Shorts Analytics That Actually Matter Most Days

Shorts analytics hand you a dozen numbers and no instructions. The two that predict your next upload are average view percentage and swipe-away rate, and the rest is context worth checking once a week. Most creators open the dashboard, glance at likes, and close it feeling informed. That is the problem.

Likes and comments feel like votes, so they get trusted more than they deserve. They arrive after the decision that matters, which is whether someone stayed past the point where leaving was easy. By the time a viewer taps the heart, the graph already knows whether the clip held them. I keep the emotional numbers in a separate column so they cannot crowd out the ones I can act on. A like is a friendly note from a stranger. It is a poor instruction for the next script.

Watch time percentage beats raw views

Views count starts. They tell you the hook survived the first frame and nothing after that. A clip with a million views and a two-second average was a thumbnail people clicked, not a video people watched. The number is exciting and almost empty. I stopped leading with it, because it answers a question nobody in the room is actually asking.

Average view percentage tells you how much of the clip the average viewer consumed. Sixty percent on a forty-second Short is a healthy sign. Fifteen percent on the same clip means the promise and the payoff live far apart, and no amount of captions will close that gap.

A view is a handshake. The percentage is whether anyone stayed in the room.

Where swipe-away rate fits

Swipe-away rate is the share of viewers who left before the end. It is the honest cousin of retention, because it counts departures instead of folding them into an average. When it climbs, the clip is losing people in the middle, not at the first frame. A rate near zero on a very short clip is normal, since people reach the end before they get a chance to leave.

I read swipe-away next to average view percentage. Together they tell me whether the opening held and whether the ending kept its promise. One number alone can flatter you for a week before the trend shows up in reach.

The numbers I stopped opening

Follower count per clip taught me nothing, because a Short can travel far beyond the people who follow me. Shares matter for reach but they are a result, not a lever I can pull while writing. Impressions without watch time are just a count of people who did not care.

None of those numbers are useless. They are downstream. When the two core signals move, the downstream ones follow, which is why I check them after the fact instead of chasing them during the edit. Treating them as the headline is how a channel grows its reach while its audience quietly stops watching.

Signals worth a weekly check

• Average view percentage above fifty on clips you built to be watched twice.
• Swipe-away rate that drops after you shorten the middle of a similar clip.
• New viewers from the Shorts feed, which tells you whether strangers found it.
• Returning viewers on a series, which tells you the format is worth repeating.
• The gap between your best and worst clip, which shows how much the topic decides.

Key takeaways

• Track two numbers daily and the rest once a week.
• Read average view percentage and swipe-away rate together.
• Treat likes as a lagging signal, not a quality score.
• Compare clips of similar length before you trust a difference.