How to Recognize Censorship on Social Platforms — and What to Do About It

As we  close this year, we want to put one unresolved issue clearly on the record: algorithmic censorship.  For many people, the idea of being “shadow banned” still sounds paranoid or abstract. For those of us who run mission-driven organizations and small businesses online, it is neither.

Censorship in the digital age rarely announces itself. It arrives quietly — through collapsed reach, vanished discovery, and distribution that disappears without explanation. It leaves a footprint. It is visible. It is measurable. And by the end of this year, we have learned how unmistakable it can be.

This article is offered as a year-end reflection and a practical guide. It is written to help others recognize when something is wrong — and to respond with discipline, documentation, and persistence rather than panic, anger, self-censorship, silence or defeat.

The Metrics That Matter

The first sign of censorship is not a drop in likes.  It is a collapse in non-follower reach.  On healthy social accounts — especially those with large or established followings — most views normally come from non-followers. That is how discovery works. That is how social platforms themselves describe growth.

When that ratio abruptly flips, something is wrong.  By the end of this year, these are the metrics we have learned to watch most closely.

Follower vs. Non-Follower Views

At scale, content should reach far more non-followers than followers. When censorship is active, a consistent pattern appears:

• Non-follower views drop to near zero
• Even highly shared posts show no external reach
• Paid advertisements are shown almost exclusively to existing followers

This does not reflect audience choice. It does not reflect content fatigue. It reflects interference with distribution.

Shares That Don’t Travel

One of the clearest warning signs looks like this:

• Hundreds of shares
• No corresponding increase in non-follower views

Shares are meant to move content outward. When they do not, distribution is being interrupted somewhere upstream.

Paid Ads That Behave Like Organic Posts

Advertising exists for one purpose: to reach people who do not already follow you.  When paid ads are (a) Approved, (b) Charged/Paid for and (c) delivered almost entirely to followers it means that the system is not functioning as represented.

Sudden Restoration Followed by Sharp Decline

Another hallmark of censorship is temporary restoration followed by rapid collapse.  This pattern is especially revealing because it shows what normal reach looks like — and how abruptly it can be removed again.  In the case of the chart below, a shadow-ban was lifted, the film was widely discussed in the media, the follower growth sky-rocketed, and the growth itself was taken as a sign of something wrong, so the engine put back the shadow-ban within one week of lifting it.  It remains so as we close the year.

What Not to Do When You See the Signs

When creators first notice these patterns, the instinct is often to post less, post “safer” content, avoid certain topics, assume the audience has lost interest, or assume that the social platform wins and you lost.  These are wrong responses.

Censorship works best when it convinces people to censor themselves!  The correct first response is documentation, not retreat.

What You Should Do:  Document Everything

From the moment suppression is suspected, we begin collecting evidence:  screenshots of daily reach, follower vs. non-follower breakdowns, paid ad delivery data (comparison to healthy prior activities is useful), audience growth charts and post-by-post comparisons (comparing Facebook performance to Instagram performance, for example).

This documentation matters. Even with it, the platform operators will often say ‘No restrictions detected.’  We know that there are no restrictions, because we know that if there were, the platforms are not coy about telling us, for one thing.  But the primary point here is that they don’t finish their sentence.  They declare that there are  ‘no restrictions detected’, but the rest of the sentence is ‘but we wouldn’t see any temporary or soft bans, they are done by the algorithms and we have zero visibility into knowing if that is happening or not, no knowledge, no control, no reach or tools to help that situation.’  That is the rest of the sentence that they don’t say.

If you have documentation, however, you have a case.  If they can’t explain what’s happening in your numbers, try though they might to make you go away, if you have the documentation, they will investigate.

Make Advertising (or Not) Part of the Process

An effective tool for diagnosing suppression is paid advertising. In this case, it is not about selling products or promoting content, but rather about gaining access to human support and generating measurable data to prove what is happening.

On many platforms, advertising is the only reliable pathway to technical assistance. For that reason, ads should be used strategically and temporarily.  The approach is deliberate:

If you are already spending on advertising campaigns, pause them until you get your shadow ban lifted.  If you are not spending on advertising, consider doing so on a temporary basis as part of the strategy for assistance.

• Turn on a small daily ad budget (often as low as $10 per day)
• Allow the ad to run long enough to generate delivery data
• Use that data to contact advertising support directly
• Pause or cancel the ad once support engagement has been initiated

The purpose of the ad is not reach. It is evidence.  When an ad is approved, paid for, and then delivered almost exclusively to existing followers — or shows suppressed reach inconsistent with platform norms — it creates a measurable discrepancy between what the platform claims ads do and what is actually happening.  This discrepancy is often the only way to force an investigation. Once support has been contacted and evidence submitted, we do not continue spending. If support cannot help, the ad is paused or canceled.

