Why hidden replies hurt the most
For most accounts under 10k followers, replying to bigger accounts is the number-one growth channel: it is free reach to thousands of strangers who already proved they care about your niche. If your replies are hidden from non-followers, that channel silently reads zero — you keep spending the effort and never see the return.
The cruel part: you cannot see it from your own account. Your replies look perfectly normal to you. Detection always has to come from the outside.
The seven signs
1. A friend who does not follow you opens the same thread and cannot find your reply — while strangers' replies are visible. This is the classic, strongest signal.
2. A logged-out (private window) view of the thread omits your reply while showing replies from accounts with similar or lower engagement.
3. Your reply only appears after clicking "show more replies" — the low-visibility bucket X uses for filtered content.
4. Engagement on your replies comes almost exclusively from your existing followers. Strangers never react, because they never see it.
5. You are consistently among the earliest replies, yet you always sort to the bottom of the thread. One instance is ranking; every time is filtering.
6. The reply count visible on the post does not grow even hours after your reply goes up (the count may only include replies the public can see).
7. The pattern repeats across many different threads and days. A single hidden-looking reply can be a glitch; a consistent cross-thread pattern is a restriction state.
How to confirm it
Combine the tests: open one of your recent replies from a private window, and ask one friend who does not follow you to do the same. If both of you cannot find the reply that you can see logged in, the pattern is confirmed for that thread.
Manual tests are noisy, though — regional views and caching interfere. An automated check measures the same signal in a repeatable way and documents it, so you can compare over time instead of guessing.
What helps
Hidden replies are usually spam-filter class and reversible. Stop anything automation-like, avoid posting identical replies under multiple posts, and keep engaging genuinely for a few days — then re-test. Many creators report recovery within a week or two of clean behavior.
If it persists, document a failing check and appeal through official channels. And until it clears, shift effort to your own posts and your newsletter — channels whose reach you can actually verify.