Brainstorming That Doesn't Just Give You Obvious Ideas

Difficulty: Intermediate · Estimated time: 15 min

What You'll Learn

Why This Matters

If you've ever asked Claude for ideas and got back a list that feels like it could apply to any brand, any client, any project — that's not a model limitation, it's an under-specified prompt. Sharper inputs get you past the obvious first layer and into ideas you'd actually use.

The Concept

Claude's first response to an open brainstorm prompt tends to reach for the most common, broadly-applicable ideas — the ones least likely to be wrong for anyone. That's useful as a warm-up, not as a final list. Two things fix it: constraints (budget, audience, format, what's already been tried) force more specific, less generic thinking; and explicitly asking for a second pass that goes past the obvious ideas pushes Claude to dig further instead of stopping at its first instinct.

Asking Claude to take on a persona ("brainstorm this as a sceptical CFO," "as a Gen Z social media manager") is another way to get a different angle than its default, neutral one. And once you have a long list, a structured framework — scoring each idea against criteria like cost, effort and impact — turns a brainstorm into an actual decision instead of a list you stare at.

Step-by-Step

Example

Prompt:

"Give me 10 giveaway hook ideas for a UGC content brand targeting small e-commerce owners. Budget is low, audience is time-poor. Then give me 5 more that push past the obvious ones — more unexpected angles. Score all 15 on effort vs likely engagement."

Good response looks like: a first list that's fine but fairly predictable, a second list that's visibly sharper and more specific, and a scored table making it obvious which 2-3 ideas are worth pursuing first.

Try This

Run a real brainstorm for a current project using the two-round structure above, then score the combined list against the criteria that actually matter for that project.

Copy This

Brainstorm [X] ideas for [GOAL], for [AUDIENCE].
Constraints: [BUDGET / FORMAT / WHAT'S ALREADY BEEN TRIED]

Round 1: give me [X] ideas.
Round 2: give me [X] more that push past the obvious ones — more specific, more unusual.

Then score every idea against: [CRITERION 1], [CRITERION 2] — and tell me which 2-3 to pursue first.

Common Mistakes

Level-Up Tip

For stuck brainstorms, try two contradictory personas back to back (e.g. "a cautious brand manager" then "an ideas person with no budget limits") — the gap between their answers usually surfaces the genuinely interesting option.

Action Checklist

Next Guide

Next: 10 — Stop Reinventing the Prompt Every Time, on turning your best prompts into reusable templates.