How I Found My Untapped YouTube Niche With 6 Claude AI Prompts

Good content isn't enough. You can post consistently, get your lighting right, and follow every YouTube growth tip you've seen — and still watch your videos flatline at 40 views. The problem usually isn't quality. It's that you're making the right content in the wrong lane, invisible in a crowded market where bigger channels already own the topic you're trying to cover.

I spent over $15,000 on YouTube growth coaching trying to solve this. It didn't. Then one conversation in Claude Cowork helped me reverse-engineer the exact gap in my market that nobody else was owning — not by guessing, but by analyzing my own channel data alongside 100 competitor channels simultaneously. The system runs six prompts, in order, and produces a niche assessment report with a Venn diagram at the end that tells you exactly where to plant your flag.

Here's the full 6-prompt system.

Before You Run Any Prompts: The Setup

Run all six prompts in the same Claude conversation — don't open a new chat for each step. Each prompt builds on what the previous one established. Three things to have ready before you start.

To give Claude access to your YouTube Studio data, you have two options: connect the VidIQ MCP connector in Claude Cowork (the faster route if you have a VidIQ account), or install the Claude Chrome extension so Claude can access your browser directly. Both are free, and both methods are covered in the video above.

Prompt 1: Prime the AI — Teach Claude Who You Are

Before Claude can find your gap, it needs to understand your business, your story, and your offer. The first prompt feeds it everything: your landing page URL, your positioning statement, the backstory that led you here, and links to your LinkedIn and YouTube channel so it can read your actual public presence.

The format that works: your backstory as a three-beat sequence — the relatable pain you started with, the tipping point that forced change, and the aha moment that led to your current system. Then: "I worked my process into a system and now I help [audience] achieve [outcome]." Claude reads your LinkedIn and YouTube data directly, then summarises what it understands about you in five bullets. Read them. Correct anything that's off. This is the foundation every subsequent prompt builds on.

Prompt 2: Audit Your Channel — Find the Hidden Bright Spot

This prompt asks Claude to dig into your YouTube Studio analytics: subscriber growth, top 10 videos, bottom 5, traffic sources, top countries, average view duration, and retention by video. The goal is to separate two very different things — vanity views (high numbers, wrong audience) versus buyer-aligned views (lower numbers, people who'd actually pay you).

The most valuable output is the "hidden bright spot": the topic or format that quietly outperforms on your real audience, even if the view count looks small. Running this on my own channel surfaced a hard truth: YouTube had me typecast as a video production tips channel — not a client-getting channel — because one teleprompter video was pulling strong numbers. That video was the signal the algorithm used to decide who to show all my other videos to. My new videos were being served to aspiring creators, not service founders willing to invest $1,500 in a strategy session.

"Your actual buyers respond to failure → number → rebuild stories. 'I lost $8K in a coaching program. Here's what I learned.' Specific number, specific method. That's the pattern."

Also check: where is your traffic actually coming from? My data showed that YouTube Shorts were inflating my raw view count by 38% — while contributing almost nothing in watch time or business outcomes. If Shorts are doing the same for you, the data says stop making them.

Prompt 3: Map the Market — Cluster What's Actually Winning

Paste your competitor channel URLs into this prompt (or tell Claude to pull them from your VidIQ or Velio list). Claude clusters the overperforming content into three to five themes, identifies which channels are winning each theme, and — critically — flags which themes have small channels breaking out.

Small channel breakout energy is the signal you want. If a channel with 2,000 subscribers has a video pulling 50,000 views, it means the topic is doing the heavy lifting, not the audience. Those are the themes you can compete in right now without waiting to build a following first.

In my analysis across 100 tracked channels, two themes had the strongest small-channel breakout energy: Claude AI as a real money/client tool (step-by-step, solve something in your business), and challenging dogma ("you don't need [popular thing], do this instead"). Both themes were pulling significant audiences for tiny channels — which means the topic carries, and you don't need a big platform to win.

Prompt 4: Find the Overlap — The Gap Nobody Owns

This is the prompt that changes things. Claude is now asked to cross two datasets: your unique advantage (story, skills, credibility — from Step 1), and the market's proven demand (the breakout themes from Step 3). The output is the white space at the intersection: what you can own that nobody else is currently owning.

Here's what it produced for me, verbatim:

"Service founders with real experience are searching for a way to use AI to get clients — but everyone teaching this right now is either selling AI services, chasing generic make-money-with-traffic content, or doing YouTube/content strategy without the AI system. Nobody is standing at the intersection of: I'm a non-technical service founder, I use Claude to build my entire client-getting system, here's how you can do the same — without becoming an AI person, becoming a YouTuber, or spending $15K on coaching that doesn't work. That seat is empty."

Three data points converged on the same gap: my highest-engagement content came when I combined a personal financial loss story with a specific outcome number ($8K lost, $19K from a VSL script, $48K from my bedroom). The market's breakout themes were Claude-as-a-business-tool and contrarian takes. The intersection — a video that didn't exist yet — was: you don't need subscribers to get clients; here's the Claude system I built instead.

Claude also mapped every named competitor and showed exactly which slice of the audience each one serves — and confirmed that the specific combination of rebel founder origin story + Claude + inbound client-getting belongs to nobody. The gap was real and unoccupied. If you want to see how to build the client system that fills it, see what working together looks like here.

Prompt 5: Validate the Gap Before You Commit

A gap you can't validate is a nicer guess. This prompt runs four checks on the niche you just identified, and it needs VidIQ connected to pull real keyword data.

Verdict from Claude: the gap is real, competition is low for now, all four checks pass — but the framing risk is that it's too invisible as a positioning line. The tweak: lead with the belief shift, not the mechanism. The contrast isn't "AI + YouTube." It's stop hunting clients, start attracting them. Claude for inbound. That's the line every video and description should reinforce.

Prompt 6: Name It in One Line

The last prompt compresses everything into a single sentence you can say in a breath, using the shape: "I help [who] achieve [outcome] using [vehicle] without [pain]." Claude produces three versions, from safest to boldest, then recommends one. You don't have to take the recommendation — pick the one that actually sounds like you.

My three options:

• Safe: "I help service founders get clients from YouTube using Claude — without needing to go viral or do cold outreach."
• Punchy: "I help service founders use Claude to build a YouTube channel that attracts clients — without chasing views, building a big following, or sending a single cold DM."
• Bold: "I help non-technical service founders use Claude to become the only obvious choice for their ideal client on YouTube — without going viral, becoming a creator, or doing cold outreach."

Claude recommended the punchy version. I like the bold one. The output isn't gospel — it's a starting point you edit to match how you actually talk.

After the six prompts, run one final command to pull everything into a single niche assessment report. Claude produces it as a clean document, plus a downloadable SVG Venn diagram showing your exact positioning overlap. Feed both back into future conversations as your positioning brief — so every script, LinkedIn post, and email starts from the same foundation.

Frequently Asked Questions