Your Followers Still See You, Nobody New Does: Instagram’s Aggregator Penalty

Your likes look fine. Your comment section is not dead. Your follower count has not dropped. And yet you have not gained a meaningful new follower in weeks, and your reach chart looks like a flat line drawn with a ruler.

Most people at this point start searching for "Instagram shadowban." That is usually the wrong diagnosis, and chasing it wastes months. There is a different mechanism that produces almost exactly these symptoms, and it has a name: the aggregator classification.

What the aggregator penalty actually does

Instagram classifies accounts that primarily post content they did not create or meaningfully edit as aggregators. The consequence is specific and unusually surgical:

• You are removed from recommendation surfaces — Reels, Explore, and feed suggestions to non-followers.
• People who already follow you keep seeing your posts normally.

Read those two lines together and the confusing symptom snaps into focus. Your engagement rate against your existing audience stays healthy, because that audience is untouched. What disappears is the pipeline of strangers. As one way to say it: it is not that nobody likes you, it is that nobody can be recommended to you.

A shadowban suppresses your content. The aggregator classification suppresses your discovery. The metrics look nothing alike once you know where to check.

The scope changed in 2026 — this is why it hit you now

Instagram first started penalizing recycled Reels back in 2024. Photos and carousels were untouched, so a lot of accounts kept running repost-heavy image feeds without consequence.

That gap closed. As of April 30, 2026, the originality enforcement was extended to all content formats, including images and carousels. If your reach fell off a cliff sometime after that date and you mostly post other people's photos, memes, or clips with a caption added, you have your answer.

The timing detail matters for self-diagnosis. If you can point to a week where non-follower reach dropped and stayed down, line it up against that date. A gradual decline over months is usually a content or saturation problem. A step change is usually a classification.

On the Facebook side, Meta published an update in March 2026 covering the same territory: refreshed written guidance, new enforcement mechanisms, and expanded protection tools for original creators. Accounts crossing the unoriginality threshold lose access to monetization programs for a period and can be marked non-recommendable — the recommendation system stops surfacing them to new audiences.

Who this catches that did not expect it

The label "aggregator" suggests meme pages and repost farms, and those are the obvious cases. But the classification is behavioral, not intentional, and it catches several account types that consider themselves legitimate creators:

• Curation accounts. Design inspiration, quote pages, industry news roundups — valuable to followers, and structurally indistinguishable from reposting.
• Brand accounts running user-generated content. If the majority of your feed is customer photos and creator clips, the account is mostly publishing material it did not create.
• Clip channels. Podcast highlights, stream cuts, and reaction snippets, especially where the source is someone else's recording.
• Multi-account operators sharing one library. Every account drawing from the same pool looks like an aggregator to each platform independently.

If you are in one of these categories, the practical question is not whether reposting is allowed — it is whether a stranger scrolling your last thirty posts would conclude a specific person makes this. That is roughly the judgment the classifier is approximating.

How to tell an aggregator classification from a real shadowban

Run these four checks before you change anything:

1. Check follower vs non-follower reach in Insights. This is the decisive test. Aggregator classification collapses non-follower reach specifically while follower reach holds. A broad suppression hits both.
2. Search your own account and a recent hashtag from a logged-out device. If you are findable, you are not hidden — you are just not recommended.
3. Audit your last 30 posts honestly. How many are things you filmed, shot, or built versus things you found? If the second number is the majority, stop looking for a mysterious penalty.
4. Look for low-value edits. Adding borders, slapping captions on someone else's clip, or changing playback speed are explicitly named as the kind of changes that do not make content original.

If all four point at aggregation, see what actually counts as original content in 2026 for the positive definition. If they do not, the three-layer shadowban breakdown is the better path.

Getting recommended again

The recovery is unglamorous and mostly about ratio, not volume:

• Shift the mix. The classification is based on what your account mostly does over a rolling window. You do not have to purge reposts; you have to stop being predominantly reposts.
• Put yourself in the frame. Reels that incorporate third-party material still count as original when a creator is visibly present and adds something genuinely new — fresh information, analysis, or a real improvement to the storyline. Presence plus contribution is the test.
• Stop treating a caption as transformation. Commentary in the caption does not change what the algorithm sees in the media itself.
• Optimize for shares, not likes. 2026 analyses of Reels ranking put shares — especially DM shares — among the heaviest signals, with some estimating DM shares carry several times the weight of a like.

Set expectations about the recovery curve, because this is where people give up too early. The classification is evaluated on your recent posting pattern, so it does not lift the moment you publish one original post — it lifts once the recent window is dominated by original work. That means weeks of consistent output before recommendation surfaces reopen, and it means resisting the temptation to fill gaps with reposts during the wait, which resets the clock.

A second trap worth naming: many accounts respond to a reach collapse by posting more often. If the added volume is more of the same reposted material, higher frequency accelerates the classification rather than escaping it. Fewer, more original posts beat more of the same — a pattern that repeats across every 2026 platform change.

One structural note for anyone running more than one account: if several of your accounts are aggregating from the same pool of sources, you are not diversifying risk — you are reproducing the same classification across every account you own. The durable version of multi-account growth is each account holding its own niche and producing its own material. That is the design principle behind tools like NoobClaw, where every account generates its own content from its own topic pool rather than sharing one feed of reposts. See also how to build a social media matrix strategy.

FAQ

How long does the aggregator classification last?

It is evaluated on a rolling basis rather than issued as a fixed-term penalty, so it lifts when your recent posting pattern changes rather than when a timer expires. Reports describe near-duplicate posting drawing restrictions ranging from about 24 hours to 30 days depending on severity, but those figures come from third-party analysis rather than Meta.

Does Instagram tell you when this happens?

Generally not in a way that is easy to notice, which is the core frustration. Account status tools flag some issues, but "you are no longer recommended to new audiences" is not delivered as a clear alert. You are expected to read it off your own reach data.

Can I still repost anything at all?

Yes. The rule targets accounts whose identity is reposting. Sharing a clip occasionally and building on it with your own commentary on camera is a different thing from running a feed made of other people's work. Credit does not change the classification, but genuine contribution does.