What to do before a harmful post disappears

A post can be deleted by the person who wrote it, removed by a platform's own moderation process, edited to soften or change its meaning, or simply buried once an account is renamed. None of these events give advance notice, and once any of them happens, the original is very rarely recoverable in full.

This is the single most important fact to plan around: you usually get one reliable chance to capture a post as it actually existed, and that chance is right now, while you are looking at it.

Act on what is in front of you before doing anything else, including reporting the post to the platform. Reporting first can prompt faster removal, which then means you have lost your own chance to capture the original.

• Take a full-page screenshot or, for video, a screen recording, uncropped, so the URL bar, timestamp, and surrounding content stay visible in the frame.
• Copy the exact source URL as text and save it separately, since a screenshot of a URL is harder to verify later than the URL itself.
• Note the date and time you captured it, along with your time zone, since the platform-displayed time and your capture time are two different facts worth keeping separate.
• Capture the account or profile as it currently appears: handle, display name, and bio, since these can change independently of the post itself.
• If the post has replies, quote-posts, or a comment thread, capture that context too, not just the post in isolation.

Only after this capture is complete should you consider reporting the post to the platform, contacting a lawyer, or taking any other action that might accelerate its removal.

Very little survives reliably, and understanding the limits of what might survive is more useful than assuming a safety net exists.

• Nothing, if no one captured it. Deletion by the author removes the post from the platform entirely, and there is no default recovery path for a stranger or the affected party.
• A cached snippet, sometimes. Search engines occasionally retain a short cached version of a page for a limited time after it changes, but this is inconsistent, often partial, and not something to plan around.
• An archived copy, if one was made before deletion. Archiving tools like the Wayback Machine only hold a copy of pages that were crawled or submitted before the content changed. They do not automatically archive every post on every platform, and social media coverage in particular is uneven.
• A repost or screenshot made by someone else. If another account reposted, quoted, or screenshotted the material before it was removed, that copy may still exist, though it usually lacks the original's full context and metadata.

Archiving tools are useful for two things: confirming that a page existed at a given point in time, and recovering a version of a page you failed to capture yourself. They are not a substitute for capturing material as you see it.

Limits worth knowing:

• Archived pages sometimes fail to capture dynamic content, such as comment counts, embedded video, or content that loads only after interaction, so an archived version may look thinner than the live page did.
• Timestamps on an archived page reflect when the archiving tool visited the page, not necessarily when the content was first published or last changed.
• Not every platform is consistently crawled or archivable, and some platforms actively block automated archiving tools, so absence from an archive does not mean a page never existed.
• A cached search-engine snippet is typically partial and temporary, and disappears once the search engine re-indexes the page.

None of this makes archives useless. It means they are a supplementary check, not a primary capture method, and they work best when you also submit the URL yourself at the time you see the content, rather than assuming a tool already has a copy.

Once a post starts spreading, the original is often not the only version that matters. Reposts, quote-posts, and mirrors on other platforms frequently outlive the original, and each one needs its own capture rather than being treated as covered by the first.

• Search periodically for exact phrases or distinctive wording from the original post, since verbatim reposts are the easiest to find this way.
• Capture each repost or mirror with its own source URL and timestamp, and note its relationship to the original where you can establish one.
• Watch for the same content resurfacing after a period of dormancy. A gap followed by a new appearance is itself worth recording, not just the reappearance.
• Keep each instance separate in your records rather than merging them into a single entry, since a reviewer may need to assess each one individually.

For a fuller workflow on tracking reposts and mirror sites specifically, see the mirror-site evidence preservation workflow. For the timestamp and time-source details that make a capture credible later, see the timestamping online evidence workflow. And for the complete method, from first capture through to a chronology a lawyer can use, see the complete guide to documenting online harassment.