[AI labs posts]

How To Know If A Video Idea Will Work Before You Film It

This post is part of the Complete AI YouTube Workflow series, a step-by-step breakdown of the exact system I use to run my entire YouTube channel with AI. You are on Article 5. This one completes the ideas and research section of the series. Articles 3 and 4 cover finding and stealing ideas, this is where you find out which of those ideas is actually worth filming.

‍

You have an idea. You think it is good. You spend a day filming it, two days editing it, and then you post it.

Thirty-seven views.

If you have been making YouTube videos for any length of time, you have lived that experience. And the brutal thing is it is not always the execution that was the problem. Sometimes the idea itself was never going to work for your audience, and nobody told you before you wasted the week on it.

That is the gap Task 3 fills. It is a validation layer that sits between having an idea and committing to it. A way of stress-testing a concept before a single second of footage is shot. And it takes about two minutes.

Why Most Video Ideas Fail Before They Start

Here is what I have learned from making over 1,300 videos and helping brands generate tens of millions of views. The failure point for most underperforming videos is not the production. It is not the editing. It is not even the title or thumbnail, though those matter enormously.

It is that the idea was never properly interrogated.

Most creators have a vague sense that an idea is good because it feels interesting to them. But interesting to you and compelling to your specific audience are two very different things. An idea might be fascinating but too niche to travel beyond your core subscribers. It might be broad enough to reach new viewers but too generic to be worth anything to the people who already follow you. It might be great in theory but missing the one angle that would make someone actually click.

Task 3 catches all of that before you film a single frame.

What Task 3 Actually Checks

The Video Idea Validator puts your idea through a structured assessment across three areas.

ICP fit. ICP stands for Ideal Client Profile, which in this context means your ideal viewer. The agent reads your Master Context Doc and checks whether this idea is genuinely relevant to the person you are trying to reach, or whether it is slightly off-target in a way that would attract the wrong audience or fail to resonate with the right one. This matters more than most people realise. A video that pulls in the wrong viewers actively damages your channel by confusing the algorithm about who to serve your content to.

The CCN framework. This checks whether the idea works across three different viewer types: your Core fans who watch everything you put out, your Casual viewers who dip in and out, and New viewers who have never seen your content before. The best video ideas serve all three groups simultaneously. They give your loyal audience something deep and satisfying, they are accessible enough not to alienate someone who just discovered you, and they have enough broad appeal to reach people who did not know they needed this content yet. An idea that only works for your core audience will not grow your channel. An idea that only appeals to new viewers will not build the loyal base you need.

Growth potential. This is the question of whether the idea can travel beyond your existing subscriber base. Could it get picked up by the algorithm and served to people who have never seen your channel? Does the topic have search volume? Does it tap into something people are actively curious about right now, or is it evergreen enough to keep working for months after you post it?

At the end of the assessment, Task 3 gives the idea an overall score out of 10 with two or three specific improvements. Not vague suggestions like "make the hook stronger." Concrete, actionable changes that would increase the score.

How to Run It

Open Claude Cowork and type:

Run Task 3

Cowork will ask you to describe your video idea in one or two sentences. Give it that and it comes back with a full verdict directly in chat. No file is saved for this one. It is a quick decision tool, not a document you file away.

That directness is intentional. You are not building a report. You are making a go or no-go call before you invest your week in something.

The Move That Multiplies Its Value

Task 3 is useful on its own. But it becomes genuinely powerful when you run it on the output from Task 1.

Here is the workflow. Every Monday, Task 1 automatically generates ten video ideas based on what is trending in your niche. You open that list and pick your top three, the ones that feel most relevant and exciting. Then you run each of those three through Task 3.

What comes back is a comparative picture. Idea A scores a 6 with suggestions to tighten the angle. Idea B scores an 8 but needs a stronger hook. Idea C scores a 4 because it is too niche to grow beyond your existing audience.

You make Idea B. You already know before you film it that it has strong ICP fit, works across all three viewer types, and has genuine growth potential. The only question left is execution.

That is a fundamentally different starting position than just going with whatever felt most interesting on Sunday night.

What a High-Scoring Idea Actually Looks Like

In my experience, the ideas that score highest tend to share a few consistent characteristics.

They solve a specific problem for a clearly defined person. Not "how to grow on YouTube" but "why your YouTube shorts are killing your long-form views." The more precise the problem, the more the right viewer feels like the video was made specifically for them.

They have a tension or a counterintuitive angle. Something that makes the viewer think "wait, I thought it worked the other way." That friction is what drives clicks. It is the gap between what someone assumes to be true and what the title implies.

And they are genuinely useful to someone who has never heard of you. The best videos on YouTube do not require any prior relationship with the creator. They earn the click from a complete stranger and then earn the subscription by delivering on the promise.

Task 3 is essentially checking for all three of those things without you having to consciously run through them yourself.

The Bigger Point About Validation

There is a mindset shift that separates creators who grow consistently from those who plateau.

Beginners treat every idea as equally valid until the analytics prove otherwise. They post, wait, see what happens, and adjust after the fact. That is reactive content strategy. You are learning from failure after the investment has already been made.

The better approach is to build validation into the process before you commit. Task 3 is the mechanism for that. It is not going to guarantee that every video performs. Nothing will do that. But it significantly shifts the odds in your favour by filtering out the ideas that were never going to work and strengthening the ones that might.

I have helped clients get 14 million views on a single video and 200,000 views in sixteen days. In every case, the idea was interrogated before it was executed. Not obsessively, but rigorously. There is a difference between overthinking and being deliberate. Task 3 is the deliberate version.

One Last Thing

The most common mistake with this tool is using it as a reason not to act. You run the validation, get a 6 out of 10, and decide to keep searching for the perfect 10 before you film anything.

That is not the point. The point is to make better decisions faster, not to wait for certainty that does not exist.

If an idea scores a 7 or above and the suggested improvements are things you can address in the concept, make the video. Done is better than perfect, but done with a validated idea is better than both.

Free guides on using AI to grow on YouTube

Subscribe Below

[WORK WITH OSCAR]