You Got a Mean Anonymous Message. Here Is Exactly What to Do

Someone typed something cruel into a box, hit send, and closed the tab. It cost them maybe ten seconds of effort. You are going to carry it around for days.

That asymmetry is the ugly math of anonymous messaging, and no amount of filtering fully solves it. Secret Messageruns a moderation pipeline that pulls threats, slurs and self-harm bait out of your main inbox before you see them. But a message can be genuinely vicious without containing a single word a filter can match. "Everyone is nicer to you than they actually feel" has no bad words in it. It still lands.

So here is a playbook rather than a pep talk. What to do in the first ten minutes, how to work out which of four very different things you are actually holding, and a separate response path for each one.

The first ten minutes

Whatever you do in the next few minutes, you do it with adrenaline in the driver's seat. Your judgment is temporarily worse than usual and it does not feel worse. That is the trap. Four rules cover it.

1. Do not respond while you are still hot

There is no reply button pointed at the sender. Inbound messages on Secret Messageare one-way by design, and that is true of most apps in this category, so "responding" in practice means going to your Story or your feed and firing something off to everyone you know at once. That is a far bigger action than it feels like at 11:40 at night, and it is the one people most often wish they could take back.

The urge to answer immediately is not weakness or thin skin. It is the completely normal response to being hit by something you cannot hit back. It just makes for bad decisions.

2. Screenshot it before you touch anything

Take the screenshot now, with the timestamp visible, even if you are ninety percent sure you will never need it. People delete in a rage and then, two weeks later when the same person is still going, realize they destroyed the first entry in what turned out to be a pattern. A screenshot costs you nothing and it is the only step that is genuinely hard to redo later.

Worth knowing while you decide: the message itself auto-deletes after 90 days whether you touch it or not, and the coarse sender metadata we hold for abuse handling is purged at 180 days. Both clocks are running from the moment it arrived, so the window for doing anything formal is finite even if you do nothing.

3. Do not post it publicly yet

We build a 1080x1920 Story image out of any message with one tap. It is a great feature for the funny ones and a loaded gun for this one. There is a whole section below on why clapping back usually backfires. For tonight, just leave it in the inbox.

4. Give yourself a deadline instead of a decision

You do not have to resolve anything right now. Close the tab, go do something physical for twenty minutes, and look again tomorrow. Anything you do inside the first hour is a reaction. Anything you do after sleeping on it is a decision. The message will still be there in the morning, you will read it in a different voice, and you will be choosing a response rather than being dragged into one.

If you want a rule you can actually follow at midnight, make it this: nothing leaves your phone tonight. No Story, no group chat, no screenshot forwarded to the friend who will get you worked up. You can do all of that tomorrow if it still seems like a good idea, and the test of whether it was a good idea is precisely whether it still seems like one.

Triage: four different things wearing the same face

"Mean anonymous message" describes four situations that need four completely different responses. Sorting yours correctly is most of the work.

Rude

Most mean anonymous messages are this. An opinion about your face, your voice, your content, your relationship, delivered by a person who felt clever for four seconds. There is no plan behind it and usually no history.

The correct response is boring on purpose: delete it, block the sender, close the app. Do not analyze the phrasing for clues about who wrote it. That analysis is the most addictive and least productive thing you can do, and it is almost always wrong. We wrote a whole piece on why in can you find out who sent an anonymous message, and the short version is that you will convict an innocent friend in your head and never know it.

One reframe that actually helps, if you want one: the message is not a verdict from the world. It is one person with a phone, a grudge or a boring evening, and no obligation to be fair. The anonymity is not evidence they were being brave enough to tell the truth. It is evidence they were not willing to say it with their name attached.

Harassment

This one is defined by pattern, not intensity. A single brutal message is rude. Six mediocre ones over four days, each showing that the sender is watching what you post, is harassment. The markers are repetition, escalation when you ignore it, references to things happening in your actual life, and a sense that the sender is keeping score.

Be honest with yourself about the rate limits here. We cap sending at 5 messages per minute and 20 per recipient per day from one sender, which stops floods but does not stop persistence. A determined person can come back tomorrow. Blocking works off the network connection a message arrived on, which handles the casual case and not the motivated one, because switching from wifi to mobile data is enough to look like somebody new. We would rather say that plainly than let you find out on day three.

So the strategy for harassment is not one button. It is: keep every screenshot in one place with dates, report each message so there is a record on our side, block after each one anyway, mute the keywords that keep appearing, and tell one person in real life what is happening. That last step is the one people skip and the one that helps most.

Credible threat

A threat is credible when it gets specific. Generic hostility ("you should watch your back") is horrible and usually empty. A message that names your street, your school, your shift at work, your car, the people you live with, or a time and a method is a different category entirely and needs a different reaction.

