You don't have an ideas problem.
You have a finding problem.
Open any "100 YouTube video ideas" article and you'll see the same thing. Tutorials. Reviews. Reaction videos. Day-in-the-life. Top 10s. Challenges. Q&As. Storytimes. Versus videos. Behind-the-scenes.
These aren't ideas. They're categories.
"Do a tutorial" doesn't tell you a tutorial on what. "Do a reaction video" doesn't tell you reacting to what. The reader is handed 100 empty buckets and told to be creative.
That's not help. That's a worksheet.
Here's the truth nobody writing those lists wants to say:
The best YouTube video ideas already exist. Other people are making them. And they're getting millions of views while you're brainstorming.
This post is about how to find them.
Why "brainstorming" YouTube ideas is broken
Most creators sit down with a blank doc and try to think their way into an idea. Vibes. Inspiration. Maybe a list of 100 formats from a blog post.
Then they pick something that sounds interesting and hope it works.
Here's why that fails:
You don't know what your audience wants. You know what you think they want. Those are two completely different things.
You don't know what's trending in your niche right now. You know what was trending six months ago when you last paid attention.
You don't know which formats are saturated and which still have oxygen. You're guessing.
You don't know what the algorithm is currently rewarding for your topic. Nobody outside YouTube does — except for the videos that are already winning.
So you make a video. It gets 200 views. You blame your editing, your thumbnail, your title. You make another. 200 views again.
The video wasn't the problem. The idea was wrong before you ever pressed record.
The data is already on YouTube. You're just not reading it.
Here's the shift that changes everything:
The best YouTube video ideas for your channel are not ideas you generate. They're ideas you find.
Every minute, hundreds of hours of video get uploaded. Most of it dies. A small percentage takes off. The videos that take off are not winning by accident — they're winning because they hit the intersection of an audience's interest, a moment in time, and a format that retains attention.
That intersection is what an "idea" actually is. And YouTube has already tested millions of them for you.
The videos in your niche that are getting outsized views — relative to the channel size, relative to how recently they were posted — are the data. They're telling you what's working right now, for the audience you want.
You don't have to brainstorm. You have to look.
What separates a viral video from a flop (it's not what you think)
It's not production quality. It's not the camera. It's not even the title most of the time.
The single biggest predictor of a video taking off is whether the idea itself was already proven before the video was made.
A great idea executed at 60% beats a mediocre idea executed at 100%. Every single time.
Look at any channel that grew fast in the last two years. Pull their breakthrough video. Then search YouTube for the same topic. You'll find that within the year before their breakthrough, three to five other channels had hits on the same exact idea. They didn't invent it. They modeled it.
This isn't cheating. This is what every winning creator does. They watch. They notice patterns. They make their version.
The creators who don't grow are the ones still trying to be original from scratch.
How to actually find good YouTube video ideas
Here's the method. Five steps. No brainstorming.
Step 1: Define your topic in three to five words
Not your channel. Not your brand. Your topic.
"My productivity channel" is not a topic. "Productivity for software engineers" is a topic. "Notion templates for freelancers" is a topic. "Stoicism for men in their 20s" is a topic.
The narrower, the better. Narrow topics have less competition, more specific audience intent, and the wins are easier to spot.
If you can't say your topic in 3–5 words, your channel doesn't have a topic yet. It has a vibe. Fix that first. (If you're still picking your niche, the faceless YouTube channel ideas playbook covers that step first.)
Step 2: Search your topic on YouTube. Filter by "This year."
Open a fresh incognito window. Type your topic. Set the filter to "This year" (or "This month" if your niche moves fast).
You're now looking at what's worked in your topic recently. Not 2019. Not the same evergreen videos that have been ranking for five years. Right now.
This is your dataset.
Step 3: Find the outliers
Scroll through. Look at view counts relative to channel size.
A 2-million-view video on a channel with 10 million subscribers? Expected. Ignore.
A 500,000-view video on a channel with 8,000 subscribers? That's a winning idea. The audience exploded that video past the channel's normal reach. Whatever it was about, in that exact framing — it worked.
These outliers are the ones you want. Write down the title, the channel, the view count, the upload date, the topic angle.
Find at least 10 outliers in your niche. More if you can.
Step 4: Group them by pattern
Look at your 10 outliers and group them. You'll see patterns emerge fast:
- Several videos taking off with the same hook angle ("I tried X for 30 days," "Why nobody talks about X," "The truth about X")
- Several videos taking off on the same sub-topic (a specific tool, a specific question, a specific debate)
- Several videos taking off in the same format (deep-dive explainer, quick-tutorial, opinion piece)
The patterns are your idea pool. Each pattern is essentially the audience saying "more of this, please."
Step 5: Make your version of the pattern
Now pick the pattern that fits your voice and channel. Make your version.
