. I'll then bring in the scans into Xara Designer Pro or CorelDraw and trace over them and then refine the shapes, especially the angle of the serifs and their length/thickness.

Then I'll export as SVG files and import them into Font Creator from High Logic. Because in a drawing application a curve doesn't really have to be on a grid, there will often be curves I need to fix once imported because a font has to adhere to a grid based upon font units (don't worry about what that means).

Then in the font editor, I'll run through the basic glyphs I draw to make stem thicknesses equal and curves taper at the same angles, etc.

Once that is done I'll copy the various glyphs to other code points, draw the accents, make composites of the base glyphs and accents and ... and ...

At some point I work on setting what are called the side bearings. The goal here is to lessen the amount of kerning work. Then I'll kern the font, then work on writing the opentype feature code.

Mike