# How to Brainstorm a Speech Topic: A 4-Step Method

Canonical: https://sakura5.app/blog/how-to-brainstorm-a-speech-topic/  
Published: 2026-11-16  
Author: the Sakura5 team · 7 min read  
Updated: 2026-11-16  
Publisher: Sakura5 (https://sakura5.app) · support@sakura5.app

> **TL;DR:** How to brainstorm a speech topic: set your constraints, dump ideas from four angles, cluster them and test the finalists. A 25-minute method with an example.

![How to Brainstorm a Speech Topic: A 4-Step Method](https://sakura5.app/img/blog/inline/how-to-brainstorm-a-speech-topic-1.jpg)

To brainstorm a speech topic, start from your audience and occasion rather than from a blank page: write down every idea you could talk about, test each against who is listening and how long you have, and keep the one you can say in a single sentence. That order matters. Most people pick a topic first and discover it doesn’t fit the room only when they start writing.

A speech topic isn’t a subject like “leadership” or “the ocean.” It’s a specific claim or story that a specific group will care about for five minutes. Brainstorming is how you get from the first to the second without staring at a cursor.

Below: why the usual approach stalls, a four-step brainstorm you can finish in twenty minutes, a full walkthrough for a five-minute graduation toast, what to do when every idea feels boring, and a short FAQ.

## Why staring at a blank page doesn’t work

When you ask yourself “what should I talk about?” you’re asking one question that hides three: what do I know, what does this audience need, and what fits the time slot. Your head tries to answer all three at once, rejects every idea for failing one of them, and you end up with nothing on the page.

Two habits make it worse. The first is judging while you generate — crossing out an idea before it has a chance to lead to a better one. The second is starting too broad. “Technology” has no edges, so there’s nothing to push against. The fix is to separate the jobs: generate first, filter second, and give yourself constraints before you start filtering. If you tend to stall at the generating stage, our guide to [brainstorming alone effectively](https://sakura5.app/blog/how-to-brainstorm-alone-effectively/) covers the habits that keep ideas coming when nobody else is in the room.

## How to brainstorm a speech topic: four steps

### 1. Write down the constraints first

Before any ideas, fill in four lines: who is listening, why they’re there, how long you have, and what you want them to do or feel afterwards. “Forty colleagues at a Friday offsite, ten minutes, I want them to try one small habit” is a real constraint. “General audience” is not. These lines cost two minutes and they do most of the filtering for you later.

![How to brainstorm a speech topic: four steps — how to brainstorm a speech topic](https://sakura5.app/img/blog/inline/how-to-brainstorm-a-speech-topic-2.jpg)

*“Forty colleagues at a Friday offsite, ten minutes, I want them to try one small habit” is a real constraint.*

### 2. Dump ideas from four angles

Set a ten-minute timer and write fast, without judging. Use four prompts so you don’t just circle the first idea:

- **Things you’ve lived.** Jobs, mistakes, trips, a day that went wrong. Personal stories are the hardest thing for anyone else to give.
- **Things you know that others don’t.** Skills, trades, small expertise you take for granted.
- **Things the audience is worried about or curious about.** Imagine standing at the door and asking three of them what’s on their mind.
- **Things you’ve changed your mind about.** A belief you used to hold is a ready-made arc.

Aim for twenty to thirty fragments. Quantity matters here; the good one is rarely the third idea. The same [brainstorming](https://en.wikipedia.org/wiki/Brainstorming) principle applies as anywhere else: defer judgment until the list exists.

### 3. Cluster and branch the strongest ones

Now read the list and group items that are really the same idea. Two or three clusters usually emerge, and the biggest cluster is often your topic, because it’s what keeps coming up when you let yourself think freely. Pick the two most promising and expand each one level deeper: what’s the story, what’s the lesson, what would the audience do differently.

A visual layout is much better than a list for this step. In [Mindmap AI](https://sakura5.app/mindmapai/) you can paste your raw dump as text and the AI turns it into a mind map with grouped branches in a few seconds. That gives you a first draft of the clusters to argue with: rename branches, drag stray ideas to where they belong, delete what you won’t say. The outline view then shows the same map as a collapsible list, which is a rough speech skeleton.

### 4. Test each candidate against the constraints

Take your two or three finalists and run each through four questions:

1. Can I say it in one sentence with a verb in it?
2. Would the people in this room care, today, for this reason?
3. Can I do it justice in the time I have, with one story and two or three points?
4. Do I have something only I can say about it?

