Everyone wants to know if their numbers are “good.” The honest answer is that it depends on the platform, the clip length, and what you’re measuring. But you can still anchor yourself against realistic ranges — and once you know which metric actually moves the algorithm, you know what to fix.
A note on the numbers below: these are industry ranges and rules of thumb, drawn from commonly cited platform observations and the patterns we see in creator data. They are not exact findings from a single named study, and they shift as platforms change their algorithms. Treat them as a calibration tool, not gospel.
The five metrics that matter
Before the tables, here’s what each metric means and why it counts.
- Average view duration (AVD) — how many seconds, on average, people watch before scrolling. The raw fuel for most algorithms.
- Completion / retention rate — the percentage of viewers who reach the end (or who watch some defined chunk, like 75%). On short clips this is the metric platforms reward most, because a finished short often loops.
- Save rate — saves divided by views. A save is the strongest signal a viewer can give that isn’t a follow: it says “I want this later.”
- Share rate — shares divided by views. Shares pull in new, cold viewers, so platforms weight them heavily for reach.
- Follow-through rate — new follows divided by views. The clearest sign a single clip converted a stranger into an audience member.
Average view duration and completion by platform
Completion is easiest on the shortest clips and hardest on anything that asks for a minute-plus commitment. These ranges assume a competent hook and captions.
| Platform | Typical AVD (short clip) | Typical completion rate | What “good” looks like |
|---|---|---|---|
| TikTok | ~50–80% of clip length | 40–60% on sub-15s clips | 70%+ completion on short clips |
| Instagram Reels | ~45–70% of clip length | 35–55% on sub-15s clips | 60%+ completion, strong replays |
| YouTube Shorts | ~50–75% of clip length | 40–60% on sub-20s clips | 65%+ retention, low early drop-off |
The pattern holds across all three: the shorter and tighter the clip, the higher the completion rate, because there’s simply less runway for someone to bail.
How clip length changes the math
Length is the single biggest lever on completion. The longer the clip, the lower the percentage that finishes it — but longer clips can still win on total watch time if the content earns the runtime.
| Clip length | Typical completion range | Notes |
|---|---|---|
| Under 15s | 50–70% | Easiest to fully watch; loops inflate it further |
| 15–30s | 35–55% | The sweet spot for most hooks; still very loopable |
| 30–60s | 20–40% | Needs a genuine story arc to hold the back half |
| 60–90s+ | 10–25% | Reserve for content that truly justifies the length |
If you’re cutting one long video into shorts, this table is your argument for ruthless trimming. A 22-second clip that finishes strong will almost always outperform the same idea padded to 45 seconds. Cutting cleanly on the audio — never mid-word, never on dead silence — is what keeps a short tight enough to land in the higher completion bands.
Save and share rates
Saves and shares are smaller numbers by nature — most viewers do neither — so the “good” thresholds look tiny next to completion rates. Don’t be fooled; a 1% save rate is a strong clip.
| Metric | Typical range | ”Good” benchmark |
|---|---|---|
| Save rate | 0.2–1.5% of views | 1%+ is excellent |
| Share rate | 0.3–2% of views | 1.5%+ signals real reach potential |
| Follow-through | 0.1–1% of views | 0.5%+ on a cold clip is strong |
Saves skew high on reference and tutorial content — anything someone might want to act on later. Shares skew high on emotional or surprising content — the stuff people send to one specific friend with “this is so you.” Knowing which kind of clip you’re making tells you which number to watch.
What actually moves each metric
You don’t optimize all five at once. You pick the weak one and pull the right lever.
- To raise completion, fix the first two to three seconds. Most drop-off happens before the third second — if the hook doesn’t land, nothing downstream matters. See why clips fail in the first three seconds.
- To raise average view duration, shorten the clip or tighten the middle. Cut the throat-clearing. Every dead beat is a place a thumb keeps scrolling.
- To make a clip loop (which quietly inflates both completion and AVD), end on a beat that flows back into the opening, or close on a line that makes the first frame land harder the second time.
- To raise save and share rates, give people a reason: a list, a number, a “save this for later,” or a payoff worth sending to someone.
- To raise watch time on muted feeds — which is most of them — caption everything, word by word. Roughly the majority of feed video is watched on mute, so uncaptioned clips forfeit watch time before they start. More on the mechanics in how muted-viewing optimization works.
The metric most worth obsessing over is completion on short clips. It’s the one platforms reward most directly, it’s the most fixable, and a clip that finishes is a clip that loops.
How to benchmark yourself honestly
- Compare like with like. Judge a 30-second talking-head clip against other 30-second talking-head clips, not against a 9-second trend.
- Use your own median as the baseline. Your last 20 clips’ median completion rate is a more useful target than any industry table — beat yourself first.
- Watch the trend, not the single post. One viral clip and one flop tell you little. The slope across a month tells you everything.
- Track the funnel, not just views. Views without saves, shares, or follows is reach that didn’t convert. Pick the leak and patch it.
This is also why cutting from one source into many ranked candidates pays off: when you can post several clips from the same recording, your median stabilizes fast, and you learn which hooks and lengths your audience actually finishes. Videotrim scores each segment on hook strength, story arc, and quotability so the candidates most likely to clear these completion bands surface first — see how the scoring and clipping work or the lengths and formats it targets.
Benchmarks tell you where the bar is. Your own median tells you whether you’re clearing it. Chase the second number and the first takes care of itself.