YouTube SEO has its own vocabulary, and half the confusion new creators and marketers run into comes from not knowing what the platform’s own metrics and algorithm concepts actually mean. Below is a working glossary — the terms you will encounter in YouTube Studio, in third-party tools, and in any serious conversation about ranking and growing a channel — defined the way a practitioner would explain them to a client, not the way a textbook would.
This is not a dictionary of generic SEO terms. Every entry here is specific to how YouTube ranks, recommends, and reports on video performance. Bookmark it and come back when a term in Studio or a strategy conversation doesn’t quite make sense.
Discovery & Ranking Terms
These are the terms that describe how a video gets found in the first place — the mechanisms YouTube uses to surface content to viewers who never typed a search query at all.
- Impressions: The number of times your thumbnail was shown to a viewer on YouTube, whether on the homepage, in search, or in the “up next” sidebar. An impression is not a view — it’s an opportunity for one.
- Traffic source: The category YouTube assigns to where a view came from — Browse features, Suggested videos, YouTube search, External, Direct or unknown, Playlists, Notifications, and others. Diagnosing a channel’s growth or stagnation almost always starts by reading the traffic source breakdown in Studio Analytics.
- Browse features: Views that came from YouTube’s homepage or the subscriptions feed, as opposed to being served alongside another video. A channel that earns strong browse-feature traffic has convinced the algorithm its content deserves proactive placement, not just reactive suggestion.
- Suggested videos: The “up next” and sidebar recommendations shown while a viewer is watching another video. This is often the single largest traffic source for established channels, and it rewards videos that keep people on YouTube rather than sending them away.
- Session duration / session watch time: How long a viewer stays on YouTube, across multiple videos, after landing on one of yours. YouTube optimizes recommendations partly around keeping the overall session going, not just maximizing time on a single video, which is why videos that lead into other watchable content tend to get recommended more.
- Search ranking: Where a video appears in results when a viewer types a query directly into YouTube search. This is the piece of YouTube SEO that behaves most like traditional Google SEO — title, description, and relevance signals matter heavily here.
Metadata & On-Page Terms
Metadata is everything a creator directly controls that tells YouTube (and viewers) what a video is about. Getting this layer right is the foundation everything else builds on.
- Title tag (video title): The clickable headline of a video, capped at 100 characters, though the first 60–70 are what reliably display in search and suggested placements. It is the single highest-weighted metadata field for both ranking and click-through.
- Description: The text field beneath a video, up to 5,000 characters, used by YouTube’s algorithm to understand context and by viewers to find timestamps, links, and additional detail. The first two to three lines display before a “show more” click, so front-load the value there.
- Tags: Keywords entered in the upload flow that help YouTube catch misspellings and closely related searches. Tags carry far less ranking weight than title, description, and actual viewer behavior, but they’re a low-effort insurance policy against typo-variant queries.
- Chapters: Timestamp markers added in the description (or via the editor) that break a video into named sections, displayed as a scrubber bar in the player. Chapters improve navigability, which supports retention, and the chapter titles themselves are searchable text.
- Closed captions / subtitle file: Text transcripts of the spoken audio, either auto-generated or uploaded manually. Accurate captions give YouTube’s systems a clean, literal read of every word spoken in the video — arguably the richest single content signal available to the algorithm.
- Video schema (VideoObject structured data): Markup added to a webpage that embeds a YouTube video, telling search engines the video’s title, duration, upload date, and thumbnail so it can appear as a rich result in Google search, separate from YouTube’s own search.
- Playlist: A curated, ordered grouping of videos that autoplays sequentially. Playlists are themselves searchable and indexable, and they’re one of the most reliable levers for extending session watch time.
Engagement & Retention Metrics Terms
These are the numbers Studio Analytics surfaces to describe how viewers actually behave once they land on a video — arguably more influential to ranking than any metadata field.
- Click-through rate (CTR): The percentage of people who saw a video’s thumbnail (an impression) and clicked to watch. CTR is measured per traffic source and is the primary signal YouTube uses to judge whether a thumbnail-title pairing is compelling.
- Average view duration: The average length of time viewers watch a given video, shown in both minutes and as a percentage of total runtime. It’s a blunt but useful proxy for whether a video delivers on its promise.
- Audience retention graph: A second-by-second line chart showing what percentage of viewers are still watching at every point in a video. Sharp drops reveal exactly where viewers lose interest — the single most diagnostic tool available for improving future edits and scripts.
- Watch time: The cumulative number of minutes all viewers have spent watching a video or channel. YouTube has historically weighted total watch time heavily in ranking decisions because it directly serves the platform’s goal of keeping people watching.
- Engagement rate: A composite measure (varies by tool) combining likes, comments, shares, and sometimes subscribes relative to views. It’s a secondary signal compared to retention and CTR, but strong engagement correlates with content the algorithm is comfortable recommending broadly.
- Relative audience retention: A benchmarked version of the retention graph that compares a video’s retention curve against similar-length videos on YouTube, showing whether a given section is over- or under-performing relative to typical viewer behavior.
