YouTube keyword research helps you find the questions people search for, understand the video they expect, and decide which topic you can cover well. Start with YouTube’s search suggestions and your own channel data, then use trends and keyword tools to narrow your choices.
I’d aim for one useful video with a clear promise. A long keyword list won’t help much if you’re still unsure who the video is for or what it needs to show.
In this guide, I’ll walk you through finding ideas, checking demand, judging competition, and turning a topic into a video brief. I’ll also explain which numbers come from YouTube, which are estimates, and which belong to Google Search.
Key takeaways
- Use YouTube Studio → Analytics → Trends for topic research, where available. Older tutorials may call this the Research tab.
- Your video’s YouTube search terms report shows searches that brought viewers to your content. It doesn’t provide the total search volume or your universal ranking for each phrase.
- In Google Trends, choose YouTube Search when you want to examine YouTube interest. Its 0–100 index is relative interest.
- Treat third-party volume and competition scores as research aids. Check their platform, location, time period, and definitions before comparing them.
- Choose a specific viewer problem, inspect competing videos, and make the title, thumbnail, and opening deliver the same promise.
What is YouTube keyword research?
YouTube keyword research is the process of finding the language viewers use when looking for videos and connecting those searches to content you can make.
A keyword might be a broad subject such as “home studio setup.” It can also be a specific question such as “how to reduce echo in a home studio.” Those searches need different videos, even though they belong to the same topic.
Why keyword research matters
Research gives you a better starting point for your script. You can identify the viewer’s problem, the words they recognize, and the information they’ll need to act.
For me, the most useful output is a sentence: “This video helps this person do this thing.” If you can’t finish that sentence, spend a little longer choosing the angle before recording.
How keyword research fits your channel’s growth
Search is one way people discover videos. Home recommendations, Suggested videos, and the Shorts feed serve different viewing situations. A search-led tutorial can succeed by answering a narrow problem, while an entertainment video may depend more on being recommended.
YouTube says its search results consider relevance and engagement, including how the title, description, and video match a query. A popular phrase alone won’t guarantee a ranking, a view target, or channel growth.
Decide what success would mean before choosing keywords. A useful tutorial might bring a smaller, highly relevant audience. A broad topic may attract more people but give them little reason to watch another video on your channel.
Methods for finding YouTube keywords
Write down three to five subjects you can explain or demonstrate credibly. Then collect specific questions from comments on your videos, recurring customer questions, your own search reports, and YouTube suggestions.
You can keep this in a simple document. Record the phrase, where you found it, the date, the intended viewer, and what you’d show on screen. Add numbers only when you know what they measure.
Use YouTube Analytics for native audience insights
On desktop, open YouTube Studio’s Trends tab through Analytics → Trends. Explore the available cards or enter a topic in the search bar.
YouTube began gradually rolling out an updated Studio experience in July 2026. The layout can differ, and some insights depend on country, language, device, or available data.
Available cards can include top searches, breakout videos, recent videos, and content gaps for Shorts. Several insights use your audience, saved searches, and activity over the previous 28 days.
I’d use these suggestions to build a shortlist. Ask whether each idea belongs on your channel and whether you can make the promised video. An interesting topic can still be a poor fit for your audience.
For searches that already brought viewers to a particular video, use Studio → Content → select the video’s Analytics → Reach. Find the YouTube search terms report. This is a different job from exploring future topics in Trends.
YouTube’s Reach documentation explains these reports. The terms describe how viewers found your video; they don’t list every keyword it could appear for.
Use YouTube autocomplete to collect wording
Type a seed topic into YouTube’s search box and pause before submitting it. Note suggestions that match a question your audience would reasonably ask.
Try useful modifiers, such as “for beginners,” “on a budget,” “without,” or the name of a device or app. For a home-recording channel, “microphone setup” and “microphone setup for a noisy room” lead to quite different briefs.
YouTube explains that predictions reflect factors including popularity, similarity, and, depending on settings, past viewing or searches. Their order isn’t a published search-volume ranking.
Keep the promising ideas, then inspect the results. A suggestion gives you wording to investigate. It doesn’t establish that the topic is easy to rank for or suitable for your channel.
Use vidIQ or a keyword tool for comparison
vidIQ’s keyword tools offer related phrases, estimated search volume, competition scores, and trend information. Its browser extension can surface keyword information while you browse YouTube.
I’d use those metrics to compare a small set of related ideas within the same tool. Keep the location and time window consistent, and inspect the search results before choosing a winner. vidIQ describes its volume figures as estimates.
You can also use an autocomplete-based tool to collect more suggestions. Keyword Tool’s YouTube page provides a keyword field and country/language controls. Check which features require a paid plan before building your workflow around them.
