Making money on YouTube with AI videos, long form and Shorts, Quentin IA Marketing

The dream they sell you is always the same: faceless AI videos running on autopilot, printing money while you sleep. The method is real, and it has become simple. But the part creators cut in the edit is the math.

The real question is not how to generate the video. It is which niche you pick and which format actually pays you.

You can produce a clean video in three minutes today, that was never the bottleneck. The bottleneck is that most people aim at the worst-paying niche on the platform and confuse the format that brings the audience with the one that brings the money. We are going to do both, in the right order: the math first, the production machine second.

Claude AI Lab

The Claude AI Lab is my Skool community where I share my Claude systems and the more advanced modules. Access is $67/month.

Join the Lab →

✅What you need

Claude. To find a niche that pays, pull the video ideas, and get the scene-by-scene breakdown.

A ViewMax account (viewmax.io). The suite that generates the video, the voiceover and the captions. Free credits to start, no card required.

A YouTube channel that is monetised, or the goal to get there.

An honest view of the numbers. The prerequisite nobody lists, and the one that decides whether you make money or not.

The principle in one sentence

Shorts bring the audience, long form brings the money. Pick a niche advertisers pay for, and you have solved 80 per cent of the problem before producing a single video.

The math nobody does

🧮 What each format really pays

Before producing anything, look at what YouTube actually pays in 2026, per million views, by format. The gaps are huge and they decide everything.

🎬

Long form faceless

Around 5,000 to 12,000 dollars per million views, higher in finance or tech niches. This is where the money is, especially past 8 minutes where you can place several mid-roll ads.

⚡

Shorts

Around 10 to 100 dollars per million views only. The creator takes 45 per cent of the revenue against 55 per cent on long form. A Short pays 10 to 50 times less than a long video at equal views.

🧸

Kids and nursery rhyme content

The LEAST paid niche: 300 to 2,000 dollars per million views. COPPA disables personalised ads on made-for-kids content, so advertisers pay far less.

Take two things from this table. First: the famous nursery rhyme niche that circulates in every tutorial is the worst one for money, not the best. Second: a Short is not a revenue source, it is a free billboard that sends people to your long video and your channel.

The right read

Use Shorts to get discovered, and long form in a real niche to cash in. Aiming for a million views on nursery rhymes to buy a car is the math that does not hold.

Picking a niche that pays

🎯 Claude finds it for you

A niche comes down to two things: what advertisers pay on it, and your ability to feed it without burning out. Claude crosses both in one query.

⚙️The framing prompt

Ask for a ranking, not a list. "Give me 10 faceless YouTube niches I can produce as AI video, ranked by ad revenue (RPM), with the typical RPM and why advertisers pay on them."

Filter by what you can sustain. Add: "Keep only the ones I can feed with [your field or interest] without becoming an expert."

Decide fast. A niche you hold for six months beats a perfect niche you drop after three videos. Pick one, do not scatter.

The trap to avoid

Do not pick your niche on the views it generates, pick it on what it earns per view. A finance channel at 200,000 views out-earns a nursery rhyme channel at 2 million.

The script and the scene-by-scene breakdown

✍️ Ideas, then scene by scene

This is where Claude does the heavy lifting. You go from a niche to a video ready to generate in two prompts.

⚙️The flow

Pull the ideas. "Give me 10 viral video ideas in the [your niche] niche, each with an angle and a first-three-seconds hook."

Ask for the breakdown. Take the best idea and follow with: "Give me a scene-by-scene breakdown for a one-minute video on this idea. For each scene: the visual description, the on-screen text and the voiceover line."

Do the long version too. For the format that pays, ask for the same breakdown over 8 to 10 minutes. The long video is your asset, the Short will be an extract of it.

Why scene by scene

A video generator produces what you describe to it. A scene-by-scene breakdown is exactly the input format ViewMax needs, so you paste and generate, with no rewriting.

Producing the video in ViewMax

🎥 Generation, voice, captions

ViewMax is the suite that turns Claude's text into a finished video. The generator runs on Sora 2, and the voiceover and captions are built in.

⚙️The production

Open ViewMax and start a creation. On viewmax.io, click create a video, then choose the AI video generator.

Paste Claude's breakdown. The scene-by-scene becomes your generation prompt. Run it, and give the tool two to three minutes to prepare the video.

Add the voice and captions. Choose a voiceover, apply a viral caption style. The watermark remover is there if you start from a source to clean up.

Export vertical for the Short, horizontal for the long form. One breakdown gives you both formats.

What ViewMax really removes

It removes the production bottleneck, not the strategy one. The video comes out in three minutes, but your niche choice and your hook decide whether it gets watched.

Publishing and holding the cadence

🚀 One long form, several Shorts

The structure that works is always the same: one long video as the asset, several Shorts as distribution.

⚙️The rhythm

Publish the long video first. Aim for 8 minutes or more to unlock multiple mid-roll ads, which is what doubles or triples your revenue per view.

Cut it into Shorts. Pull two or three Shorts per long video. They live their own life in the feed and point back to the asset that pays.

Hold the cadence. Posting two Shorts a day is doable with this machine. But consistency in a good niche beats volume in a bad one. Do not confuse activity with progress.

The signal that it works

Your Shorts push up the views on your long videos. If your Shorts run but your long form stays flat, your distribution works and your monetisation sleeps: the problem is in the niche or the offer, not the production.

What this method will not do for you

Two points of honesty, because this is exactly what the dream-selling videos cut in the edit.

The first: this is not passive income, and definitely not a sports car in 28 days. At the real numbers, paying yourself 150,000 dollars in one month with kids content would take more than 100 million views over the period. Production is instant, traction is not.

The second: the tool removes the production work, not the thinking work. The niche choice, the first-three-seconds hook, the consistency over several months, all of it stays your job. ViewMax and Claude make you ten times faster on the mechanical part, they do not decide for you what is worth producing.

Where to start

Open Claude, ask it for the 10 faceless niches ranked by revenue, and pick one you can hold over time. Pull one idea, get the scene-by-scene breakdown, generate the first video in ViewMax. Publish the long form, cut two Shorts, and watch which ones send traffic back.

The takeaway

Production is no longer the problem. What separates a channel that pays from a channel that grinds is the niche you pick and the format you treat as your cash register.