Weekly AI Content Ideation System
Twelve prompts. One workflow. A system for generating better content ideas every week in Claude or ChatGPT.
Create, optimize, and measure content that actually drives leads with Content Hub.
The hardest part of producing great content isn't writing — it's deciding what to write about, consistently, without recycling the same angles or running dry mid-week.
This system, built on Kieran Flanagan's approach to AI-powered content ideation, helps solve that. You'll get a repeatable weekly workflow for generating ideas already filtered against your audience, your platform's proven patterns, and what's trending right now, plus 12 prompts to run the full process in Claude or ChatGPT.
The Workflow
The system runs in five steps. The first two are setup — you do them once and refine over time. Steps three through five are your weekly session.
This is not your ICP. It's a detailed picture of what makes your audience react — their emotional triggers, what turns them off, who they trust, and how they see themselves. This is what the AI checks before generating anything.
Look at your best-performing content across each platform and identify the patterns behind the performance — not the topics, the formats and angles. A contrarian take, a numbered how-to, a personal build story. These become your pattern library.
Feed the AI your audience profile and pattern library. Ask it to generate ideas mapped to both. This is where the quality gap opens up between this workflow and generic AI ideation — the ideas come back pre-filtered against what actually works for your specific audience on your specific platforms.
Cross-reference your ideas against what's gaining traction right now — on Reddit, X, LinkedIn, and across the web. When an idea maps to a proven pattern and a trending signal, that's your high-confidence pick for the week.
12 Content Ideation Prompts
Use these in Claude or ChatGPT. Run Steps 1 and 2 once to build your foundation, then use Steps 3–5 every week.
"Here is my current audience content profile: [paste profile]. Based on this, what gaps or weak areas do you see? What questions should I be able to answer about my audience that aren't covered here? Help me strengthen it."
"I don't have detailed performance data, but I want to build a platform pattern library for [platform]. Based on what typically performs well for [describe your content niche and audience], help me draft an initial pattern library with 6–8 ranked content patterns. I'll refine this over time as I gather real performance data."
"Here is my current platform pattern library for [platform]: [paste library]. Here is my content performance data from the last 30 days: [paste data]. Update the pattern library based on this new data — adjust the rankings, flag any patterns that are declining, and identify any new patterns emerging in my top performers."
"Here is my audience content profile: [paste profile]. Here is my platform pattern library for [platform]: [paste library]. Generate 10 content ideas for [platform] that are mapped to both. For each idea, note which pattern it maps to, why it fits this audience, and a one-sentence angle I could take. Do not recycle ideas from previous sessions."
"Here are the content ideas I'm considering this week: [paste ideas]. Search for trending conversations on LinkedIn, Reddit, and X related to each idea. For each one, tell me whether there's meaningful trending signal, what the signal is, and whether it strengthens or weakens the idea. Flag any ideas where the trending signal is strong enough to move them to the top of my queue."
"Search for the most-discussed topics in [your niche] across LinkedIn, Reddit, and X in the last 7 days. Cross-reference these trending topics against my audience content profile [paste profile] and my platform pattern library [paste library]. Surface the 5 ideas where there's the strongest overlap between trending signal and proven pattern."
"Here are this week's content ideas: [paste ideas]. Score each one from 1–10 based on: fit with my audience profile, match to a proven platform pattern, and strength of trending signal. Then categorize each idea as one of the following: contrarian take, data-led, educational, personal story, or numbered how-to. Present the results as a ranked table."
"I want to develop this content idea further before I write it: [paste idea]. Give me: the core argument in one sentence, three supporting points, one relevant data point or statistic worth researching, one story or example that would make this land, and the strongest one-sentence hook I could open with. Do not write the post — give me the building blocks."
Expert Insights From Kieran
When you ask AI for content ideas without context, it optimizes for topic relevance. That produces ideas anyone in your space could make. The better input is your platform's proven engagement patterns — the formats and angles that consistently perform well for your specific audience. When AI generates ideas mapped to those patterns first, the output is filtered against what actually works for you before a single topic is chosen.
Chasing trends pulls you toward what everyone else is already making. The more useful move is to generate ideas against your audience profile and pattern library first, then check them against trending signals. When an idea maps to a proven pattern and there's strong trending signal behind it, that's your high-confidence pick for the week. Trend without pattern fit is usually noise.
The prompts in this system are designed to give you research, pattern matching, scored ideas, and deep-dive outlines — not finished content. The value of AI in a content workflow isn't in the writing it produces, it's in the thinking work it accelerates. Your judgment, voice, and taste are what make the content worth reading. Use the prompts to get to a strong brief faster, then write from there.