Kwebby’s overview of YouTube keyword tools can help you shortlist products to investigate. Confirm current features, pricing, and whether each metric describes YouTube or Google before subscribing.
For examples of turning a seed topic into related phrases and questions, see Kwebby’s vidIQ keyword research walkthrough. Its screenshots and example numbers reflect an older version; check the current controls and metric definitions in vidIQ.
Use Google Trends with the YouTube Search filter
Open Google Trends, explore a term, and change the search-type selector from Web Search to YouTube Search, as shown in Google’s search-type tutorial. Choose your audience’s country and a useful date range.
Start with a longer period to spot seasonality, then inspect a shorter period for recent movement. Compare closely related terms under identical settings. Keep search terms and broader topics clearly labeled because they can cover different sets of searches.
Google Trends normalizes its data on a 0–100 scale. A value of 100 marks peak relative interest within the selected comparison. It doesn’t mean 100 searches, 100,000 searches, or a guaranteed audience size.
Low-volume searches can appear as zero. Treat an empty chart as insufficient information rather than proof that nobody cares about the problem.
A seasonal topic needs enough lead time to research, record, edit, and publish. I’d rather make a useful video before a predictable busy period than rush a weak one after the spike.
Identify a content gap you can actually fill
YouTube describes a content gap as a search where viewers lack enough suitable results. Missing, outdated, or low-quality answers can all contribute. Its current Trends guidance specifically includes gaps for Shorts.
For your own research, finish this sentence: “Existing videos explain X, but my viewer still needs Y.” That missing piece should be something you can demonstrate, explain more clearly, or bring up to date.
- Missing situation: the tutorials assume a quiet room, while your viewer records beside traffic
- Missing step: everyone shows installation, but nobody explains how to check the result
- Outdated process: a popular walkthrough uses menus that have changed
- Wrong level: the results assume knowledge a beginner doesn’t have
Check the videos before claiming there’s a gap. A smaller view count, an old upload date, or a low tool score doesn’t establish that the answer is poor.
Then consider whether your format fits. A quick before-and-after demonstration may work as a Short. A setup with several troubleshooting branches may need a longer walkthrough. Choose the length around the explanation you owe the viewer.
Evaluate search volume and keyword difficulty
Before sorting a keyword list by volume, check the source. “Searches,” “views,” “interest,” and “competition” can describe very different things.
- YouTube search terms
-
Queries associated with viewers finding your own content. These help you understand existing traffic.
- YouTube Trends insights
-
Native topic and audience signals for developing ideas. Availability and scope vary.
- Google Trends, filtered to YouTube Search
-
Relative search interest over the selected place and period. Use it for comparison and timing.
- Third-party YouTube keyword volume
-
A vendor’s estimate of demand. Record the source, location, date, and period alongside the number.
- Google keyword-tool volume
-
A different search market. A Google volume figure doesn’t establish how often a phrase is searched inside YouTube.
How to judge competition without a magic score
Open the results for the exact viewer problem and look at the videos competing for attention. Note the quality of the answer, the specificity of the title, the production required, and whether the information is still accurate.
Tool difficulty scores are proprietary. A score of 40 in one product needn’t mean the same thing as 40 in another. Advertising competition also measures a different contest from earning an organic video view.
A lower-volume phrase can be a good choice when it describes a problem you can solve especially well. I wouldn’t impose a universal minimum monthly-search threshold on a small channel or a specialized tutorial.
If tools disagree, keep the disagreement in your notes. Check the platform and settings first. Then use the numbers as a rough comparison and let the viewer problem, visible results, and your ability to deliver guide the decision.
Understand what the viewer wants to watch
Search intent is the job behind the words. “Best microphone for YouTube” suggests a comparison. “Connect USB microphone to iPhone” suggests a demonstration. “Microphone not detected” suggests troubleshooting.
For each shortlisted phrase, identify the expected format, the viewer’s starting knowledge, and the point where they’d consider the question answered.
Read the leading titles and watch relevant parts of the videos. Look for a pattern, but leave room to make a clearer answer. If every result is a hands-on comparison, an opinion-only video may struggle to satisfy the same need.
I’d also list what the video should leave out. A tutorial about removing room echo doesn’t need a long history of microphones. That boundary helps you write a tighter opening and avoid burying the answer.
Create a keyword map and a usable video brief
Group phrases by the problem they describe. Close variants can belong to one video, while questions that need a different demonstration deserve separate briefs.
For example, “reduce echo in a room” and “stop room echo when recording” may fit together. “Remove echo from an existing recording” needs a different workflow and should be evaluated separately.
