Finding Tags Worth Using: vidIQ & Competitor Mining

You've got the title-based tags down, so the only question left is where the rest come from. And you don't invent them. You go and look at what's already working.

The tool for this is vidIQ. Install the browser extension and it puts the tag list of any video you're looking at right there on the page, which turns every video on YouTube into research you can read for free.

Start with the reference channels in your niche. The ones making the thing you're making, at the size you'd like to be. Open a few of their videos, look at what they've tagged, and you'll notice something quickly: the good ones aren't using anything clever. They're using the specific, obvious language of the topic. That's the point. You're not looking for a secret keyword. You're checking that you're describing your video the same way the channels that already rank describe theirs.

Then get more targeted. Take your own video title, search it on YouTube, and look at what comes back. Those results are the videos you're directly competing with for that phrase, on that exact topic. Pull up their tags with vidIQ and take the ones that genuinely apply to your video. Those tags are proven, in the sense that they're attached to videos YouTube is already surfacing for the thing you're about to publish.

vidIQ will also make suggestions for you. The Boost Video feature gives you a list of recommended tags with a score attached to each one. Sort that list by the overall score and work from the top. What you're looking for is the combination of high keyword volume and low competition. Lots of people searching it, not many videos fighting over it. That's the pocket worth being in, and the scoring is there so you don't have to eyeball it.

Now the filter, because this is where people undo all of it.

Be specific. Every tag you add is a signal about who this video is for, and a tag that's too broad tells YouTube to show your video to people who were never going to want it. They see it, they don't click, or worse they click and leave in eight seconds. Either way you've taught the algorithm that your video doesn't perform. A generic tag doesn't get you a wider audience. It gets you a worse one.

So the test on every tag is the same: does this accurately describe what's actually in my video? If it's vaguely adjacent, or it's there because it's popular rather than because it's true, leave it out.

Specific and true, or not at all. That's the only rule that matters in this field.