The pattern people call a shadowban is what a normal reel looks like

What people mean by it

A shadowban is the belief that a platform has quietly limited your reach without telling you. The experience behind it is real: views drop, the account feels invisible, and nothing in the app explains why. The explanation is usually wrong, and the reason is that almost nobody knows what a normal reel does.

Normal is lower than you think

Of 2,543 Reels we analysed, 1,667 reached fewer people than the account already had followers. That is 66 percent. Landing under your own follower count is not evidence of suppression. It is the single most common outcome in the sample.

It is worse at the account level. 302 of 534 accounts, 57 percent, have no reel in our sample that beat three views per follower at all.

Three things that look like a shadowban and are not

1. Your Reels got shorter. Under 15 seconds has a median of 48,548 views against 217,962 for 15 to 30 seconds. Drifting short produces exactly the graph people screenshot as proof of suppression.
2. You had one outlier and reset your baseline to it. A single good reel makes every normal reel afterwards look like a penalty.
3. Your openings converged. Ten Reels that start the same way will perform like one reel repeated.

How to actually check

1. Compute views divided by followers for your last 20 Reels, and take the median. That is your baseline.
2. Compare this month to that number, not to your best ever.
3. Look for a step change, not a low level. Suppression would look like a sudden cliff across every reel; being average looks like a flat line that was always there.
4. Check length first. It is the largest controllable effect we can measure.

None of this proves shadowbans never happen. It says the evidence most people have is also exactly what ordinary performance looks like, and ordinary is the more likely explanation.

Figures on this page come from 2,543 Instagram Reels across 534 accounts, analysed by Scrptly and deduplicated so no reel is counted twice. Read 2026-10-13.