Youtube Comment Scrapper
Pricing
$1.80 / 1,000 comments
Youtube Comment Scrapper
Scrape comments on Youtube Video Pages. Just specify the video page urls to visit and how many comments you would like to collect. We'll get all meaningful information: comment text, author, publication date, number of likes and replies and video+channel url and id
5.0 (1)
Pricing
$1.80 / 1,000 comments
14
Total users
415
Monthly users
44
Runs succeeded
>99%
Issues response
2.3 hours
Last modified
a month ago
I’m planning to scrape 50k+ comments/video across multiple videos.
Open
Could you clarify:
- The max comments/video and how to avoid interruptions?
- Best practices for input configuration to handle consecutive runs?
- How checkpointing/resumption works for long jobs?
Hi,
I would suggest you avoid long runs. Better to multiply short runs than having a long run failing because of a timeout or whatever.
It also depends what you want to achieve, some users wants to remain updated regarding new comments/new reactions on video. In such case, a good pattern is:
- scrap channels to collect only the video details...
- it depends a bit how far you want to go in the past, but 10 channels per run is probably a good configuration
- for each video collected in 1), scrap the comments (you can set the limit on comments regarding what you've already collected in the past)
- it depends on the limit, but 10 videos per run is probably a good configuration
You need to keep track of what you've already collected and when, so you can possibly iterate and restarting from step 1) with an appropriate dateFilter, to collect only the new videos and or update the videos details up to a certain point. I typically stop tracking comments/reactions 2 weeks after a video has been posted. But that depends on your goal.
You will have failing runs... or runs with empty results... Unfortunately, I can't do much about it, sometimes youtube is just not answering or not the way the scrapper expects it. Shuffle your list of video/channel to scrap to avoid the same videos are always scrap within the same run, so you minimize the risk of missing data because of a failing run.
Other advice, start small, make sure that everything works well and then continue.
Feel free to reach out in case of needs, I am here to help (as much as I can)
Cheers,
