YouTube Comment Scraper: From Comments to a Video Brief
Use a YouTube comment scraper to collect a sample, tag the questions it contains, and choose one piece of content to test. Save the video link and retrieval date with your notes so you can trace the idea back to its source.
The useful result is a brief you can act on: what someone struggled with, what you could explain, and what remains uncertain. A comment can suggest a lesson. It cannot tell you how everyone who watched the video felt.
Read beyond the preview
On September 11, 2026, we ran The Spawn's comment tool on freeCodeCamp.org's course, Claude Code for Beginners Tutorial [Full Course]. The request returned 20 comments through Treg via ScrapeCreators. The interface showed 10 comments; its JSON export includes all 20 received comments.
Our manual reading found four comments discussing prerequisite knowledge or difficulty following a course presented for beginners. We also found one comment requesting dubbing because English was a barrier. These are paraphrased observations from that page, with no commenter profiles reproduced.
Three of the four prerequisite-related comments, plus the language request, were outside the preview. Reading only the visible rows would have missed them. Downloading the received page mattered before we chose a topic.
This run fetched one page. The response included next_cursor, but we did not retrieve another page, and the sampling order was unknown. The observations do not establish how common these concerns are across the audience or whether the course is effective overall.
Tag what the comment actually asks
Copy a worksheet like this into your notes or spreadsheet. Give each comment a local row reference so you can revisit it in the saved JSON. Keep a paraphrase beside the tag; a category without the underlying meaning is hard to check later.
| Tag | Observation in this page | Question for a content brief |
|---|---|---|
| Prerequisites | Four comments discuss prior knowledge or beginner difficulty | What should a learner already know before starting? |
| Language access | One comment asks for dubbing because English is a barrier | Which language or format would help this person follow the lesson? |
For your own video, add columns for the row reference, proposed response, and unresolved detail. You can add tags such as technical obstacle or general feedback when the comments warrant them. Those are worksheet options, not extra findings from our example.
Allow a comment to carry more than one tag. Count the comments within each category without treating the categories as a partition of the audience. If someone says they are lost without naming a step, record that missing detail. Do not invent a software error, missing skill, or reason they stopped watching.
Before selecting a topic, reread the comments behind it. A request for a prerequisite explanation and a request for a slower demonstration might need different answers, even if both initially received a beginner-difficulty tag.
Choose a small lesson to test
For a creator planning a similar tutorial, I would first draft a prerequisite check and a small worked example. The proposed title could be Before your first Claude Code project: check what you need to know. That choice follows a specific uncertainty in this page: what knowledge the lesson assumes.
The draft brief would contain the prerequisites, a way for the learner to check each one, and a small task showing how those skills enter the workflow. Verify the actual course requirements before filling in the checklist. The comments alone cannot supply them.
Try the draft with people who describe the same difficulty. Record whether they can finish the worked example and which instruction stops them. Ask what they expected to know before starting. If they remain blocked at the same point, revise that explanation before expanding the lesson. Keep the original comments and these new observations separate in your notes.
This is a proposed experiment. We have not tested whether it improves completion or retention. The language request deserves its own follow-up question before committing to dubbing; one request does not establish demand for a translation project or willingness to pay for one.
Check the collection method before making broader claims
The Spawn currently retrieves a bounded page through Treg's YouTube comments endpoint. The web flow provides JSON export of what it received. It does not offer a CSV export, automatic tagging, or automatic collection of every comment and reply.
If your research needs controlled ordering or additional pages, YouTube's official API documentation describes time or relevance ordering and pagination for comment threads. Those documented API options do not establish the order used in our Treg response. Assess collection coverage before using a larger dataset to compare videos. Disabled or unavailable comments can also leave you without data; that does not answer whether viewers have questions.
Start with one public video in The Spawn's YouTube comment research tool. Preview the request, approve the run, and download the received JSON before completing your worksheet. For an assistant workflow, connect MCP at https://thespawn.io/mcp. Use spawnr_tools to find the operation, then spawnr_tool_quote and spawnr_tool_run after approval. Both paths support the same bounded task: gather the comments you received and decide what to investigate next.