Tags used to be one of the main signals for how a video was matched to an audience. That has changed substantially, and if you are putting effort into optimising them it is worth understanding what they now do.
Where tags sit in the system
YouTube largely works out what a video is about from the title, the description and, most importantly, what people actually do after it appears. Someone clicks, watches a while and does not leave, and that behaviour is a stronger signal than anything you typed into a field.
Tags are a minor input by comparison. They are not worthless, because they are free and they occasionally help, but optimising them at the expense of the thumbnail, the title or the first thirty seconds of the video is optimising the wrong thing.
How to use them sensibly
Start from your main keyword and generate related phrases, then keep the ones that genuinely describe your video. Ten to fifteen is plenty, and most of the value is in capturing the vocabulary people actually search with rather than in the number.
The most useful habit is looking at what comparable videos in your niche use. Not to copy wholesale, but because their tags tell you the words your audience reaches for, which is difficult to guess from your own perspective.
Irrelevant tags do not help
Adding broad popular tags to a video about a narrow subject is a common tactic and it does not work. Tags that do not describe the video tell YouTube the video is about something it is not, and the mismatch affects who it is shown to. The audience it reaches is the audience that will not watch, and that shows up in the numbers afterwards.
Related is better than popular. Someone searching for your actual subject is worth more than a large number of people searching for something adjacent.
Where the effort belongs instead
If you are optimising one thing for discovery, make it the packaging: a title that names the problem precisely, a thumbnail legible at small size, and an opening that answers the question the title promised. Those three together do more for a video than any amount of tag work.
Try next
Related guide: free text to speech online
Where tags are worth the effort anyway
Despite the reduced importance, there is a small set of cases where filling the field in is worthwhile. New and low-view channels benefit from every weak signal available, because the system has less behaviour to work from. Channels in a niche with quiet or seasonal search see tags contribute where viewing data is thin. And channels that make series benefit from consistent tagging, because it helps the videos be associated with each other.
There is also a case in translation. Tags can carry a keyword in a language you are not targeting, which occasionally surfaces a video to an audience that searches in that language.
None of this makes tags a growth strategy. It makes them a five-minute task that occasionally helps and never hurts, which is a reasonable thing to do at the end of an upload rather than a project to plan around.
The genuine waste is the long copy-pasted list. Thirty tags of varying relevance adds nothing over five accurate ones, and the effort is better spent on the two lines of description, which do get read by both people and machines.
If you want to know whether tags made any difference, compare two uploads where everything else is similar and the tagging differs. That is a weak experiment, but it is better than assuming, and it takes one upload to run.