Hashtags vs tags on YouTube: what actually matters

They sit in different boxes, do different jobs, and are constantly confused with each other. This guide separates them — and shows how to build both with our free Hashtag Generator.

Two different systems

Hashtags are the clickable blue links that appear above your title. They live in the video's description (or title), viewers can see and click them, and each one leads to a feed of videos using the same hashtag. Tags are invisible metadata in the Studio "Tags" field. Viewers never see them; they exist to help YouTube understand the video's topic — including common misspellings and alternate phrasings of it.

Neither is a primary ranking factor in 2026 — watch time, click-through rate and satisfaction drive distribution. But both are free, both take two minutes, and both help YouTube place a video in front of the right first audience, whose behaviour then teaches the system who the video is for. That first audience is why metadata still matters, especially for smaller channels without a large subscriber base to seed the data.

Hashtags: three good ones beat fifteen random ones

YouTube shows the first three hashtags from your description above the title. Use more than fifteen and YouTube ignores all of them. The practical recipe: three to five hashtags total, chosen like categories a viewer would actually browse — the topic (#dyatlovpass), the broader niche (#historyfacts), and the format or angle (#unsolvedmystery). Put them at the end of the description, not the start: the first lines of a description are valuable space for the summary and links a human reads.

Hashtag mistakes to avoid: using a competitor's channel name as a hashtag (it can be removed and looks spammy), tagging a trending topic the video isn't about (the clicks it attracts leave immediately, which hurts the video), and repeating the same hashtag set on every video regardless of topic — each video should earn its own set.

Tags: exact phrases, natural variants, misspellings

You have 500 characters of tag space. Spend it on phrases, not single words: "dyatlov pass incident", "dyatlov pass explained", "what happened at dyatlov pass" — the actual strings a viewer might type — rather than "dyatlov", "pass", "incident" separately. Include the video's main search phrase first, then close variants, then the broader topic terms, then any common misspelling of the main phrase (tags are the one place misspellings are useful, because viewers type them).

Tags carry the most weight for search-driven and evergreen videos and for topics with ambiguous names — a place, a person or an event that could be confused with something else. For a purely browse-driven entertainment video they matter less, but the two minutes cost nothing.

Building a set from your topic (a worked method)

1. Write down the one exact query the video targets — the phrase you would type to find it.
2. Add three to five variants of that query: the question form, the "explained" form, the name-plus-year form.
3. Add the broader category terms a browser of this niche would use.
4. Choose hashtags from the top of that list: the specific topic, the niche, the angle.
5. Check the total: tags under 500 characters, hashtags between three and five, nothing irrelevant, nothing copied wholesale from a bigger channel's video (their tags describe their video and their authority, not yours).

What about Shorts?

Shorts live in a feed, not in search results, so hashtags do relatively more work there as topic signals, and tags relatively less. Three relevant hashtags in the Short's description, including the niche term a scroller's feed is built around, is a sensible default. The same honesty rule applies: a Short tagged with an unrelated trend may get a burst of wrong viewers and die faster than it would have on its own topic.

The honest bottom line

Hashtags and tags will not rescue a video nobody clicks, and they will not sink a video everybody loves. They are the last 2% — worth doing properly because they are cheap, and worth ignoring as an explanation when a video underperforms (that answer is almost always in the thumbnail, the title or the first thirty seconds). Do them once, do them from the topic, and move on to the work that matters.