How to Brainstorm Ideas With AI

Quick Answer

Start with a constraint-heavy prompt, not a vague one. Ask for 20 ideas, not 5 — the first 10 are obvious, the last 10 start getting interesting. Then use follow-up prompts to expand the strongest directions. AI generates volume and diversity; you select and develop.

What This Workflow Helps You Do

• Overcome the blank page and generate options quickly
• Find unexpected angles by prompting from different perspectives
• Evaluate and develop the strongest ideas systematically

Step-by-Step Process

Step 1: Set up a constrained brainstorm prompt

Vague prompt: “Give me business ideas.” → Produces generic output.

Better prompt:

Generate 20 ideas for [specific thing you want to brainstorm].

Context:
- Who this is for: [specific audience, not "everyone"]
- Goal: [what a good idea would achieve]
- Constraints: [budget, time, resources, format, scope]
- What I've already considered and rejected: [list if applicable]
- What has already been done that I want to be different from: [competitors, existing solutions]

Start with the most obvious ideas, then push into less expected territory. Label them (Obvious / Unexpected / Contrarian).

Step 2: Generate ideas from unexpected angles

After your first pass, generate from different perspectives:

Now generate 10 more ideas for [same topic] but from these specific angles:

1. What would this look like if it were 10x cheaper or simpler?
2. What would a competitor in a completely different industry do?
3. What would you do if you could only reach [very specific niche audience]?
4. What's the opposite of the conventional approach?
5. What would a first-time customer or beginner actually want here?

Flag which of these feels most original.

Step 3: Expand the most promising ideas

Pick 2–3 of the strongest ideas and expand them:

Take this idea: [paste the idea]

Expand it with:
1. A one-paragraph description of what it would look like in practice
2. Who specifically would find this most valuable
3. The main obstacle or challenge with this idea
4. What would need to be true for this to work
5. A realistic version and an ambitious version
6. One specific first step to test it cheaply

Step 4: Evaluate and rank ideas

Here are the ideas I'm considering: [paste your shortlist]

Help me evaluate each one against these criteria:
- Feasibility: How hard is it to execute with limited resources?
- Impact: What's the potential upside if it works?
- Differentiation: How different is this from what already exists?
- Speed to test: How quickly could I run a small experiment?

Score each 1–5 on each criterion and suggest which 1–2 are worth pursuing first based on the scores. Explain your reasoning.

Step 5: Generate names, titles, or angles

For content and creative brainstorming specifically:

Generate 15 [headline / title / name / angle] options for [describe the project or piece].

Requirements:
- Each should be distinct — not just slight wording variations
- Mix: direct, curiosity-based, outcome-focused, contrarian
- Audience: [who this is for]
- Tone: [the feel you're going for]

Star the 3 you think are strongest and explain briefly why.

Common Mistakes

Stopping at the first list. The first round of ideas is almost always the obvious ones. Push past it with follow-up prompts and angle changes.

Asking for too few ideas. “Give me 3 good ideas” produces 3 safe ideas. “Give me 20, including 5 that seem ridiculous” produces a broader range where the best ideas surface.

Not adding your own context. AI generates patterns from existing data. The most valuable ideas come when you combine AI’s suggestions with your specific knowledge of your customers, market, and situation — knowledge AI doesn’t have.