Keyword & Search-Demand Mining

Search is the one place on YouTube where the audience tells you what they want before you've made anything. They typed it into a box. All you have to do is go and read it.

So the job is to collect the real questions and real problems people in your niche are searching for, rank them by how many people are searching, and turn each one into a video. vidIQ will build that list for you. So will an AI prompt, as long as you ask for the actual questions and problems people search in your niche and tell it to order them by search demand. Either route gets you the same thing. A queue of topics with proof of appetite already attached.

Then use the free tool that's been sitting in front of you the whole time. Type your niche word into YouTube's search bar and stop before you press enter. Read what drops down. Those are phrases people are typing right now. Now type your word and the letter A. Read those. Then B. Then C. Work the alphabet end to end and copy down anything that fits. It costs nothing, it takes a morning, and you'll walk out of it with a better topic list than a full day of brainstorming produces. Google's People Also Ask box does the same work from a different angle. Every line in it is a question somebody actually asked.

Reddit is the third well. Find the subreddits for your niche, sort by top, and watch for the question that gets asked over and over again. People go and ask on Reddit precisely where YouTube's coverage is thin. A thread with two hundred replies and no decent video answering it is about as loud a signal as you'll get.

Now the discipline, because collecting keywords is easy and wrecking the list is easier.

Relevance beats raw volume. Every time. You'll find phrases with enormous search numbers and only the faintest connection to what your channel actually does, and they will be tempting. Cut them. A big keyword that isn't yours drags in people who leave in twenty seconds, and it teaches YouTube to show you to the wrong room. Sniper, not spray and pray. You're hunting the most valuable keywords, not the biggest ones.

If you're working in vidIQ, open the Questions tab. It hands you the phrasing word for word. Then answer that question as literally as you can stand, because YouTube rewards the exact match. If the query is seven specific words, your video is about those seven words. Not your cleverer rewrite of them.

A few habits that quietly pay off. Spacing and spelling move the numbers, so check both versions of a phrase and build on whichever carries more volume. If AI throws you a piece of jargon you've never heard, don't take its word for it. Run it through vidIQ and confirm humans are searching for it before you sink a day into a video nobody wants. And you can point vidIQ at a competitor's video to see the exact keywords it ranks for and what position it holds for each. That's their search strategy, itemised.

You're looking for the same shape every time. Real volume, competition that isn't brutal. The Pomodoro technique is the textbook version of it, fifty thousand searches a month sitting at medium competition, which is a door left open. If your niche looks crowded, bolt a modifier onto the main keyword and look again. Something like "reimagined" will often surface a pocket of demand nobody's serving. You can also take two keywords you've already researched, fuse them into a single idea, and hand the pair to AI for concepts.

Then pick your first topic where the highest-demand query meets the concept you can explain most easily. Demand you can actually deliver on. Start there.