AVD (average view duration) is the only YouTube Analytics metric that consistently predicts whether the algorithm will push a video harder. CTR gets you the impression; AVD keeps you in the recommended feed. Subscribers don't matter. Views don't matter. AVD matters.
But "good AVD" doesn't exist in a vacuum. A 6-minute video and a 60-minute video can't be compared by raw AVD. So here's the niche-by-niche breakdown from the channels we've edited in the last 12 months.
Benchmark table
| Niche | Typical length | Healthy AVD | % of length |
|---|---|---|---|
| Podcasts (video) | 60–120 min | 18–32 min | 25–35% |
| Business / startup | 8–15 min | 4–7 min | 40–55% |
| Education / tutorials | 6–12 min | 3.5–6 min | 45–60% |
| Tech reviews | 8–14 min | 4–7 min | 45–55% |
| Video essays | 15–35 min | 7–18 min | 40–60% |
| Vlogs / lifestyle | 8–12 min | 2.5–5 min | 30–45% |
| Gaming (no commentary) | 15–30 min | 5–10 min | 30–45% |
| Gaming (commentary) | 12–20 min | 5–9 min | 40–55% |
Caveat: these are the ranges we see on channels between 5k and 200k subscribers. Smaller channels usually start below the bottom range; very large channels (1M+) often beat the top. The math compounds — the algorithm pushes harder once you break a threshold, which lifts AVD further.
The metric that matters more than raw AVD
YouTube actually weights the 50%-of-video-length retention point more than raw AVD. The percentage of viewers still watching at the midpoint of your video is the signal that gates whether YouTube pushes you to the recommended feed.
Healthy 50% retention by niche (2026):
- Business / tutorials: 55–70% at the 50% mark
- Video essays: 45–65%
- Vlogs: 35–50%
- Video podcasts: 40–55%
If you're under these numbers, the algorithm is going to throttle distribution regardless of how many subs you have. The fix is structural editing changes — see our how-to-grow-a-YouTube-podcast playbook for the 7 things every stuck channel gets wrong.
What moves AVD the most
- Cold opens. Lead with the most interesting moment of the video, not your intro. Single biggest lever — we see 20–40% AVD improvements just from adding cold opens.
- B-roll density. Pure talking head loses retention at the 30-second mark. Adding B-roll, on-screen receipts, and motion graphics keeps eyes engaged. Aim for B-roll on at least 40% of total screen time.
- Pacing cuts. Cut every "um", every pause, every sentence that doesn't serve the next 15 seconds. We typically cut 15–25% of raw runtime.
- Chapter markers. Lets viewers skip to what they care about — which sounds bad for AVD but actually improves it because they stay for the section they wanted instead of leaving entirely.
- Sound design. Audio cues, music ducking, SFX. Increases perceived production value and keeps the brain stimulated. We see 10–15% AVD lifts from sound design alone.
"AVD is the metric YouTube actually pays for. Cold opens and B-roll density move it more than anything else."
How to actually measure
Every quarter, pull your top 5 and bottom 5 videos by AVD. Don't look at views or subs — look at the AVD curve in YouTube Studio. What did the top 5 have in common (length, cold open style, B-roll density)? What did the bottom 5 share?
We do this analysis at the start of every monthly engagement — it's free, it's data you already have, and it changes how we edit your future videos. If you want us to run it on your channel, book a call.
Common AVD myths
"Shorter videos always have better AVD." False. Length doesn't determine AVD — pacing and structure do. We've seen 35-minute videos beat 7-minute videos on the same channel because the long one was structured better.
"You should never go over 10 minutes." False in 2026. Long-form is back, driven by the audience-overlap with podcasts. YouTube actively pushes longer content if AVD stays healthy.
"Click-baity titles tank AVD." Sort of. Clickbait that the video doesn't deliver on tanks AVD. Clickbait that does deliver actually raises AVD because curiosity drives stay-time.