Paid keyword tools sell you a spreadsheet of last month’s averages. Viewers type their questions today. The free method below reads that live demand directly — the same autocomplete data every big channel’s research ultimately rests on — and turns it into a title, a description and a plan in about twenty minutes.
Why is YouTube keyword research different from Google?
Direct answer: intent. Google serves readers; YouTube serves watchers. “Best budget mic” on Google wants a comparison table; on YouTube it wants to hear the mics. Phrases with watchable intent — how to, versus, test, honest review, day in the life — carry the demand, and phrasing that wins on one platform can be a ghost town on the other.
This is why copying keywords from a blog-SEO tool into a video title underperforms: the numbers were measured on the wrong audience. Research where the viewers are, in the language they use when they intend to watch.
What is the free autocomplete method, step by step?
Direct answer: seed → expand → validate → choose. Type one seed phrase, harvest every real suggestion around it (prefixes, questions, a–z variations), confirm which suggestions have live demand, and pick one primary phrase whose competition you can actually beat.
- Seed. One plain phrase for the video idea — “home workout”, “resume tips”, “budget mic”.
- Expand. The free Keyword Research Tool fires ~110 expansion queries at live autocomplete — the seed alone, question words in front, buyer modifiers, every letter a–z — and streams back the real phrases people type, deduplicated and scored.
- Validate. Sort by popularity, then check the intent labels: question phrases feed tutorials, versus phrases feed comparisons, long-tail (4+ words) phrases are your low-competition wins.
- Choose. One primary phrase per video. Confirm its placement plan with the free Focus Keyword Checker once the page or video description is drafted.
Twenty minutes, zero rupees, and you exit with a ranked list instead of a hunch — plus the tool’s Phase 1/2/3 plan for which phrases to attack this week versus this quarter.
How do you judge competition without paid tools?
Direct answer: read the results page like a scout. Search your phrase on YouTube and note three things — the view counts of the top results, the size of the channels holding them, and the upload dates. Small channels with old videos ranking = beatable; giants with fresh uploads = pick a longer-tail phrase.
The tell that matters most is mismatch: a top result with 200k views from a 5k-subscriber channel means the phrase itself carries demand — the video is ranking on relevance, not channel power. That is your green light. Conversely, if every result is a mega-channel, the phrase is not wrong, just premature; note it for the quarter you have more authority and take its four-word cousin today.
How do you pick keywords for Shorts specifically?
Direct answer: shorter queries, trend-adjacent phrasing, and topics that resolve in under sixty seconds. Shorts search skews to two-to-four word phrases (“desk setup ideas”, “excel tricks”), and hashtag feeds double as a live index of what phrasing your niche browses.
A workflow that compounds: when your long-form research surfaces a question answerable in 45 seconds, that is a Short — same primary phrase, condensed. The Short’s first description line uses the phrase verbatim, links to the full video, and shares its series hashtag (placement rules in the hashtag strategy guide), so the two rank as a pair instead of competing.
Where do the keywords go once you have them?
Direct answer: four places, in priority order — the title (primary phrase, early), the first 150 characters of the description (phrase + promise), your spoken first thirty seconds (YouTube transcribes everything), and the tags field (phrase + variants, two minutes maximum).
Each placement has its own guide and tool: draft title angles with the free SEO Title Generator (formulas in the titles guide), structure the description with the descriptions template, and fill the hidden field with the YouTube Tag Generator’s ranked list. The spoken placement costs nothing except remembering to say the phrase naturally while you still have the viewer’s first minute.
Which keyword research mistakes waste the most effort?
Direct answer: chasing raw volume over beatable competition, researching after recording, targeting five phrases with one video, treating research as one-and-done, and ignoring what the results page is plainly telling you about who currently owns the phrase.
The after-recording mistake hurts twice: the phrase you find rarely matches the words you said on camera, so title and transcript disagree. Research first, say the phrase on camera, and the whole metadata stack reinforces itself. Then let the free YouTube SEO Checker audit the result across your recent uploads.
How do you turn one seed into a month of videos?
Direct answer: cluster the expansion list. One good seed returns 80–120 real phrases; group them by intent — questions become tutorial videos, versus phrases become comparisons, “for beginners / for X” phrases become audience-specific remakes — and a month’s calendar falls out of a single research session.
The export makes this practical: pull the CSV from the Keyword Research Tool, sort by intent label, and mark each phrase as long-form, Short, or both. A cluster of eight related phrases is stronger than eight random ideas, because every video you publish in the cluster makes the others easier to rank — YouTube learns your channel owns the topic.
Sequence inside a cluster follows competition: long-tail first (wins fastest, builds topical authority), mid-tail in month two, and the head term last, when the cluster’s watch time is already vouching for you — the same Phase 1/2/3 logic the tool prints under every search.
What does a real research session look like, start to finish?
Direct answer: twenty minutes, four artifacts. Minute 0–2: seed typed, expansions streaming. Minute 2–10: sort, star 10–15 phrases, read the results page for the top three. Minute 10–15: pick one primary + variations, draft three titles. Minute 15–20: description opening written, tags generated, phrases logged in the cluster sheet.
The artifacts are the point — a chosen phrase, three title drafts, a description opening and a tag list — because they flow straight into the upload checklist from our upload guide. Research that ends in a spreadsheet nobody opens again is entertainment; research that ends in artifacts is production.
Frequently asked questions
Run the method right now
Open the free Keyword Research Tool, type one seed phrase, and watch a hundred real queries stream in — with popularity, intent and a phased plan attached.