Make room for the conversation you want
Twitter reply settings: choose who can join the conversation
A question can get views while some readers cannot answer it. Check who is allowed to reply before judging the response to your post.
On this page
Check who can answer before judging the silence
You ask a question on Twitter, get views, and receive almost no replies. Before rewriting the question, check whether the people you wanted to hear from were allowed to answer.
On X/Twitter, reply settings control participation under a post. Choose them around the conversation you want: broad feedback, a discussion with familiar accounts, or an exchange with specific people. A smaller conversation can be intentional. It does not tell you whether the writing was good or bad.
What each Twitter reply setting allows
X's conversation controls documentation describes these choices:
| Setting | Who can reply | Trade-off to consider |
|---|---|---|
| Everyone | On a public post, anyone; on a protected post, your followers. | New readers can contribute, but you may have more replies to review. |
| Accounts you follow | Accounts you follow, plus accounts mentioned in the post. | Someone with useful feedback may be outside your existing circle. |
| Only accounts you mention | Accounts named in the post. | This suits a specific exchange, not a broad request for feedback. |
| Only verified accounts | Verified accounts, plus accounts mentioned in the post. | Verification is an access condition, not evidence that a reply will be useful. |
"Accounts you follow" means people you have chosen to follow, not everyone who follows you.
Change the setting before or after posting
In X's composer, select the reply control beside the globe icon before publishing and choose the intended audience.
For an existing post, open its More menu, choose the option to change who can reply, and select a new audience. You do not need to delete and recreate the post just to change this setting.
The restriction concerns replies. It does not itself disable likes, reposts, or poll votes, and quote posts do not inherit the original reply restriction. Other account restrictions still apply.
Reply controls do not make a public post private
Choosing a narrow reply audience is not a substitute for protecting your posts. X's public and protected posts guide distinguishes public visibility from access limited to approved followers.
Decide separately who should read a post and who should answer it. Do not publish sensitive information on the assumption that a limited reply audience makes it private. Keep your chosen privacy protections in place when reviewing participation.
For a public question intended to reach people you have not met, check that your reply audience matches that goal. For a conversation you need to keep bounded, fewer eligible participants may be exactly what you wanted. You do not have to widen access to make an analytics number look better.
Hiding one reply is a different action
A reply setting determines who can participate. Hiding a reply changes how an existing response appears in the conversation.
X explains in its replies and mentions guide that hidden replies can still be accessed through the hidden-replies icon, subject to visibility restrictions. The post author can unhide them. Hiding is not deletion or a way to make information confidential.
If your goal is to manage an existing response, identify that problem before changing who can join the whole conversation.
Keep reply settings beside your performance notes
Suppose you publish two similar questions, one open to everyone and one limited to accounts you follow. Comparing their reply counts alone leaves out a meaningful difference: the pool of people who could answer.
This is a hypothetical comparison, not a claim that one setting produces better reach. Topic, timing, wording, and audience also vary. If you change the setting midway through a conversation, note when you changed it instead of treating the whole post as one unchanged test.
Keep a short record in your own notes:
Post URL:
Purpose of the question:
Who I wanted to hear from:
Reply setting at publication:
Later change and time, if any:
When I reviewed the results:
What the replies helped me decide:
In PilotMyX, review the post text beside its captured impressions and reply count. Record the reply setting yourself; the exposed post data does not include it or a history of changes. Keep missing metrics distinct from zero, and compare results at a similar post age. The guide to when to judge a Twitter post explains why timing matters.
Also separate replies you write from replies your posts receive. They answer different questions about your activity.
Ask your agent for a review, not a verdict
If you use an AI agent with PilotMyX, give it your setting notes alongside the available results:
Review my original posts from the last 7 days using stored PilotMyX data.
Keep replies I wrote separate from replies received by my posts.
Use my supplied notes for reply settings and changes. Mark unknown settings
and unavailable metrics as unknown; do not infer them from low reply counts.
Show post text, publication time, metric freshness, impressions, and replies
where available. Explain which comparisons are limited by audience or age.
Suggest one clearer question to test. Do not claim a setting caused reach
or change settings, refresh data, create drafts, schedule, or publish.
If the right people could answer but the question still fell flat, work on the request you made of readers. If they could not answer, decide whether that boundary still serves the conversation you want.
Use your own account