When I'm writing a speech, the fastest way to the answer is a conversation with the host:
- Tell me about the worst speech ever to this audience and why it went so badly, and
- Tell me about the best speech ever to this audience, and why it worked.
I probe on these, getting the specifics (without revealing who the hapless speaker was), and I have a wealth of information on what does not work. There is always something that would have appealed to me had I been in the audience, and that is a good reminder that a speech written that only delights me can be an ineffective speech.
Getting a response on the best speech usually is not as automatic. Here I really need to exercise my skills as an interviewer, often helping them put themselves back into the audience on that happy day. The specifics of what worked are usually easy for me to abstract, and they are gold nuggets for my own work.
My final step is to imagine who I know who would have hated the worst speech and loved the best. This person is my audience.
I've taken similar approaches with nonfiction books and articles, and editors have been extremely good at providing advice for these. Reading successes and failures helps for both fiction and nonfiction. Ultimately, you want to find someone you know who:
- Has similar interests to my ideal audience,
- Is comfortable with their vocabulary,
- Has about the same attention span,
- Is similarly knowledgeable,
- Has about the same "hot button" tolerance (regarding sex, religion, violence, politics),
- Has similar values and perspectives, and
- Will accept you as the source on this topic to a comparable degree.