Not a copy — your angle. Your insight. Your delivery.
If the pattern is "I tried X for 30 days," and X in your niche is currently a productivity method, you don't have to do the same productivity method. You can do an adjacent one. Or do the opposite ("I tried X for 30 days and quit on day 7 — here's why").
The pattern is what's working. Your insight is what makes it yours.
A worked example
Let's say your topic is "AI tools for small business owners."
Search YouTube. Filter to This Year. Sort by view count.
Within the first two pages you might spot:
- A 14-year-old channel hitting a 600K video on "I replaced my entire team with AI for a week"
- A 12K-sub channel landing a 220K video on "5 AI tools I use every day to run my business"
- A 30K-sub channel pulling 410K on "ChatGPT prompts that actually save me time"
- A 4K-sub channel scoring 180K on "I built a business using only AI tools"
The pattern is obvious within 90 seconds of looking. The audience for "AI tools for small business" right now is rewarding:
- First-person experiments ("I tried," "I replaced," "I built")
- Specific time windows ("for a week," "every day," "in 30 days")
- Concrete outcomes (replaced a team, saved time, built a business)
Your video idea isn't "AI tools for small business." That's a topic. Your video idea is something like: "I ran my consulting business with only AI for 14 days — here's what broke and what didn't."
You didn't brainstorm that. You found it. The audience already told you the shape.
That's the method.
Where this gets hard — and why most creators give up on it
Doing this manually for one video takes 60 to 90 minutes if you do it right. Searching, scrolling, noting outliers, grouping patterns, picking your angle.
That's fine if you publish one video a month. It doesn't work if you publish weekly. It really doesn't work if you publish 2–4 times a week, which is what most growing channels do.
The math kills it. Most creators try this method, do it once or twice, and revert to brainstorming because the manual version is too slow.
This is where automation matters.
Where AI helps with YouTube ideas (and where it doesn't)
Two categories of AI tool. Don't confuse them.
Prompt-based idea generators. You type "give me YouTube video ideas about X" into ChatGPT or a generic AI tool. It generates 20 ideas. The ideas are not modeled on what's working — they're modeled on what the AI's training data thinks "YouTube video ideas" sound like. The output is generic. The same prompt produces similar slop for everyone using it. None of those ideas have been validated against the audience.
That's why AI-generated idea lists feel interchangeable and rarely work.
Ranking-video-modeled tools. A different category. Instead of generating ideas from a prompt, these tools pull the videos already ranking and winning for your topic, surface the outliers, identify the patterns, and give you ideas based on what the algorithm is currently rewarding.
This is what ScriptFaster does. You give it your topic. It pulls the videos already winning in that topic. It shows you the patterns. Then it generates a script for your version of the pattern — modeled on what's actually working.
ScriptFaster works for any YouTube creator serious about ranking — narration, lore, history, mystery, deep-dive explainers, face-led essays, tutorials. Channels that publish weekly and depend on consistent output get the most out of it, but the analysis and script generation work for any format. Once you have the idea, the next step is the script — we wrote the manual version of that method in writing a YouTube script.
The point isn't that you can't do this manually. You can. The whole method is in this post. The point is that doing this for every video you publish, every week, year after year, only works if you compress the loop. Otherwise it's a great method you used twice and abandoned.
How to never run out of YouTube video ideas
The reframe: you're not trying to come up with ideas. You're trying to build a system for finding them.
A system that runs every time you sit down to plan a video. A system that doesn't depend on inspiration. A system that gets you to "this is what I'm making next" in 10 minutes instead of two hours.
The system has three parts:
- A clear topic. Three to five words. Narrow.
- A regular scan. Once a week, search your topic with the "This year" filter. Note new outliers. Watch the patterns shift.
- A pattern bank. A doc, a Notion page, a spreadsheet — somewhere you log the patterns you find. So when you sit down to plan, you're picking from your bank, not staring at a blank page.
Do this for three months and you'll never have an "I don't know what to make" moment again. Not because you got more creative. Because you stopped trying to be creative and started being observant.
The summary
YouTube video ideas aren't generated. They're found.
The videos already winning in your niche are the data. The outliers — videos getting views well past their channel's normal reach — are telling you exactly what the audience wants right now.
Your job:
- Define your topic narrowly.
- Search YouTube with the recent filter.
- Find the outliers.
- Group them by pattern.
- Make your version of the pattern.
The creators who grow fast aren't more creative than you. They're more observant than you. They watch what's winning, and they make their version while the pattern is still hot.
The blank page is a choice. The data is sitting right there.
Stop brainstorming. ScriptFaster pulls the YouTube videos already winning for your topic, identifies the patterns, and generates a script for your version. Any niche, any format. Free to try.