Longer reels get more views. Bigger accounts make longer reels. Those are not the same finding.
Longer reels get more views in our data. They also come from much bigger accounts. Read per follower, the pattern reverses. Measured 16 September 2026.
What the numbers say
Read straight down the views column, a reel of 90 seconds or more has a median of 112,112 views against 13,663 for a reel under 8 seconds, which is eight times more. Read the follower column beside it and the reason appears: the accounts posting those long reels have a median of 168,765 followers against 26,232 for the short ones. Divide one by the other and the ranking inverts. Reels under 8 seconds return a median of 0.77 views per follower, the highest of any band. The 30 to 59 second band, which is also the most common at 616 of the 2,053 reels, returns 0.42, the lowest.
The practical reading is that length is not a lever on its own. The most crowded length in the data is also the weakest relative to audience size, which is what you would expect if creators pick 30 to 59 seconds by default rather than because the content needs it. Pick the length the idea needs and spend the effort on the first three seconds instead.
Sample of 2,053, measured on 2026-09-16. Scrptly production database, queried 2026-09-16. Distinct Instagram reels, deduplicated by shortcode, keeping the highest view count where one reel had been analysed more than once.
The measurement
Each point is one duration band. Median views and median follower count are both medians within that band, not averages, because a single 149 million view reel would drag any average it appeared in. Views per follower is the median of the per reel ratio, which is not the same as dividing the two medians above it, and is the more honest of the two.
| Band | Reels | Median views | Median followers | Median views per follower |
|---|---|---|---|---|
| Under 8 seconds | 141 | 13,663 | 26,232 | 0.77 |
| 8 to 14 seconds | 221 | 59,302 | 131,356 | 0.58 |
| 15 to 29 seconds | 275 | 91,065 | 153,229 | 0.69 |
| 30 to 59 seconds | 616 | 89,053 | 225,412 | 0.42 |
| 60 to 89 seconds | 459 | 110,121 | 204,726 | 0.48 |
| 90 seconds and over | 341 | 112,112 | 168,765 | 0.62 |
One Reel, beat by beat
One reel from the 30 to 59 second band, the band that performs worst per follower, chosen because it is an example of that length being earned rather than defaulted to. 7,577,545 views, 50.8 seconds, hook score 10 of 10, retention score 88. Timings are exactly as our transcriber returned them.
| At | What happens |
|---|---|
| 0:00 | The moment you start working harder than the people around you, they stop calling it dedication, they |
| 0:04 | start calling it obsession. They tell you to slow down, to balance, to not lose yourself, |
| 0:09 | to remember that life is more than work, and you start to believe them because it sounds logical. |
| 0:14 | But you forget that this is not about them. Their dreams are not at stake here, yours are. |
| 0:19 | If you want a life of your choice, you cannot live like them, and if your |
| 0:23 | goals do not make you a little inhuman, a little too intense, a little uncomfortable to |
| 0:28 | be around, they are too small. Because ambition is supposed to feel uncomfortable. |
| 0:33 | It is supposed to ask too much of you. It is supposed to scare the people who want you to stay normal. |
| 0:38 | If your dream does not make you a little insane, you are thinking inside your comfort zone. |
| 0:43 | The world calls everyone who refuses to settle obsessed. |
| 0:47 | Take pride in being one of them. |
What to do next
- Stop treating 30 to 59 seconds as the safe default. It is the most crowded band in this data and the weakest one per follower, which is what a default looks like from the outside.
- Before you cut a reel to hit a length, check whether the idea survives the cut. The 50 second example above holds attention because every four seconds turns the argument, not because 50 seconds is a good number.
- If you want to test this on your own account rather than ours, take your last twenty reels, divide each view count by your follower count on the day it posted, and sort by length. Twenty reels is not a study, but it is your audience rather than somebody else s.