Tweets are load-bearing citations now. They appear in news stories, court filings, academic papers, and arguments with your brother-in-law — and they are also among the most perishable objects on the web. Posts get deleted, accounts get suspended or renamed, and the platform's own access rules change yearly. If a public post matters to you — as evidence, as a source, as history — the time to preserve it is the moment you see it. Here are the legitimate ways to do that, and what each one is actually worth.
Why screenshots aren't enough
A screenshot is instant, convenient, and evidentially almost worthless. Fake tweet images are trivial to generate, which means your genuine screenshot inherits the credibility problem of every fake one — it proves only that pixels exist. Screenshots also omit the machine-readable context: the URL, the timestamp's time zone, whether the post was later edited. Keep taking them for personal reference; just understand they're a memo to yourself, not a record anyone else must believe. (The verification flip side — how to judge other people's screenshots — is covered in our embeds explainer.)
Web archives: the real preservation tools
Two public services do the heavy lifting for the whole internet:
- The Wayback Machine (web.archive.org). The Internet Archive's crawler — the closest thing the web has to a public records office. Paste a tweet URL into its "Save Page Now" feature and it captures the page, timestamps it, and serves it forever at a stable URL. Coverage of individual tweets is imperfect (the platform's login walls have periodically hindered crawling, a saga we recount in the login wall story), so always verify the snapshot actually rendered the content before relying on it.
- archive.today. An independent service that takes on-demand snapshots and has historically handled login-walled and JavaScript-heavy pages better than automated crawlers. It produces both a browsable copy and a screenshot, timestamped, at a permanent short URL.
The habit that makes these work: archive at the moment of encounter. An archive link created before a controversy erupts is dramatically more credible than one created after — and deletion usually follows controversy within hours.
The platform's own tools
Bookmarks are private, unlimited, and invisible to the post's author — genuinely useful as a reading queue. But a bookmark is a pointer, not a copy: when the post dies, the bookmark dies with it. The same applies to likes-as-storage. For anything you'd mind losing, a bookmark is a reminder to archive, not the archive. Your own posts are a different story — the platform will hand you a complete export of everything you've written, which we walk through in our data archive guide.
Citing tweets properly
If you're preserving posts for research or journalism, capture the full citation while it exists: the author's handle and display name, the complete text, the timestamp with time zone (the platform renders times in your local zone — note what that is), the direct URL, and the date you accessed it. Academic styles (APA, MLA) have tweet citation formats; all of them assume you recorded these fields at capture time. Pair every citation with an archive link and you're immune to deletion.
The etiquette and ethics question
Archiving public posts by public figures on matters of public interest is the digital equivalent of clipping newspapers — essential accountability infrastructure, practiced by journalists, researchers, and courts. Archiving a private individual's posts to build a dossier is something else, even when technically legal; the fact that someone's posts are public doesn't make every use of them fair, a distinction we explore in what your public profile reveals. A reasonable compass: archive for accountability, evidence, and history; don't archive to harass. The tools don't know the difference, so you have to.
The thread problem
One trap worth knowing: archiving the first post of a thread does not archive the thread. Each post in a chain is its own page with its own URL, and an archive snapshot of post one typically captures only whatever replies happened to render — often nothing. For a thread that matters, archive each post's URL individually, or at minimum the first and the most important ones, and note the sequence in your records. The same page-by-page logic applies to a profile: a snapshot of someone's profile page preserves only the posts visible at capture time, not their history. Archives preserve pages, not accounts — scope your captures accordingly.
A practical workflow
- See something that matters → copy the tweet URL immediately.
- Submit it to the Wayback Machine and archive.today (both — redundancy is the point; each service has different blind spots and failure modes).
- Verify both snapshots rendered, then save the archive URLs wherever you keep notes, alongside the original URL and capture date.
- Optionally screenshot for quick personal reference, treating it as a convenience copy only.
Ninety seconds, total. The deleted-tweet news cycle runs weekly, and the difference between "I saw it, trust me" and a timestamped public archive link is the difference between an anecdote and a record.