A talk starts long before you open your mouth, and it almost never starts with the text. The most common mistake is to sit down and write the first sentence. Half an hour later you have a paragraph you dislike and a conclusion that you're bad at speaking. In fact you skipped the stage that decides everything else: why you are getting up at all.
Purpose, not topic
A topic answers "what about". A purpose answers "what changes in the room". "Quarterly report" is a topic. "So the team stops dreading the new tracking system and logs at least one task in it this week" is a purpose. The difference is practical, not cosmetic: a topic can't tell you what to cut, and a purpose can.
Classical rhetoric sorted speeches by time: about the past (what happened), about the future (whether to act), and about the present (what we value). Name yours honestly. If you think you're talking about the future but you're actually defending the past, the room notices before you do.
Write your purpose as one sentence: "After my seven minutes, a listener will be able to / will want to / will understand ____." If the sentence won't come, it's too early to speak — not because you lack skill, but because you haven't decided what this is for.
Who is in the room
An audience is never "people in general". Three questions get you most of the way: what do they already know, what worries them about it, and what can they actually do afterwards. The first saves you from explaining the obvious and from skipping the necessary. The second shows you where resistance lives. The third stops you from ending on a call to action nobody in the room is able to perform.
If you don't know the answers, ask. Two short conversations with future listeners beat two hours of guessing.
Gather more than fits
A workable rule: collect three to five times more than your time allows. The surplus isn't there to be spoken. It's there so you have something to choose from, and so you stand in front of the room as someone who knows more than they say. That is audible in the voice.
Draw from four sources: your own experience (specific cases, with dates and details), checkable facts with a source you could name, other people's observations, and — in a separate pile — objections. Collect objections as diligently as arguments. You'll need them later; for now they show you where your position is thin.
Put one fact or one story on one card, paper or digital. Cards can be reshuffled later. A wall of prose cannot.
Exercise (25 minutes, today)
- 5 minutes. Write your purpose as one sentence, using the template above. Rewrite it at least twice — the first version is nearly always a topic in disguise.
- 10 minutes. Set a timer and produce 12 cards: facts, cases, numbers, things people said. No judging, no editing. Volume only.
- 5 minutes. Mark each card yes, no, or maybe against a single test: does it serve the purpose. Fewer than four yeses means the material isn't gathered yet — return to step 2 tomorrow.
- 5 minutes. Record 90 seconds of yourself answering, unscripted, "why am I talking about this at all", then listen back. Note the second where your voice came alive. That is your real purpose, and it may not match the one you wrote.
Repeat before every talk. By the fifth time it takes ten minutes instead of twenty-five.
A group helps at exactly one point: hand your purpose sentence to someone outside the subject and ask them to say it back in their own words. The gap between what you meant and what they heard is the cheapest edit you will ever make.
In closing
Preparation doesn't make a talk flawless. It makes it yours: you walk up knowing why you're standing there, holding more material than you'll use. Most of what you gathered won't be spoken. That isn't waste — it's the reserve that lets you stop fearing questions.