Three real Reels, all with millions of views, and the reason they are not comparable
Three real Reels with millions of views each, and why one reached 14 times its audience and another a fifth of it. Measured 19 September 2026.
The honest version first
Pick any three reels with millions of views and you would call all three a success. Measured against the audience behind them they are not remotely the same thing. The first reached about fourteen and a half times its following. The second reached about half of its own. The third reached a fifth. Every one of them was scored 9 out of 10 on the hook by our analysis, which is the other half of the point: the score did not separate them either.
Side by side
All figures read from Scrptly production data on 2026-09-19. Hook and retention are scores our own analysis assigned. Times its audience is views divided by the account's follower count on that reel.
| Reel | Views | Times its audience | Our scores |
|---|---|---|---|
| A 13.1 second reel from an 8.0 million follower account | 115,756,571 | 14.50x | Hook 9, retention 92 |
| A 179.2 second reel from a 56.2 million follower account | 29,235,195 | 0.52x, almost exactly the corpus median | Hook 9, retention 89 |
| A 43.9 second reel from a 50.2 million follower account | 10,750,474 | 0.21x | Hook 9, retention 82 |
What the numbers say
The reel with the most views on this page, 115,756,571, came from the smallest account of the three at 8.0 million followers. The reel with the fewest, 10,750,474, came from an account with 50.2 million. So the ranking by views and the ranking by account size are almost exactly inverted, and the ranking by our own hook score does not exist at all, because all three scored 9.
The useful habit is to stop reading examples as league tables. A reel you admire on a large account may be underperforming its own audience, and a modest looking number on a small account may be the best thing in the feed that week. Before copying anything, divide its views by its creator's followers. It takes ten seconds and it changes which reels you decide are worth learning from.
Sample of 2,166, measured on 2026-09-19. Scrptly production database, queried 2026-09-19. Instagram Reels only, deduplicated by shortcode, one query run.
The measurement
Where those three sit in the whole corpus. Every deduplicated reel with a view count and a follower count, grouped by reach relative to its own following. The bands add to 100 percent.
| Reach | Share | The example here |
|---|---|---|
| Fewer views than followers | 64% | The 0.52x and the 0.21x reels both sit in this band |
| One to ten times | 26% | None of the three |
| Ten to a hundred times | 8% | The 14.50x reel |
| More than a hundred times | 2% | None of the three |
What to do next
- Take the last three reels you saved as inspiration and work out each one's views per follower. At least one of them will probably turn out to be an ordinary reel from a big account.
- Then look at what the high ratio example does differently. The 14.50x reel here runs 13.1 seconds against 179.2 for the median one, and the corpus median is 48.1.
- Run one of them through Scrptly if you want the breakdown rather than the ratio. Two analyses are free, and the structure is where the copyable part lives.