Worked example: choose an angle before recording
This is an illustrative planning example, not a live keyword-volume test. Imagine you make beginner home-recording tutorials and are considering these three ideas:
- “Home studio setup”: a broad overview requiring choices about room, microphone, recording software, and budget
- “Reduce room echo for voice recording”: a focused problem you could demonstrate with before-and-after audio
- “Best studio microphone”: a comparison that requires credible access to the products and a fair testing method
I’d investigate the echo idea first if you can record a clear comparison using materials your viewers already have. That is a production-fit decision. You’d still validate demand and competing answers before committing.
Your research notes might reveal two different audiences: people improving the room before recording and people repairing audio afterward. Keep those needs separate instead of forcing both into a vague “better audio” video.
Next, draft a brief with these fields:
- Viewer: a beginner recording speech in a room with noticeable reflections
- Main question: what can I change in the room before I buy more equipment?
- Working title: “Reduce Room Echo for Voice Recording: Simple Changes to Try”
- Opening: play a short untreated sample and explain what the demonstration will compare
- Evidence to show: the room arrangement, microphone placement, and comparable audio samples
- Useful limitation: room treatment, outside noise, and software cleanup solve different problems
- Next video: a separate tutorial for improving an existing recording
Fill the demand field with your actual sources and observations. If you don’t have reliable volume data, write “unknown” and record the other evidence. An honest blank is more useful than an invented estimate.
Keep research notes another person could follow
Save enough context to repeat the check. For a suggestion, record the seed phrase and the wording you saw. For a tool estimate, record the product, country, time period, and capture date. For a competitor video, keep its URL and the specific section you reviewed.
Separate observations from judgments. “Three videos show the same outdated menu” is an observation you can verify. “A new tutorial should be easier to follow” is your editorial judgment. Both can help, as long as you know which is which.
Also record what would change your mind. You might postpone the echo tutorial if you find a recent, clear demonstration that already answers the exact question. You might proceed if viewers repeatedly ask about a room layout that existing videos overlook.
That small habit prevents a spreadsheet full of keywords from becoming a list of assumptions. When you return to an idea later, you can see why it looked promising and which facts need another check.
Choose what to make next
Compare your shortlist using four questions: Does this matter to my audience? Can I show a useful answer? Is there a meaningful improvement I can make? Can I produce it well with the time and resources I have?
Use a simple high, medium, or low judgment if it helps. Add a reason beside each judgment. The explanation will be more useful than an unexplained total when you revisit the decision.
Put one video into production and keep the remaining ideas in a backlog. You don’t need to turn every keyword variation into a separate upload.
Analyze competitors without copying their work
Review both established channels and smaller channels serving the same viewer. Record what each video promises, how it demonstrates the answer, and what questions remain.
Compare videos at similar ages where possible. A five-year-old upload has had much longer to collect views than a video published last week. Public view totals also don’t reveal how much traffic came from Search, recommendations, ads, or an external audience.
Look beyond tags. YouTube says tags usually play a minimal role in discovery, although they can help with common misspellings. Titles, thumbnails, and descriptions matter more as metadata.
A useful competitor note might read: “The demonstration is clear, but it skips checking the audio levels.” That gives you something concrete to improve. “Copy this title and all its tags” gives you very little insight into why viewers watched.
Use your research in the video and its packaging
Once the angle is clear, write a title that tells the right viewer what they’ll get. Include the main phrase naturally when it helps that explanation.
The thumbnail should reinforce the same promise. If the video is a beginner setup, show the setup or the result. Avoid suggesting a dramatic transformation you don’t demonstrate.
Write the title, description, and opening together
Draft all three before filming. This is a quick way to spot a mismatch between the search question and the video you’re about to make.
In the description, summarize the specific lesson early, then add relevant resources, equipment details, chapters, or links. Use related wording where it helps explain the content, without repeating a list of near-identical phrases.
In the opening, establish the problem and begin delivering on the promise. You can use the viewer’s language naturally in the script. There’s no need to repeat a keyword on a timer or force it into every sentence.
Kwebby’s YouTube SEO guide also covers titles, thumbnails, and playlists. Use it for planning ideas, and follow the current YouTube Help links in this article for settings and requirements.
Check Google separately if it matters to your topic
Some questions also surface video results in Google. Search the phrase and inspect the actual results to see whether videos are useful for that query at the time you check.
Keep this as a separate note in your brief. A Google video result is an additional discovery opportunity, and its presence can change. It doesn’t establish demand inside YouTube.
Where Semrush Keyword Magic Tool fits
Semrush defines Keyword Magic Tool volume as average monthly searches on Google at the national level. Its keyword difficulty refers to website SEO, while its competition metric relates to advertisers.
Use that tool to investigate Google searches and video-related search features. If you use a separate YouTube-specific product, record its name and metric definitions rather than carrying the Google figures into your YouTube column.
