Every "100 YouTube video ideas" post has the same problem: you use four of them and you are back where you started. Ideas are not the asset. Sources are.
Key takeaways
- Build sources that produce ideas weekly, not lists that run out.
- Validate demand before filming, not after.
- A video that beats its own channel average is worth more study than a video with a big number.
- Titles and thumbnails decide the click. The idea only decides whether the click was worth it.
The nine sources
1. Questions your audience already asks
Sales calls, support tickets, DMs, comments. Every question asked twice is a video. This is the highest-converting source and it costs nothing.
2. YouTube's own search suggestions
Type your topic into YouTube search and read what autocompletes. Those are real queries, ranked by how often people type them, given to you free by the platform.
3. Competitor videos that beat their channel average
Not their most-viewed video, which is usually old and lucky. The video from the last 90 days that outperformed that channel's own baseline. That gap is a signal about the topic, not about the channel size.
4. Your own comments section
The question asked under video 12 is video 20. Sort by newest and read fifty.
5. Search demand from Google
A term with real Google volume usually has YouTube volume too, and Google increasingly shows video results for how-to queries. Our post on search intent and keyword types covers how to read which queries want a video.
6. Formats that work in your category elsewhere
A carousel that performed on Instagram or a thread that worked on LinkedIn is a script that has already been tested on a smaller audience.
7. The thing you keep explaining internally
If you have explained the same process to three colleagues, that is a video with a proven need.
8. Objections from lost deals
"Too expensive", "we already have a tool", "we can build it ourselves". Each is a video, and each one converts.
9. Public data nobody has visualised
Ad libraries, public filings, pricing changes across a category. Original data is the rarest content type on YouTube and it earns links and mentions long after publishing.
Validating an idea before you film it
Two days of filming deserves twenty minutes of checking:
- Type it into YouTube search. No suggestions means no search demand. That is not a veto, but it changes the plan: you now need a title and thumbnail that create the demand.
- Check the top results. If every video is three years old and under 5,000 views, the topic is either dead or unserved. Both are worth knowing.
- Check ratio, not volume. A 40,000-view video on a 20,000-subscriber channel travelled. A 200,000-view video on a 4M channel did not.
- Check Google volume. Video results appear in Google for a lot of how-to terms, which doubles the return on the same script.
The titling rule
The idea does not decide the click. The title and thumbnail do.
Three rules that hold across categories:
- Promise a specific outcome, not a topic. "How to check competitor ad spend in 10 minutes" beats "Competitor research tips".
- Put the number where it is visible on mobile, which crops titles hard. Google applies the same truncation to title links in search results.
- Do not overpromise. A clickbait title inflates click-through and destroys watch time, and YouTube weighs the second one more.
Our post on copywriting for search results covers the same principle for titles that have to earn a click in a SERP.
Why this matters beyond YouTube
Video work now feeds AI visibility. Ahrefs' analysis of around 75,000 brands found YouTube mentions were the strongest correlate of brand visibility across ChatGPT, AI Mode and AI Overviews (Ahrefs, 2025). Assistants are trained on and cite transcripts, so a video is now a distribution channel into answers, not only into YouTube.
Our GEO playbook covers how to turn that into a system, and YouTube SEO rules covers the platform side.
Turning sources into a pipeline
Once a month, one hour:
- Pull ten questions from source 1 and 4.
- Check YouTube suggestions for each, keep the six with demand.
- Look at what competitors published in the last 90 days and which posts beat their baseline. Pengu Insights tracks that automatically from €39 a month, along with YouTube and social performance for a competitor set.
- Write six titles before six scripts. If a title is not compelling, the video will not save it.
- Batch-record. The calendar dies without batching, exactly as it does for social content.
Frequently asked questions
How do I come up with YouTube video ideas?
Build sources instead of brainstorming: questions your audience asks, YouTube search suggestions, competitor videos that beat their channel average, your own comments, and formats proven on other platforms.
How do I know if a video idea will get views?
Check YouTube suggestions, look at whether existing videos outperformed their channels' averages, and check search volume before filming.
What is YouTube keyword research?
Finding the phrases people type into YouTube search and checking competition. Intent skews toward how-to, review and entertainment, and volume differs from Google for the same phrase.
How often should I publish on YouTube?
Consistently rather than frequently. One a week sustained for a year beats three a week for a month.
Do YouTube videos help SEO?
Yes. They rank in Google results and video carousels, and YouTube mentions correlate strongly with brand visibility inside AI assistants.
Written by
Christian Monge
Founder of Pengu Insights. Competitive intelligence practitioner for DTC brands and marketing agencies.
