Archive / Start a YouTube channel from zero with seven prompts

Start a YouTube channel from zero with seven prompts

From picking a niche to a 90-day plan, with a faceless production step for anyone who'd rather not be on camera.

For: Solo founders, Small business owners, New to AI

Topics: Marketing, Video

Why I'm sharing it

I'm planning video for one of my projects, and I won't be on camera. A post from gptcheats on Threads laid out a seven-step path for a new channel, and it included the step most guides skip: how to produce a video when nobody appears in it.

I've rewritten the prompts in my own words and tightened them so the AI has to explain its reasoning instead of handing you a list.

What makes it work

It starts with demand. Step 1 compares niches on audience interest, how crowded they are and how hard the videos are to make, before you commit to one.

Positioning comes before content. Step 2 forces an answer to the question every viewer asks: why this channel and not the big one I already watch?

Ideas need a reason. Each of the 30 video ideas in step 3 has to say why it deserves to be made.

Packaging gets its own step. Titles and thumbnails decide whether anyone clicks, and step 4 asks for curiosity without misleading anyone.

Faceless is a real production plan. Step 6 breaks a script into scenes with voiceover, stock footage keywords and text on screen.

The prompts

Step 1: Pick a niche

Prompt

Compare YouTube niches that have strong viewer demand, a way to earn money and room for a new creator. For each, weigh the competition, the usual video formats, what viewers want and how hard the videos are to produce. Given my [skills, interests and the time and budget I have], recommend five niches ranked by long-term potential and tell me why each one made the list.

Step 2: Position the channel

Prompt

Define a channel about [niche] that stands apart. Tell me who it's for, its angle, five name ideas, a one-sentence positioning statement, a visual style and the specific reason a viewer would pick it over the channels they already watch.

Step 3: Find topics people want

Prompt

Look at what's working in [niche]: the topics, the questions viewers ask and the gaps nobody covers well. Give me 30 video ideas, each with its angle, the viewer it's for and one sentence on why it deserves to be made.

Step 4: Write titles and thumbnails

Prompt

For a video about [topic], give me 10 title and thumbnail pairs. Use curiosity and a clear benefit to the viewer, and never promise something the video doesn't deliver. Tell me which three pairs are strongest and why.

Step 5: Write a script that holds attention

Prompt

Write a script about [topic] for [who the viewer is]. Open with a hook in the first few seconds, set up a question the video answers, break the middle into short sections with natural transitions, and cut anything that doesn't keep the viewer watching.

Step 6: Plan a faceless production

Prompt

Turn this script into a production plan for a video where no one appears on camera. Split it into scenes with rough timestamps, and for each one give me the voiceover line, B-roll ideas, stock footage search terms, text on screen, sound effects and editing notes.

Step 7: Plan the first 90 days

Prompt

Make a realistic 90-day plan for my [niche] channel. Set a publishing schedule I can keep with [hours] a week, three content pillars, one small experiment per week and the numbers to watch. Explain how I should use my analytics to decide what to make next.

How to use it

Run steps 1 and 2 once, carefully. Steps 3 to 6 are the loop you repeat for every video. Step 7 sets the pace.

Be honest about your hours in step 7. A schedule you can keep beats an ambitious one you drop in week three.

If you're going faceless, collect your B-roll as you go: hands at work, a screen recording, a walk outside. Step 6 is easier when you already have clips to choose from.

Inspired by a seven-step YouTube series from gptcheats on Threads. The steps follow their order; the prompts are my own rewrite.

Use it for: Deciding what a new channel should be about and getting the first three months planned before you film anything.

Back to the archive