Keyword Research for YouTube: How to Find Video Ideas People Actually Search For (Free Tools First)

Most advice on keyword research for YouTube tells you to copy whatever big channels rank for, which works if you have 500,000 subscribers and a production budget. For everyone else, it is a recipe for publishing videos that get buried on page three.

The better move is targeting specific long-tail searches where the existing videos are old, off-target, or low quality. Demand you can actually capture, not demand you admire from a distance.

What keyword research for YouTube actually is (and how it differs from Google keyword research)

Keyword research for YouTube means finding the exact phrases people type into YouTube search, then matching your videos to those phrases. It sounds like web SEO, but the mechanics differ enough that copying a Google workflow will mislead you.

First, YouTube has no reliable public search volume data. Google Keyword Planner gives you numbers for web searches. YouTube gives you nothing equivalent, so every volume figure in a third-party tool is an estimate built from sampling. Treat those numbers as directional, never as fact.

Second, viewer intent on YouTube is "watch," not "read." Someone typing "how to fix a leaky faucet" into Google might want a quick answer. The same person typing it into YouTube wants to see the fix happen. That changes which keywords are worth targeting: phrases implying a visual demonstration, a walkthrough, or a review carry stronger YouTube intent than dry definitional queries.

Third, the YouTube search algorithm weighs watch time and click-through rate alongside keyword relevance. A video that perfectly matches a query but loses viewers in the first 30 seconds will drop. A video that loosely matches but holds attention will climb. So keyword selection comes down to topic-demand fit: pick a phrase you can genuinely satisfy, not just one you can rank for on paper.

Start with YouTube Autosuggest and the search results page itself

Type a seed phrase into the YouTube search bar and do not press enter. The autocomplete dropdown is the fastest free source of real YouTube search data that exists. Every suggestion in that list is a phrase people have actually typed, ranked roughly by popularity.

The alphabet trick extends this. Type your seed phrase, then add a space and the letter "a." Note the completions. Replace "a" with "b." Work through the alphabet. Each letter surfaces a different cluster of long-tail video keywords that share your seed but diverge in intent. A creator in the meal-prep space typing "meal prep a" gets "meal prep air fryer recipes" and "meal prep and storage containers." Typing "meal prep b" gets "meal prep breakfast ideas" and "meal prep bowls." Each of those is a separate video idea with its own demand.

Modifier words work the same way. Prefix your seed with "how to," "what is," "why does," "can you," or "best" to expose informational and comparison intent. "How to meal prep" is a different audience from "best meal prep containers," and autocomplete will show you both.

After mining suggestions, look at the actual search results page for each candidate. This is where keyword research for YouTube videos gets real. Check three things: the age of the ranking videos, their view counts relative to the channel's subscriber count, and whether the videos genuinely answer the query or just happen to rank anyway.

A keyword where the top results are three years old, filmed on a phone, and only loosely related to the query is a keyword with weak competition. That is your opening.

Use YouTube Studio's Research tab and your own analytics

YouTube Studio has a Research tab (found under Analytics) that shows the top searches your existing audience runs on YouTube, plus content gaps YouTube has identified for your channel. A content gap means viewers searched for something, did not find a satisfying result, and YouTube flagged it as an opportunity. This is demand data from the platform itself, not a third-party estimate.

The catch: the Research tab requires a minimum level of channel activity to populate. If your channel is very new or very small, the tab may appear empty or not appear at all. In that case, skip ahead to the Trends validation step and come back once you have a few dozen videos published.

The other goldmine inside YouTube Studio is the Traffic Source: YouTube Search report. Open Analytics, click Traffic Source, then YouTube Search. You will see the exact search terms already bringing viewers to your existing videos. These are phrases you already rank for, which means expanding around them is the lowest-risk growth move available.

If a term like "budget meal prep for one" is sending you 40 views a month, that tells you two things. The phrase has demand, and YouTube already considers your channel relevant to it. A dedicated video targeting that exact phrase, or a close variation, is far more likely to rank than a video on a topic where you have no track record.

Validate demand with Google Trends (YouTube filter) before you commit

Google Trends has a YouTube search filter that most creators never touch. Go to Google Trends, type your candidate keyword, and under the "Web Search" dropdown, switch to "YouTube Search." Now you are looking at relative interest over time for that phrase specifically on YouTube, not on the web.

The absolute numbers are indexed to a 0-100 scale, so they do not tell you raw volume. What they tell you is direction and seasonality. A keyword trending upward over 12 months is gaining demand. A keyword that spikes every January and flatlines by March is seasonal, which is fine if you plan for it, but a problem if you expected year-round traffic.

The comparison feature is where this earns its time. Load three to five candidate keywords into the same Trends view and compare them head-to-head. You will often find that the phrase you assumed was biggest is actually the weakest, or that two phrases you thought were interchangeable have very different trajectories.

For the full walkthrough of reading Trends data for video topics, the dedicated guide on Google Trends for YouTube covers the filtering and comparison workflow in detail.

