Watch Time & Completion Rate: TikTok's Most Important Metrics
Watch time and completion rate are the two metrics that most directly determine how far a TikTok video travels. Likes and comments are visible and satisfying, but they are downstream signals; the recommendation system's primary question is simpler and colder: when this video appeared on a screen, did the person keep watching it?
The logic is economic. TikTok's product is attention, measured in seconds on the feed. A video that holds viewers is inventory worth distributing; a video people swipe away from costs the platform engagement every time it is shown. Every other ranking signal is, in effect, a correction term on top of these two.
The definitions, precisely
Watch time comes in two forms. Total watch time is the sum of all seconds all viewers spent on the video — the raw quantity of attention it generated. Average watch time is total watch time divided by views, shown in seconds in TikTok's per-video analytics. Because a "view" on TikTok counts essentially from the moment a video starts playing, average watch time includes every instant swipe-away, which is why the number often looks brutally low.
Completion rate is the percentage of plays in which the viewer reached the end of the video; TikTok reports it as "watched full video." A related and stronger signal is the loop or rewatch: when viewers watch past the end into a second play, average watch time can exceed the video's actual length — a 10-second video with a 13-second average watch time has been, on average, watched 1.3 times per view.
Why ratio metrics beat count metrics
The recommendation system compares videos shown to audiences of wildly different sizes, so it leans on ratios that are scale-independent. A video with 500 views and a 60% completion rate carries a stronger quality signal than one with 50,000 views and 12% completion — and the batch-testing system treats it accordingly, expanding distribution for the first while throttling the second.
This is also why engagement rate alone can mislead. Likes correlate with retention but lag it: viewers who swipe away in two seconds never reach the moment of liking. When creators see high retention but few likes, distribution usually continues; the reverse — decent likes from a loyal audience but poor retention among new viewers — is the classic signature of a video that stalls.
What counts as good: realistic benchmarks
There are no official thresholds, and numbers vary by niche and audience — but creator analytics across the platform cluster consistently enough to give working ranges:
- Short videos (under ~15 seconds): completion rates above 50–60% are strong, and averages over 100% (looping) are the profile of breakout clips.
- Mid-length videos (15–60 seconds): 30–50% completion is solid; the best performers hold a majority of viewers past the halfway mark.
- Long videos (1 minute and up): completion drops naturally — often 10–25% — but total and average watch time rise, which the system increasingly rewards as TikTok pushes longer formats.
- The first seconds dominate: across lengths, the steepest drop-off in any retention curve happens in roughly the first two to three seconds, before content quality has had any chance to matter.
Where to find these numbers in TikTok analytics
Per-video analytics (tap a video → the analytics or "more data" view) show average watch time, watched-full-video percentage, total play time, and a retention graph plotting the percentage of viewers still watching at each second. The retention graph is the most diagnostic chart on the platform: a cliff at the start indicates the opening lost people; a mid-video slump marks the exact second interest broke; a flat curve that bumps up at the end usually indicates loops.
The same view breaks down traffic sources (For You, followers, search, profile) and the new-versus-returning viewer split. Reading retention alongside traffic source matters because follower views retain better than cold FYP views — a video can look healthy on average while failing the only audience that determines further distribution.
The length trade-off
Completion rate and total watch time pull in opposite directions: shortening a video raises the percentage who finish it but lowers the seconds each view contributes, and lengthening does the reverse. TikTok's system values both, and its weighting has shifted toward total watch time as the platform promotes longer content — most visibly through the Creator Rewards Program, which only pays on videos over one minute. The practical interpretation: neither metric is a target in itself. They are two readings of the same underlying quantity — sustained viewer interest — measured per second and per play.