Stop shooting videos nobody is searching for. The vidIQ + Claude playbook, free.

What you end up with

You stop guessing what to make. You ask a question in plain English and get back what people are actually searching for this month, whether anyone is covering it, what is working right now for channels your size, and a title built from the words people type instead of the words you like.

Honest version: this is a research tool, not a growth tool. It tells you what to make and what to call it. It does not make you good on camera and it does not write your hook. Those are still yours.

What you need

• A paid Claude plan, so you can use connectors.
• A vidIQ account. The free one answers most of what is below.
• A channel, or an honest idea of the one you are about to start.

1. Connect it

Claude › Settings › Connectors › vidIQ › Connect › Sign in

Test it in a new chat before you plan anything: "How many people search for [something in your world] on YouTube each month?" If a number comes back, you are connected.

2. The five prompts I actually use

Find me topics in [your industry] with rising searches and low competition. Show me monthly search volume, the trend, and the competition for each.

I am planning a video called "[your title]". How many people search for this? Is there a version of this title more people are searching for?

Find Shorts about [topic] from channels under 100,000 subscribers that got far more views than usual. What do the hooks have in common?

Pull the transcript of [video link]. Break down how the first 5 seconds hook people, and how the video is structured.

Give me 5 titles for this video that match what people actually type into YouTube. Here is the script: [paste it].

3. Turn the answers into a week you can actually shoot

A list of topics is not a plan. This is the prompt that turns the research into things you can film in one sitting.

Here are the topics you found. I can film for [two hours] this week.

Pick the [five] worth making, and for each one give me:
- the search people are typing, and the monthly volume
- the title, in their words
- the first line I say out loud, which is not the title
- the one point the video makes, in a sentence
- what I have to show on screen for it to be believable

Only pick topics I can speak to from experience: [one line on what you actually do]. If a topic has big volume but I have no real story for it, leave it out and say why.

4. Take one video that went big apart

This is worth more than any list of topics. Find a video in your world that did far better than the channel normally does, and study why instead of guessing.

Pull the transcript of [video link] and the stats for that channel.

Tell me:
- how many views this got compared to what that channel usually gets
- the first 5 seconds, word for word, and what it is doing: a question, a number, a claim, a face
- how the video is built, section by section, with roughly how long each part runs
- where it would have lost me, if anywhere
- what I could steal about the structure without copying the content

Then write the version of the first 5 seconds that would work for [what you do], in my words, not theirs.

The rule I work by

The search sets the title. The hook on screen is the human line. "How to Connect vidIQ to Claude" is the title, because that is what people type. "176,000 people search this every month" is what you hear in the first second, because nobody says a title out loud.

What goes wrong, and the fix

• You chase the biggest number. Volume with a wall of established channels in front of it is worth less than a smaller number nobody is covering. Ask for both and read them together.
• It hands you topics you cannot speak to. Tell it what you actually do and give it permission to throw topics out.
• The titles come back as clickbait. Say it plainly: the title has to be true after they watch it. A title that beats the video costs you the next one.
• It quotes a number and you repeat it on camera. Search volumes are estimates. Say "about" or do not say it at all.
• Every idea starts sounding the same. Feed it your last ten videos and ask what you keep making, then ask for the one you have never made.

Before you shoot

• Say the first line out loud. If it sounds like a title, rewrite it.
• Check you have the proof. If the point needs a number or a screen, have it ready before you hit record.
• Check the title against the search, not against your taste.

Search numbers are a direction, not a promise. The research tells you what to point the camera at. It cannot tell you whether what you say next is worth hearing.