Do not delete it. Do not reply to it. Do not go looking for the sender. Report it in the inbox so it is flagged and the associated records are preserved, screenshot it, and then contact law enforcement. If the threat is immediate, that is a call to emergency services, not a report form and not an email to us.

Also tell someone physically near you today. A parent, a partner, a roommate, a manager, a campus safety office. Threats work by isolating you, and the fastest way to take the power out of one is for a second person to know about it within the hour.

Someone in crisis

Occasionally the message is not really about you. Someone is drowning and it comes out as cruelty, or as a description of their own pain handed to a stranger because they had nowhere else to put it. This is covered separately below, because it deserves more than a table row.

The buttons that actually help

Everything in this list is free and lives in your inbox. The safety page has the full reference, but here is what each one is genuinely good for.

• Report. Tap the ⋯ menu on a message, then Report. This flags it for review and creates a record. Emailing a screenshot to a friend does not. If you ever need a paper trail, this is where it starts.
• Block. Cuts off further messages from the network connection that one arrived on. Effective against the casual sender, beatable by someone who changes networks. Use it every time anyway, since it raises the effort required and most people are not that committed.
• Mute keywords. The most underrated control we ship. Any message containing a muted word goes silently to your Filtered folder instead of your inbox. If the same insult, the same nickname, or the same subject keeps showing up, mute it and the messages stop reaching your eyes without you having to police anything.
• Delete. Gone from your inbox. Do this freely for rude messages and never for anything you might report to police. Screenshot first if there is any doubt.
• Make your board private. One toggle. It stops new messages without destroying your board or your link. Think of it as a door you can close for a week rather than a decision you have to make forever.
• Take the link out of your bio. Underrated and totally reversible. Your board and your inbox survive. Your 4-digit PIN gets you back in whenever you want. Nothing is lost by disappearing for a week.

One honest note about the Filtered folder, because knowing where the line sits will save you from expecting the wrong thing. It is a pattern matcher, not a mind reader: a list of shapes somebody wrote down in advance, matched against your message on our own servers. The safety page itemises which shapes. Everything that matches lands in Filtered instead of your inbox.

What that leaves untouched is the entire category of cruelty that keeps its language clean. If something nasty reached your inbox, it is not because nobody was watching. It is because "I heard what people say about you when you leave" contains no word, phrase or shape that any rule can flag without also flagging a thousand harmless sentences. This is the honest ceiling on filtering, and it is why the mute-keyword control matters so much: you are the only person who knows which specific harmless-looking words are being used as weapons against you. A filter cannot learn that a particular nickname is the one that guts you. You can tell it.

Should you clap back publicly?

Almost always no. Not because you should be the bigger person, but for five concrete reasons.

1. It is the payoff they were waiting for. The sender is watching your Story. That is the entire reason they used your link instead of shouting into the void. A screenshot with a savage caption is the receipt confirming the message landed.
2. You amplify the insult. One person wrote it and one person read it. Post it and your entire follower list reads it, in your voice, attached to your face, with your caption vouching that it got to you. You just did the distribution for free.
3. It advertises that you react. Anyone else browsing your bio link now knows that a cruel message gets a response. That is the exact signal you do not want to send.
4. Vague callouts hit the wrong people.Post "some people need to grow up" and three friends will quietly wonder if you meant them. You will spend the next day reassuring people who did nothing.
5. You might be wrong about who sent it. If your caption implies a suspect and the suspect is innocent, you have created a second problem that outlives the first one, and you cannot un-post it.

There is a version of posting that works. It happens later, when you genuinely feel nothing, and it stands on its own without the screenshot. Naming a pattern ("heads up, someone has been spamming my link, ignore anything weird") is fine. Turning a message that hurt you into content while it still hurts is not, and the difference is obvious to your audience even when it is not obvious to you.

When to bring in a parent, a school, or an employer

The first person to tell is not an institution. It is one human being who knows you. Say the sentence out loud to somebody. It deflates faster than you expect and you stop carrying it alone.

Beyond that, escalate when there is a pattern or a real-world link. If you are under 18 and the messages are sustained, sexual, or threatening, tell a parent or another trusted adult. It is not tattling and you will not get your phone taken away for reporting something that was done to you. If you are a parent trying to figure out how to react to this without overreacting, our parent's guide to anonymous messaging apps is written for exactly that conversation.

Schools sometimes can act when the sender is plausibly a student, even when the messages happen off campus. Plenty of anti-bullying policies are written to cover digital conduct that affects the school environment, so it is worth asking what yours actually says instead of assuming the answer is no. Bring dated screenshots and describe the pattern rather than leading with the single worst message. A pattern is something an administrator can act on. One insult usually is not.

For work, the same logic applies with HR, and the thing to document is impact and repetition. If the mess started with an anonymous feedback channel your team set up, the failure is usually structural rather than personal, and how to use anonymous feedback without it turning toxic covers how to fix the channel itself.