The topic that survives all four is the one. If two survive, pick the one you’d be more embarrassed to fumble, because that’s the one you care about. Once you have it, the next job is shaping it — our post on [how to organize thoughts before speaking](https://sakura5.app/blog/how-to-organize-thoughts-before-speaking/) picks up exactly where this one stops.

## A worked example: a five-minute graduation toast

Priya has been asked to give a five-minute toast at her younger brother Dev’s engineering graduation party. About sixty guests: family, his friends, two professors. She has one week and no idea what to say, so she starts with the constraints: sixty mixed guests, five minutes, a warm room, and the goal that Dev feels seen and the room laughs once and tears up once.

![A worked example: a five-minute graduation toast — how to brainstorm a speech topic](https://sakura5.app/img/blog/inline/how-to-brainstorm-a-speech-topic-3.jpg)

*Priya has been asked to give a five-minute toast at her younger brother Dev’s engineering graduation party.*

**The dump (ten minutes, 24 fragments).** She writes what comes: Dev’s first robot built from a toaster, the year he failed organic chemistry, the night he drove four hours to fix a friend’s laptop, his terrible puns, the group project that fell apart, how he calls their mother every Sunday, the time he refused to ask for directions and they ended up in a different town, and sixteen more.

**The clusters (eight minutes).** She pastes the 24 fragments into a map. Four branches form: *things he built* (six items), *times he failed* (five), *how he treats people* (eight), and *funny family stories* (five). The biggest cluster is how he treats people, and two of its items — the four-hour drive and the Sunday calls — connect to the failure cluster: the organic chemistry year is when he started staying late to help classmates who were struggling.

**The test.** Her three finalists: “Dev’s inventions,” “the failed year,” and “the people he shows up for.” The first fails question two, since the professors know his projects better than she does. The second is good but sad on its own. The third passes everything: “Dev’s best engineering is how he shows up for people, and it started the year he failed.” That’s one sentence with a verb. It has one story (the four-hour drive), a laugh (the toaster robot as the opener) and a tear (the Sunday calls). It fits five minutes with room to breathe.

Total time from blank page to a topic: about twenty-five minutes. The actual writing came the next day, and it was easy, because the hard decision was already made.

## When every idea feels boring

If the list is full of ideas that feel flat, the problem is usually specificity, not the ideas. “I learned teamwork” is flat. “The week our whole team slept in the office and I discovered who actually reads the instructions” is not. Take the dullest idea on the page and ask three follow-ups: when exactly did this happen, who else was there, and what did I get wrong? The answers are the speech.

If the opposite is true — you have too many good ideas — use the time limit as a hard cut. Five minutes is roughly 650 spoken words at a comfortable pace, which is room for one story and two or three points, not for a survey. Park the rest in a separate note; they’re the start of your next talk.

And if you’ve been handed a topic by someone else, you can still brainstorm: the angle is yours. Run the same four prompts on the assigned subject and pick the angle that you have a story for.

## FAQ

**How long should I spend brainstorming a speech topic?**

Twenty to thirty minutes is enough for most speeches: two minutes on constraints, ten on a timed idea dump, ten on clustering and testing. If you need longer, you probably skipped the constraints. Spreading it over two days helps, because your mind keeps working on the shortlist between sessions.

**What if I don&rsquo;t have any ideas at all?**

Start with the audience instead of yourself. Write what they’re worried about, curious about, or about to decide. Then ask what you’ve lived or learned that touches any of it. Looking outward first usually unlocks the personal material you thought you didn’t have.

**Should I pick a topic I&rsquo;m an expert in?**

Not necessarily. Expertise helps, but the audience mostly needs a specific angle and a real story. A topic you care about and have lived through beats one you know thoroughly but feel flat about. If you’re both expert and invested, even better.

**Can a mind map really help me choose a speech topic?**

Yes, because choosing is mostly about seeing which ideas cluster and which stand alone. A map shows that at a glance in a way a list doesn’t. [Mindmap AI](https://sakura5.app/mindmapai/) can build the first version from pasted notes, so you spend your time judging clusters instead of drawing them.

About the app: [MindMap AI: Mind Map Maker](https://sakura5.app/mindmapai/) · App Store: https://sakura5.app/go/mindmapai?from=/blog/how-to-brainstorm-a-speech-topic/