Keep YouTube hashtags relevant
Hashtags link related content. Choose a few that genuinely describe your video, rather than adding an unrelated trending topic.
YouTube’s hashtag guidance says that more than 60 hashtags causes all hashtags on that content to be ignored. Excessive or misleading tagging can also have consequences. Treat that as a limit to avoid, not a target to approach.
Add chapters that help someone use the video
Use chapters when the lesson has distinct steps. Label each section by what the viewer will learn or do there.
For manual YouTube chapters, the description needs a first timestamp of 00:00, at least three timestamps in ascending order, and chapters at least 10 seconds long. Access and automatic-chapter availability can vary.
Write chapter names after the edit, when the timings are final. A keyword-heavy label that doesn’t describe the section makes the video harder to navigate.
Share where the video answers a real question
If a relevant community allows it, share the video with enough context for people to decide whether it helps them. A short explanation of the problem and result is more useful than dropping a bare link.
Record external promotion when reviewing performance. A burst of traffic after a newsletter or social post shouldn’t be mistaken for proof that a keyword started ranking.
Use vidIQ insights as a check on your brief
If a tool recommends another phrase, ask whether it describes the same lesson. You may improve the wording or discover a separate topic worth researching.
I wouldn’t rewrite an accurate title just to satisfy every suggestion. Keep a note of the recommendation, the change you made, and why it helped the viewer understand the video.
Track search traffic and viewer response
After publishing, return to the video’s Reach report and review YouTube search traffic and the search terms available. Use the Engagement reports to understand whether viewers stayed for the explanation.
A search-term entry is evidence that the query brought viewers to the video. It isn’t a position report. A missing term also doesn’t prove that the video never appeared for it.
YouTube limits some analytics data, including low-traffic search terms. Avoid treating an incomplete table as a complete list of demand or rankings.
Make a small, useful review log
Record the publication date, target question, title, thumbnail, format, and promotion. At your review, compare similar videos over comparable periods after publication and use the same metric definitions.
You might review an early period and then a later one, depending on how quickly your channel gathers data. These are planning checkpoints, not deadlines by which YouTube must rank a video.
- Little search traffic: revisit the topic, wording, timing, and competing results before blaming one metadata field
- Impressions with weak clicking: inspect whether the title and thumbnail communicate a useful, believable promise
- Clicks with early drop-off: check whether the opening delivers what the viewer expected
- Relevant traffic with satisfied viewers: investigate the next related question you can answer
Keep traffic sources separate when interpreting click-through rate and retention. A changing audience or source mix can change the averages even when the video itself hasn’t changed.
Manual searches and third-party rank trackers can provide additional observations. Results depend on context, so save the query, date, location, device, and method rather than treating one position as universal.
When you adjust the title or thumbnail, note the change. Avoid claiming it caused every later increase or decline; demand, competition, recommendations, and promotion may also have changed.
Frequently asked questions
Which is the best keyword tool for YouTube?
Start with YouTube’s search suggestions and your available Studio reports. Add a dedicated tool if you need larger suggestion lists, comparative estimates, or tracking. The best choice depends on the task, your budget, and whether the tool explains its data clearly.
How do I find YouTube keyword search volume?
Dedicated YouTube keyword tools can provide estimated volume. Keep the vendor and settings with each number. Studio search terms describe traffic to your content, while Google Trends provides relative interest. Neither should be presented as a universal monthly-search total.
What does a keyword or SEO score mean?
It’s a tool-specific assessment based on that product’s method. There is no universal “SEO IQ” score you need to reach before publishing. Read the score’s definition and use it alongside your own review of the topic and competing answers.
What is SEO on YouTube?
YouTube SEO is the work of making your video understandable and discoverable for the people it helps. That includes choosing a relevant topic, delivering a useful answer, and writing clear titles and descriptions with accurate supporting metadata.
How do I track YouTube rankings for free?
You can record manual search observations and use Studio to monitor search traffic. Keep those records separate: a search-term report measures how viewers found your video, while a manual search shows a position in one particular context.
What if I can’t see the Trends tab or a content-gap card?
Check the channel-level Analytics area in desktop Studio. Feature availability and the ongoing layout rollout can affect what appears. You can still collect ideas through autocomplete, audience questions, existing search traffic, and a review of competing videos.
Can the right keywords guarantee more views?
No. Keywords can help you choose and explain a topic, but viewers still need a reason to click and keep watching. Make a video that fulfills a clear promise, then use the results to improve your next decision.
Final thoughts
I’d begin with one question your audience genuinely needs answered. Collect the wording they use, check the available evidence, inspect the competing videos, and write a brief you can actually deliver.
Keep the data sources separate and the promise specific. That gives you a research process you can repeat, and a much clearer reason to make the next video.