Make it Your Devotion to Hold the Social Platform Accountable

This process cannot be handled emotionally – it must be approached with administrative devotion.  Our system is simple and repeatable:

• Two to three mornings per week
• Two to three hours at a time
• Calm, consistent, documented contact attempts with support
• Requests to speak with a human technician (regularly)
• Continued Submission of evidence with every contact event.
• Run and pause ads periodically throughout the process.  Ads that are set up to run ten days and then paused after one day will show on their metrics.

Roughly half the time, persistence results in speaking with someone – at the beginning.  As the social platform continues to close your cases without resolution or explanation, and as you continue to try to speak with someone, actual success in getting through will drop to 10 to 20% of the time requested.

When ignored completely, open a new trouble ticket. Run an ad test and gather that new evidence. Report the problem over again.  As we are ignored, we submit more reports — not fewer.  Censorship thrives on exhaustion. We all have to refuse to provide it.

A Case Study from the End of the Year

As the film One Battle After Another approached release, we already suspected our page was being censored on one major social platform.  Because we understood the importance of that public moment, we did not wait for a crisis.

Months in advance, we documented irregular metrics and worked repeatedly with support on that platform to ensure that when the film was released, our visibility would not be restricted. We were explicit that we were not seeking special treatment — only normal distribution.

For a brief period, that appeared to happen.  For approximately five days, censorship lifted.

The data during that window was unmistakable:

• More than 80% of views came from non-followers
• Shares traveled normally
• Paid content reached beyond our existing audience
• Monetization functioned as expected

Then, without explanation, censorship returned.  Reach collapsed again. Non-follower views dropped to near zero. Paid ads reverted to follower-only delivery.  The contrast between those two periods is visible in the data. It is not anecdotal. It is not subjective. It is structural.

What We Can — and Cannot — Know

Because social platforms operate behind opaque systems, creators are left to infer causes from outcomes. Technical support representatives often cannot see — or cannot discuss — distribution routing, trust scoring, or monetization logic.

Based on our documentation, timing, and repeated interactions with support, we believe there are only a few plausible explanations for what occurred this year.

We offer them here not as accusations, but as reasoned possibilities — because silence from platforms leaves creators no alternative but to guess.

Three Plausible Explanations

1. Content-Based Censorship

It is possible that one social platform did not want to promote a radical or politically challenging film — particularly one that questions power, authority, or dominant narratives — and that our association with that film triggered censorship.  If so, that would constitute content-based suppression of lawful speech.

2. Growth-Based Interruption

It is also possible that rapid audience growth triggered automated intervention. Our organic growth from the film (One Battle After Another) and the press around the Los Angeles launch naturally caused a spike (because suppression was already happening) and then the new spike led to algorithmic restrictions on the page.  This suggests that success itself triggers censorship — even in the absence of policy violations.

3. Monetization-Based Suppression

Finally, it is impossible to ignore the timing around monetization.  During the brief five-day period when censorship lifted, our content was monetized normally — and for the first time, we received a monthly payout of over $500.  Within a week of that amount being calculated, censorship returned.

For comparison, our normal monetization is $50 per month and in one month, it grew ten times that, which could be igniting algorithmic suppression.  In the greater scheme of things, this is not much money we are talking about and surely, not worth the interruptions to follower growth.

We cannot say whether monetization played a role. But the sequence raises a legitimate question: whether internal systems deprioritize or interrupt distribution once content begins generating meaningful revenue.  None of these possibilities are reassuring.  When platforms refuse to explain suppression, creators are forced to speculate

Why We Believe This Matters

Algorithmic censorship does not only harm creators.  It:

• Distorts public discourse
• Punishes lawful speech without notice or appeal
• Disproportionately harms mission-driven organizations
• Undermines trust in the platforms themselves

When visibility is removed quietly, without explanation or recourse, the issue is no longer technical.  It is ethical.

Closing the Year

As we close out the year, we are naming this for what it is.  Censorship does not require intent. It requires outcome.  If you suspect suppression, do not assume you are imagining it. Watch the ratios. Document the patterns. Persist calmly. Treat visibility as something worth defending — not begging for.

Silence imposed without explanation is censorship by default.  And it should not be a problem we carry forward into the new year.