Keywords YouTube

I open YouTube and type my core topic into the search bar. That gives me autocomplete suggestions — the phrases real users are typing. For example, I typed "how to edit" and got "how to edit videos", "how to edit photos", "how to edit in capcut", "how to edit like a pro", "how to edit videos for beginners". I screenshot the list. Those are my seed keywords. Next, I search the most relevant one, say "how to edit videos for beginners". I examine the top five videos. Three have "beginner tutorial" in the title. Two have "step by step". One includes "2025 edition". I note those patterns. Then I use the alphabet method: I type "how to edit videos for beginners a", "b", "c", and so on. That gave me "how to edit videos for beginners adobe premiere" and "how to edit videos for beginners after effects". That is specific audience intent. I repeat for each seed keyword. After 15 minutes I had 40 keyword ideas.

I then filter. Broad terms like "video editing" I drop immediately because my own data from a [Keyword Rank Checker](/keyword-rank-checker/) shows that broad keywords push videos into a pool of millions. I look for intent markers: "for beginners", "2025", "with phone", "free". Those are my targets. I also check the search results for each candidate. If the top video has 50K views and the tenth has 200 views, the niche is real. If every top ten video is from channels with 100K+ subscribers, I reconsider — that signals high [Keyword Difficulty](/keyword-difficulty/).

When writing the title, I place the primary keyword as early as possible. For "how to edit videos for beginners capcut", I lead with that exact phrase. The description opens with the same phrase in the first sentence, then adds two related keywords naturally: "I walk you through trimming, transitions, and adding text in CapCut." I use the tags field for only one or two exact-match phrases that differ from the title, because tags barely count anymore. That is my path.

Do not stuff keywords into the title and description. I tested this on a video about vegan lasagna. I used the same three-word phrase three times in the title and twice in the first paragraph of the description. The click-through rate dropped because the title looked spammy. YouTube's algorithm also penalises mismatch between metadata and content. If I promise "easy vegan lasagna" but the video shows a complicated 40-minute recipe, viewers leave early, signalling irrelevance.

A second watch-out is relying on tags. I used to fill all 500 characters with every variation I could think of. That is wasted time. YouTube’s own documentation says tags are a minor signal. The real heavy lifters are title, description, captions, and watch time. I now spend that 20 minutes improving the first 30 seconds of the video to hook viewers.

Another trap: ignoring [Keyword](/keyword/) intent. On YouTube, "best smartphone 2025" is comparison intent. "How to set up a smartphone" is tutorial intent. If I make a tutorial video and title it "best smartphone 2025", I get clicks but low retention because the [SERP](/serp/) showed a different expectation. I check the top videos for the keyword: are they listicles? How-tos? Reviews? I match my video to what already works.

One more: starting with a tool before doing manual research. Tools like VidIQ are useful later but they pull from broad data. The YouTube search bar is where actual users type. I always start there. Tools can expand the list but not replace autocomplete.

Also do not overlook the description length. Short descriptions miss opportunities for related keywords that appear in [People Also Ask](/people-also-ask/) or the transcript. I write at least 150 words that include three to five target phrases.

I got tags wrong for years. When I started my channel in 2020, I read a blog post claiming tags were the second-most important ranking factor. I believed it. I spent 30 minutes per video typing every synonym and misspelling. My tags field looked like a thesaurus had thrown up. The videos still performed poorly. I ran a test: uploaded a video with no tags. Same title, same description. It got more views in the first week than my previous five tagged videos combined. I dug into YouTube Creator Academy and found that tags are a low-weight signal used mainly for spelling corrections. The algorithm relies on title, description, captions, and viewer behaviour. I changed my process overnight. Now I add only one or two tags to catch common misspellings, otherwise I leave the field empty.

Second mistake: choosing keywords based on search volume only. I saw "how to edit videos" had 200K monthly searches and thought that was my ticket. My video got buried. I did not account for competition. I later learned to use a Keyword Difficulty metric. For "how to edit videos for beginners capcut" the difficulty was low because the phrase was specific. That video now sits at position two for that term and drives steady traffic with only 1,200 monthly searches. That taught me to prioritise specificity over volume.

Third mistake: ignoring the alphabet method. I thought it was a gimmick. Then I tried it for "best budget headphones" and discovered "best budget headphones for running", "for gaming", "with microphone". Those three phrases had less competition and matched my content better. I now do the alphabet method for every core topic.

I also used to write two-sentence descriptions. I thought viewers did not read them. But the algorithm does read them. Short descriptions miss chances to include related keywords. I now write 200-word descriptions that include the main keyword in the first sentence and three to four related terms later.