When to contact police, and what to preserve

We are not lawyers and this is not legal advice, but here is the practical shape of it. The categories that tend to get a report taken seriously are threats of violence, stalking or a sustained campaign, extortion and sextortion, doxxing, and any sexual content involving a minor. Someone calling you ugly is not one of them, and walking into a station over a single insult will mostly cost you an afternoon. If you are unsure which side of that line you are on, that uncertainty is itself a good reason to go ask, rather than a reason to talk yourself out of it.

If you are in the narrow set, preserve this before you do anything else:

1. Screenshots of every message with visible timestamps. Full screen, not cropped to the text.
2. The exact URL of your board, which looks like secretmessageonline.com/b/your-id.
3. A plain list of dates and times, written by you, in order. This is the document investigators will actually use.
4. The original messages, left undeleted in your inbox until you are told otherwise.

Then move reasonably fast, because two different clocks are running. The message body is deleted from our database and our backups 90 days after it arrives. The coarse sender metadata, which is the part a legal request would actually care about, is purged at 180 days. That is the outer edge. A request that lands after it has nothing left to attach to, and no amount of goodwill on our end can conjure back a record that has already been purged.

Which is the real argument for your own screenshots. They are the only copy of the message that survives on your terms rather than ours, and they are what lets an investigator ask us a specific question about a specific message while the metadata still exists. A police report filed in month seven that says "there were some horrible messages last spring" is not a case. Dated screenshots plus a board URL plus a timeline is.

Set your expectations honestly about what happens next. We never show sender identity, IP or location to you, and there is no feature we could sell you that would. What exists server-side is an IP address, a country, and a device type, held for abuse handling and lawful requests, which we describe in the privacy policy. An IP is a network, not a person, so turning it into a name requires legal process aimed at an internet provider. That path is real, it is slow, and it is driven by investigators rather than by you. If a formal request is coming, an officer can reach us directly at [email protected].

If the message involves self-harm

This shows up in two shapes and they are not the same problem.

The first is a message aimed at you telling you to hurt yourself. Our severe-content detection is built to catch that language and route it to Filtered before it reaches your inbox, in English and several other languages, but nothing catches everything. If one gets through, report it and block. And if it landed harder than you expected, that is not an overreaction and it is worth saying to a person you trust today rather than sitting with it. In the US you can call or text 988 for the Suicide & Crisis Lifeline, any hour. Elsewhere, your local emergency number and your country's crisis line exist for exactly this and you do not have to be in immediate danger to use them.

The second shape is a sender describing their own crisis, or threatening self-harm if you do not respond. That is genuinely hard, because the concern might be real and the pressure is real either way. Two things are true at once. Someone may be struggling, and you cannot be their crisis service. You cannot call anyone for an anonymous sender because you do not know who or where they are, and no feature we could build would change that.

What you can do is small and legitimate: report the message so it is on record, and if you have any idea who it might be, tell an adult who could plausibly check on them. If it is a stranger, the honest limit is that you can post a general note with a crisis number and nothing more. Then take care of yourself, because absorbing someone else's emergency through a text box takes a real toll. Being unable to fix it is not a failure of yours.

The week after

The thing nobody mentions is the tail. The message stops being an emergency after a day and then just sits there, surfacing while you brush your teeth. That is normal and it fades, though not on a schedule you get to pick.

Two things help. One is that the auto-delete does real work here. After 90 days it is not in a folder somewhere waiting to be reread, which means the reread stops being an option and starts being a memory, and memories get blurrier while saved screenshots do not.

The other is knowing what your own head is doing to the ratio. Go back and count your inbox. Whatever the actual split between kind messages and cruel ones turns out to be, the cruel one is the one you can quote from memory and the kind ones have already blurred into a pleasant smear. That gap is not information about what people think of you. It is just how memory weights things that sting.

Then decide about the link, and take the decision seriously in both directions. Plenty of people keep it and are glad. Plenty of people take it out of their bio and feel lighter within a day. There is no version where keeping a link open to strangers is an obligation, and there is no version where closing it means you lost. Your board, your PIN, and your Secret Message inbox will be there whichever way you go.

Where to get help outside this site

We are a messaging product, not a crisis service, and the right move when something is genuinely frightening is to talk to people whose actual job it is. These are free, staffed by trained humans, and not run by us:

• 988 Suicide & Crisis Lifeline (US) — call or text 988, or use the chat at 988lifeline.org. For you or for someone who sent you something worrying.
• Crisis Text Line — crisistextline.org, if talking out loud is not something you can do right now.
• Samaritans (UK and Ireland) — call 116 123, or samaritans.org.
• Anywhere else — findahelpline.com lists verified lines by country.

If there is a credible threat to your safety, contact local emergency services first and us second. Our safety page explains what we can preserve and hand over when authorities ask.