In your Tower point attribute table, you should have one field called OFFSETA which contains the height of the tower. In your case, 80 or 100 feet (or, 24.4 or 30.5 if the DEM is in meters). This is sometimes confusing, because the tower point is being used as the observation point. However, this is the correct way to do the analysis.
Reading your 3rd comment on Dan's answer, it looks like this is what you are doing (which is correct). So, why is there no change in the output? It might be due to the accuracy of the DEM you are using for the elevation raster. By default, OFFSETA is set to 1 (~3 feet if DEM is in meters). Assuming the tower points are all at high points in your DEM, there might NOT be any difference between a 1 meter OFFSETA and a 30 meter OFFSETA. Or, the difference might be so slight, you don't notice it.
A couple of things you could do to test: Use a really big OFFSETA (e. g. 1000') or move a tower location to a spot where you are sure a 100' OFFSETA will make a difference in the output. (e. g. 50' down-slope of a peak).