If you only track likes on your Reels, you are reading the wrong number. Likes are a consequence. The input Instagram cares about most for short video is far simpler: did people keep watching?
Why watch time outranks everything else
Instagram earns money from attention. A Reel that holds people keeps them in the app; a Reel they swipe past does not. So the system tests your Reel on a small group, measures how long they stayed, and decides whether to show it to a larger one. Watch time is the closest thing to a direct vote, which is why two Reels with identical like counts can get wildly different reach.
The two watch-time numbers
Average watch time
How long a typical viewer stayed, per view. This is your quality signal. It tells you whether your hook and pacing work, independent of how many people saw it.
Total watch time
Average watch time multiplied by views — the total attention your Reel captured. This is the number that scales distribution: a long-ish Reel that holds people can out-earn a short one with more views.
What counts as a good average watch time?
There is no universal second-count, because it depends on length. Judge it as a percentage of the Reel instead:
- Under 30% — your hook is failing. People are leaving in the first seconds.
- 30% to 60% — normal. The idea works but the middle sags.
- Over 60% — strong. Expect Instagram to push it further.
- Over 100% — yes, possible, and the best outcome there is: people are looping or rewatching. Short, satisfying Reels do this.
The first three seconds decide the rest
Most Reels are lost before the idea even starts. Common killers: a slow intro, a logo animation, a long “hey guys, welcome back,” or a first frame that does not say what this is. Fixes that work:
- Start mid-action. Cut the run-up entirely; begin at the interesting part.
- Say the payoff up front. “This one change doubled my reach” then explain.
- Put text on frame one. Most people watch on mute — the first frame must carry the hook.
- Create an open loop. Pose the question early, answer it at the end.
Make it loop
A 7-second Reel that plays three times gives you far more total watch time than a 45-second one people abandon at 10 seconds. If the ending flows naturally back into the beginning, viewers often watch twice without realising. Shorter is usually the easier win.
Diagnosing your Reels
- Lots of views, low watch time — your hook is doing its job but the content does not pay off. Tighten the middle and cut the ending.
- Low views, high watch time — your content is good and under-distributed. The hook or cover frame is not stopping the scroll. Remake it with a stronger opening.
- Both low — the topic itself is not landing. Check which subjects your audience actually responds to before making more.
Track it per Reel, not in aggregate
An account-wide average hides everything. The useful move is ranking every Reel by average watch time and studying the top two — their hook, length, and pacing are your template. Valarcha pulls watch time for each of your Reels and ranks them, so your best-performing hook is obvious instead of buried. Pair this with posting when your audience is active and the early-engagement window works in your favour.