I built an AI YouTube analytics tool because I was tired of doing this manually
Hey Indie Hackers,
I love making YouTube content, but the analytics part always felt like a second job.
Every week I'd open my dashboard, stare at the numbers, and not know what to actually do next. So I'd copy everything into ChatGPT and ask "why did this video flop?" The advice was always generic.
So I built CreatorPilot to solve exactly that.
What it does:
- AI video analysis — why each video performed well or poorly, with specific fixes
- Daily action plan — 3 concrete tasks based on your real channel data
- Trend matching — today's trending topics matched to your content style
- 20 next video ideas with predicted view counts
- Channel pattern analysis — your winning title/time/length formulas
Numbers:
- Built solo in ~2 weeks
- Tech: Next.js, PostgreSQL, Claude API (Haiku), Vercel, Stripe
- Gross margin: ~90% (aggressive DB caching on AI calls)
- Pricing: Freemium, $9.90/month PRO
- Languages: EN, KO, JA, DE, ZH
Launched today on Product Hunt. Would love feedback from fellow indie hackers.
Tried the site, the core idea is strong, the daily action plan and trend matching especially are things most analytics tools skip entirely. One thing worth fixing before more traffic hits it, your homepage nav already shows Dashboard, Settings, and your current plan status, that's app navigation, not marketing nav, so people who haven't signed up yet are seeing your product's internal structure instead of a reason to sign up. Worth splitting the public landing page from the actual app shell.
This is genuinely useful, thank you for actually digging in. You're right about the nav issue — I built the app shell first and never went back to separate it from the marketing/landing experience. New visitors shouldn't be looking at Dashboard/Settings before they've even signed up. Adding this to the fix list this week. Appreciate you taking the time to test it and point out something most people would just silently bounce from instead of mentioning.
What I'd watch carefully is the gap between explanation and confidence.
A creator can receive a very convincing reason for why a video succeeded or failed and still be no closer to understanding what actually happened.
The more useful the explanations feel, the more important that distinction becomes.
I completely agree.
One challenge with AI-generated insights is that they're often more persuasive than they are accurate.
What I'm aiming for is not prediction or certainty, but helping creators spot possible patterns faster than they could manually.
I'm currently thinking about adding confidence scores and more supporting data so users can see why a recommendation was generated instead of just trusting the explanation blindly.
Thanks for raising that.
That's what I'd be curious about.
Sometimes additional evidence helps people calibrate confidence.
Sometimes it simply makes an explanation feel more convincing.
Those aren't always the same thing.
That's a fair and honest point.
The distinction you're drawing — between feeling more calibrated
and actually being more calibrated — is something I don't have
a clean answer to yet.
What I can do is make the underlying data visible enough
that a skeptical user could challenge the explanation themselves.
Whether that's enough to avoid the trap you're describing,
I'm genuinely not sure.
It might just be a fundamental limitation of this kind of tool.
Appreciate you pushing on it.
That's exactly the kind of uncertainty I was referring to.
Probably more than I'd try to unpack properly in a thread.
Happy to continue by email if useful.
Interesting product. I like the idea of helping creators move from scattered content creation to a more repeatable system.
A lot of creators do not only struggle with writing posts. They struggle with consistency, repurposing, and knowing what to publish next without starting from zero every time.
The strongest value here would be if CreatorPilot can help turn one idea into a full content workflow: idea → outline → post → variations → schedule → review.
That is where AI becomes more useful than just “generate me a post.”
Thanks for the thoughtful feedback, Ali!
You're absolutely right — the full content workflow you described (idea → outline → post → variations → schedule → review) is exactly the direction CreatorPilot should grow toward.
Right now the focus is on the "review" and "what to publish next" parts — analyzing past performance and recommending next steps based on real channel data. But the workflow piece you mentioned is a genuinely valuable next layer.
Would love to keep you posted as we build toward that. Thanks for taking the time to share this.