Reach counts people, impressions count views. The number that matters more is how either compares to your own size
Reach counts accounts, impressions count times seen. And your reach per follower depends mostly on how big you already are. Measured 25 September 2026.
What the numbers say
The definitions first, from Meta's own developer documentation: reach is the number of unique accounts that have seen the media, and impressions are the number of times the media has been seen. One person watching four times is one reach and four impressions. That is the whole distinction, and it is smaller than the amount of advice written about it. Here is the part that actually changes how you read your own numbers. Across 2,319 Reels from 483 accounts, views measured against the account's own follower count fall away sharply as accounts get bigger. Accounts under 10,000 followers reached a median of 3.10 times their following. Accounts between 100,000 and 499,999 reached 0.38. Accounts over two million reached 0.38 as well. So a small account routinely reaches several times its own audience and a large one routinely reaches a third of it, and neither is a problem being fixed or a penalty being applied.
Stop comparing your views-per-follower to a bigger account's. The two figures are not measuring the same situation. Compare yourself to your own median instead, which is the only benchmark that holds still while you grow.
Sample of 2,319, measured on 2026-09-25. Scrptly production database, queried 2026-09-25. 2,319 Reels across 483 Instagram accounts, deduplicated by shortcode, each with a view count and its account's follower count. Definitions of reach and impressions read from Meta's Instagram Platform insights documentation on 2026-09-25.
The measurement
Views per follower is the Reel's views divided by its own account's follower count. Median views is the plain number, shown alongside so the two readings sit next to each other. The five bands hold all 2,319 Reels across 483 accounts: 262 plus 545 plus 730 plus 490 plus 292. One caveat before you use the top row as a target: a small account needs very little absolute distribution to produce a large multiple, so the 3.10 is partly arithmetic rather than achievement.
| Account size | Reels measured | Median views per follower | Top tenth reaches | Median views |
|---|---|---|---|---|
| Under 10,000 followers | 262 | 3.10 | 82.12 | 2,713 |
| 10,000 to 99,999 | 545 | 0.68 | 15.14 | 25,949 |
| 100,000 to 499,999 | 730 | 0.38 | 6.00 | 78,552 |
| 500,000 to 1.9 million | 490 | 0.36 | 4.40 | 348,055 |
| 2 million or more | 292 | 0.38 | 2.05 | 5,822,879 |
One Reel, beat by beat
Reach can run far past your following, and when it does it is the content doing it rather than the account size. One real Reel from the corpus, not a composite: 436,398 followers, seen 17,633,373 times, which is 40.41 times the following. Its band's median is 0.38, so this Reel reached about a hundred times what is typical for an account that size. It runs 47 seconds.
| At | What happens |
|---|---|
| 0:00 to 0:03 | Open on the claim that sounds impossible, stated flatly and without preamble |
| 0:03 to 0:15 | Introduce the person and the single idea at the centre of what they built |
| 0:15 to 0:30 | The specific inventions and the problems each one solved. The detail is the proof |
| 0:30 to 0:40 | Show the thing working, and place it in the story of how long it took |
| 0:40 to 0:47 | Point to where the rest of it can be seen |
What to do next
- In the app, open a Reel and tap View Insights. Reach is unique accounts. Views counts plays including repeats. Both are there and they will not match, which is expected rather than a bug.
- Find your band in the table above before you judge your own numbers. If you are over 100,000 followers, a typical Reel reaching about 0.38 times your following is the norm in this corpus, not a decline.
- Do not read the top row as a goal. A 3,000 follower account needs about 9,000 views to hit 3.10 times, and a 500,000 follower account would need over a million for the same multiple.
- Benchmark against yourself. Your own median across your last twenty Reels is the only comparison that stays valid as you grow, and it is what makes a genuine drop visible.
- One honest limit: these are Reels from accounts somebody chose to analyse, so the corpus skews towards content that already worked. Read the ordering between bands as sound and the exact figures as indicative.
- Scrptly measures every Reel against its own account rather than against a global average, which is the only way the comparison means anything. The first analysis does not need an account.