Algorithm & Recommendation Terms
These terms describe the machinery behind the scenes — the systems that decide who sees what, and why some videos snowball while others stall at a few hundred views.
- The algorithm: Shorthand for YouTube’s recommendation system, which is really a collection of machine-learning models trained to predict what a given viewer is likely to watch and enjoy next. There is no single ranking formula — it’s a probabilistic system that updates constantly based on aggregate viewer behavior.
- Video velocity: The speed at which a video accumulates views, likes, and comments relative to its subscriber base and typical performance, especially in the first 24–48 hours. Strong early velocity is a common trigger for wider algorithmic distribution.
- Topical authority (channel-level): The degree to which YouTube’s systems associate a channel with a specific subject area based on its history of uploads. A channel that consistently covers one topic tends to get recommended more readily within that topic than a channel that jumps between unrelated subjects.
- Cold start problem: The challenge new videos and channels face because the algorithm has no viewer-behavior data yet to base recommendations on. Early performance signals — from an existing audience, an email list, or a community post — help YouTube “warm up” a new upload faster.
- Shadow-testing (impression testing): The practice of showing a small sample of viewers a video before deciding how widely to distribute it, based on their response (CTR and early retention). This is why the first hours after publishing matter disproportionately.
Thumbnail & CTR Terms
Thumbnails deserve their own cluster because the terminology around them gets thrown around loosely, and precision here actually changes what a creator tests.
- Custom thumbnail: A manually designed image uploaded to represent a video, as opposed to an auto-generated frame pulled from the footage. Channels that use deliberate custom thumbnails consistently outperform those relying on auto-generated frames.
- A/B thumbnail testing: Using YouTube’s built-in testing feature (or a third-party tool) to run two or three thumbnail variants against live traffic and let real CTR data decide the winner rather than guesswork.
- Thumbnail-title pairing: The combined read of a thumbnail and title together, since viewers process both simultaneously in a feed. A mismatch between what the thumbnail visually promises and what the title states usually depresses CTR even if each element is individually strong.
Monetization & Channel Terms
SEO strategy and channel economics are tightly linked on YouTube, so a few monetization and structural terms belong in any serious glossary.
- YouTube Partner Program (YPP): The eligibility tier that unlocks monetization features (ad revenue, channel memberships, Super Chat) once a channel crosses subscriber and watch-time thresholds. SEO fundamentals — discoverability and retention — are usually what get a channel there.
- RPM (revenue per mille): Actual earnings per thousand views after YouTube’s revenue share, as opposed to CPM, which reflects what advertisers pay before that split. RPM is the number that tells a creator what their content is actually worth per view.
- Channel authority: An informal term practitioners use for the accumulated trust signals a channel has built — consistent upload history, strong average retention, low spam/strike history — that appear to influence how readily new uploads get distributed.
Frequently Asked Questions
What's the single most important term in this glossary for a beginner to understand?
Audience retention graph. Almost every other metric on this list — CTR, watch time, suggested video placement — is downstream of whether people actually keep watching once they click, and the retention graph is the only tool that shows exactly where a video gains or loses viewers.
Is watch time the same thing as average view duration?
No. Watch time is a cumulative total across all viewers of a video or channel, while average view duration is that total divided by the number of views. A video can have enormous total watch time simply from high view count even if its average view duration is mediocre.
Do tags still matter for YouTube SEO?
They matter far less than they did in YouTube's early years. Tags mainly help catch misspellings and close variants of your target query; title, description, captions, and actual viewer engagement carry the real ranking weight now.
What's the difference between CPM and RPM?
CPM is what an advertiser pays per thousand ad impressions, before YouTube takes its revenue share. RPM is what actually lands in a creator's account per thousand views, after that split and after accounting for videos that had no ads served at all. RPM is the honest number to track.
Why does YouTube care about session watch time instead of just my video's watch time?
Because YouTube's business is keeping people on the platform, not on any single video. A video that reliably leads viewers into more YouTube content — through strong end screens, playlists, or genuinely good suggested-video fit — tends to get recommended more than an isolated video with identical individual watch time.
What is a browse feature, in plain terms?
It's YouTube proactively putting your video on someone's homepage or subscriptions feed without them searching or clicking from another video. Earning strong browse-feature traffic is a sign the algorithm trusts your content enough to push it rather than wait for a viewer to find it.
Terry has 30+ years in software and SEO. He’s the founder of Salterra Digital Services and SEO Spring Training, host of the Roundtable SEO Mastermind, and lead instructor at SEO University — teaching the exact tactics his team uses on client work.
Continue learning
Ready to master this?
This guide is one lesson from the YouTube Optimization course. Get every lesson, framework and checklist — plus the full 38-course catalog — inside SEO University.
Practitioner-focused training across the full digital marketing stack — from technical SEO to conversion optimization and the AI search era. By Salterra Digital Services, since 2011.