I forgot to mention, you'll want to have some empty space left in the audio file before the word starts, just so there's time for the reverb to fade in. The best way is to just duplicate the whole vocal track, do the reverse reverb processing to it and then just delete the parts you don't want