YouTube topic research: how to find video ideas worth making
A YouTube topic research method: sample videos in your niche, group them by topic, compare views per subscriber, and choose topics with evidence.
Free app. Runs on your Mac with the Claude or Codex plan you already have.
What YouTube topic research is
YouTube topic research is the work of finding out which subjects in a niche viewers watch, which are already crowded, and which questions no video answers well yet. Done well, it gives a short list of topics, each backed by a count of videos and how they performed. Done badly, it is a brainstorm of video ideas chosen because one video on the subject went viral.
View counts on their own mislead: a large channel gets views on almost anything, and an old video has had years to collect them. The method below is how an experienced YouTube strategist researches topics by hand, using public data. Each step says what to do and why, then how 10folds carries it out.
What it does for the business
- Spend production time on topics with demandA video takes hours or days to make. Choosing topics that have already pulled viewers in your niche gives each one a better start than a guess.
- Find what smaller channels can winComparing views against subscriber counts surfaces topics where small channels beat their usual audience. Those are topics where the subject itself drew viewers, so a new channel can compete.
- Avoid crowded subjectsSeeing how many videos already cover a topic, and how they performed, tells you where another standard video will struggle and where a different angle is needed.
- Plan a series, not single videosGrouping videos by topic shows clusters of related interest. A cluster suggests a run of videos that feed each other, instead of one-off uploads. A series also gives viewers a reason to watch the next video.
YouTube topic research, step by step
Define the niche and the sample rules
A comparison is only fair when every video in it was chosen the same way.
What to do
- Write down the niche and the viewer you are making videos for.
- Fix the language, the publication period and the format: Shorts and long videos behave differently.
- List the channels in your niche, from small to large.
- Decide how many videos you will sample per search or channel.
Why
Without fixed rules, a sample drifts toward the videos that are easiest to find, which are usually the most popular ones. A consistent sample makes the comparison between topics honest, and separating formats stops a Short with a million views from distorting long-video topics. Listing channels of different sizes matters too: a niche seen only through its biggest channels hides what works for everyone else.
Gather candidate videos
Start from what viewers search and watch, not from what you want to make.
What to do
- Search your main subject on YouTube and note the search suggestions and related searches.
- Open the top results and the channels behind them.
- Browse the popular and recent uploads of the channels on your list.
- If you have a channel, check the Research tab in YouTube Studio for what your audience searches.
- Keep the link for every video you note.
Why
Search suggestions and related searches reflect what viewers type, which is the core of YouTube keyword research. Channel uploads show what has worked for creators like you. Your own Research tab adds private search data that no public page shows. Keeping a link for each video means every figure can be checked later, when the numbers have moved.
Record each video and code its topic
Code the topic from what the video covers, not from a word in its title.
What to do
- For each video, record title, link, views, publication date, format and the channel's subscriber count.
- Write the date you observed the numbers.
- Assign each video a topic in plain words, based on the title, description and content.
- Leave a figure empty if the page does not show it.
Why
Public numbers change daily, so the observation date matters. Coding the topic by hand groups videos that cover the same subject under different titles, which a word match would split. An empty field is honest; a guessed subscriber count is not. When a video covers two subjects, code it under the one most of its runtime is about, so each video is counted once.
Compare topics fairly
A topic with two strong videos may be luck; one with ten is a pattern.
What to do
- Count the videos in each topic.
- Take the median views per topic, which a single viral video cannot skew.
- Compute views per subscriber for each video to find outlier videos that beat their channel's usual reach.
- Compare videos of similar age and the same format.
- Show the sample size beside every figure.
Why
Views per subscriber is a simple way to adjust for channel size. A video with ten times its channel's subscribers in views shows that the topic pulled in new viewers. The median protects against one huge hit, and the count tells you whether a pattern is real. Age matters because views accumulate: a video from last month and one from three years ago should not be ranked side by side.
Validate the topic and plan a test
The questions viewers still ask in the comments are the next videos.
What to do
- Read the comments on the strongest videos and list the unanswered questions.
- Look for a content gap: a question that existing videos answer poorly, or an angle nobody has taken.
- Recheck the top figures against the video pages.
- Choose one or two topics to test, with the angle and format.
Why
Strong numbers show interest, but they do not prove viewers lack a good answer. Comments show where existing videos fall short. Testing a small number of topics first lets your own analytics confirm the choice before you commit to a series. Write down what you expect each test video to achieve, so you can judge it fairly against your channel's usual results.
Common mistakes
- Ranking topics by raw viewsBig channels get views on anything; adjust for channel size.
- Mixing Shorts and long videosThey are watched differently, so compare each format on its own.
- Comparing new and old videosA five-year-old video has had far longer to collect views.
- Trusting one viral hitCount the videos behind a topic before betting on it.
- Copying a video's angleA copy competes with the original on its own ground; answer the question it left open instead.
What a good result looks like
- A clear niche, language, period and format.
- One row per video with views, subscribers, date, format and link.
- Topics coded in plain words, with the number of videos in each.
- Median views and views per subscriber, each with its sample size.
- A short list of topics to test, with the evidence date.
And then what? Using the result
- Pick one or two topics and plan a video with a distinct angle.
- Draft titles that answer the questions found in comments.
- Check your own analytics after publishing to confirm the topic worked.
- Run the same research again with the same scope to see how topics move.
When to run it
- Starting a new channelIt shows which topics in your niche let small channels reach beyond their audience.
- When views on your channel stallComparing topics across your niche shows whether the subject or the execution is the problem.
- Planning a quarter of contentA topic table gives you a ranked list to schedule, with the evidence behind each choice.
- Entering a neighboring nicheSampling the new niche on the same rules shows which of its topics suit your channel before you change direction.
Questions people ask
How do you do YouTube topic research?
Define your niche and a consistent sample, collect videos from YouTube search and competitor channels, code each by topic, then compare topics by video count, median views and views per subscriber. Finish by reading comments for unanswered questions.
How do I find YouTube video ideas that get views?
Look for topics where several videos, especially from small channels, beat their channel's usual views. Then read their comments for questions still open, and make the video that answers them.
What is an outlier video on YouTube?
An outlier video is one that gets far more views than its channel usually does, often measured as views per subscriber. It suggests the topic or angle itself drew viewers.
How do I find content gaps on YouTube?
Read the comments and questions under strong videos and note what they leave unanswered. Public view counts alone do not prove a content gap.
Does 10folds use my YouTube analytics?
No. It reads public channel and video pages only. Search volume, click-through rate and retention need your own analytics.
Does it see every video on a topic?
No. It works from an observed sample of up to 25 sources and states the number of videos behind each figure. A topic with few videos in the sample is a lead to check, not a conclusion.
More jobs in Research
Research YouTube topic opportunities, on your own Mac.Free app. Runs on your Mac with the Claude or Codex plan you already have. Your files stay on your computer.
Download for Mac