Steal intelligently: competitor and outlier analysis

Pick three to five channels in your niche that are 1.5 to 5 times your subscriber count. Channels in that range are big enough to have data but small enough that their wins are still replicable. Sort their videos by "Popular" and look for outliers: videos that dramatically outperformed the channel's average.

A video with 500K views on a 50K-subscriber channel is a 10x view-to-subscriber ratio. That kind of overperformance almost always means the video is pulling in search traffic or getting recommended beyond the channel's existing audience. Either way, it signals demand for that topic that the channel's subscriber base alone cannot explain.

Note the exact title phrasing of those outliers. Not the topic, the phrasing. "I Tried Meal Prepping for 30 Days" and "Meal Prep for Beginners: 5 Easy Recipes" might cover similar ground, but they target different searches and different intents. The title tells you which query the video is actually capturing.

The mistake to avoid is copying head terms that big channels own. "Meal prep" as a keyword is dominated by channels with millions of subscribers and years of authority. You will not outrank them. But "meal prep for night shift nurses" or "gluten free meal prep under $50" are long-tail variations those same channels ignored because the audience is too small to matter at their scale. At your scale, it is exactly the right size.

When to use tools like vidIQ and TubeBuddy (and what the free versions actually give you)

Both vidIQ and TubeBuddy offer keyword research features built around a similar model: enter a topic, get related keywords with estimated search volume and a competition or opportunity score. vidIQ shows volume and a competition score. TubeBuddy's Keyword Explorer combines volume, competition, and channel strength into a single weighted score.

The scores are useful for sorting a long list quickly. They are not useful as precise measurements, because neither tool discloses exactly how the numbers are calculated. The volume figures are estimates derived from sampling, and the competition scores are proprietary formulas. Treat them as rough filters, not verdicts.

For a beginner, the free tiers of both tools are enough. You get a limited number of keyword lookups per day, which is fine when you are researching one video at a time. The paid tiers add bulk lookups, historical tracking, and more granular competitor data, which matter once you are publishing several videos a week and managing a backlog of ideas.

The honest position is that tools accelerate the workflow described above. They do not replace it. Autosuggest mining, SERP quality checks, Trends validation, and competitor outlier analysis all work without spending anything. A tool that gives you a score of 74 out of 100 is helpful. A tool that replaces your judgement about whether the current ranking videos are actually good is not.

Score and prioritize: a simple keyword shortlist system

After working through Autosuggest, Studio, Trends, and competitor analysis, you will have more candidate keywords than you can film. The scoring step turns a brain dump into a ranked list.

Use three criteria, scored roughly:

• Demand signal. Does the phrase appear in Autosuggest? Is Trends flat or rising? Did a competitor's video on this topic outperform their channel average? More yes answers means stronger demand.
• Competition weakness. Are the current top results old, off-target, low production quality, or from channels that no longer post? Weakness in the incumbents is your opportunity.
• Channel fit. Can you make a genuinely good video on this topic? Do you have existing videos that rank for related terms? A keyword you cannot serve well is a keyword that will hurt your click-through rate and audience retention even if it ranks.

Score each keyword informally against those three criteria and rank your list. Aim for a shortlist of 10 to 15 keywords, ordered by combined strength. That is your content calendar for the next two to three months.

One keyword per video is the default. If two or three keywords are close variants of the same intent ("meal prep for beginners" and "easy meal prep for beginners"), they can share a single video. If they represent different intents ("meal prep for beginners" versus "meal prep container reviews"), they need separate videos. Mapping distinct intents to distinct videos is how you build topical coverage without cannibalizing yourself.

Put the keyword to work: title, description, and the first 30 seconds

Keyword research for YouTube only pays off when the keyword shows up where the algorithm and the viewer both look. That means three places, in order of importance.

The title carries the most weight. Use the exact-match phrasing where it reads naturally. "Meal Prep for Beginners: 5 Recipes Under $3" targets the phrase directly. "My Sunday Routine" targets nothing. If the exact phrase sounds awkward, get as close as natural language allows rather than forcing it.

The description matters most in its first two lines, because that is what appears in search results and what YouTube's system reads first. Include the keyword in a natural sentence within the first 150 characters. The rest of the description can expand on the topic, list timestamps, and link to related videos.

Say the keyword out loud in the first 30 seconds of the video. YouTube's auto-captions are indexed, so the spoken word carries ranking signal. It also confirms to the viewer that they clicked the right video, which protects your audience retention in the window where most drop-off happens.

Tags matter far less than they used to. YouTube's own help documentation states that tags play a minimal role in discovery compared to title, description, and viewer behavior. Fill them in with your keyword and a few variants, but do not spend more than two minutes on them.

The same instinct that finds winning YouTube keywords, starting from the real questions your audience asks, now applies beyond YouTube. Buyers are typing those same questions into ChatGPT and Perplexity, and the brands that get recommended there are the ones whose content already answers them. If you want to see where you stand, AnswerRank's free GPT SEO Checker runs your brand against live AI answers and shows whether you come up. You can also track your AI search rankings over time as you publish.

Keyword research for YouTube is not a one-time setup. It is a repeating cycle: mine, validate, score, film, check what ranked, and mine again. The creators who treat it as a system, not a guessing game, are the ones whose channels compound.