Find an old article on a dead site

A link you saved three years ago. You click it. The page is gone. The site exists but the article is missing. Or the site itself is gone and the domain redirects to a parked-domain landing page. Or the redirect chain eventually times out.

This happens constantly. Link rot is common on the open web, and news sites that go through ownership changes tend to fare worst. Most readers, when they hit a dead link, just close the tab. We want to show you what to do instead.

Four tools, in the order we reach for them. The first one solves most cases. The last one solves the hardest cases.

Tool 1: the Wayback Machine

The Internet Archive's Wayback Machine is the first place to go. It crawls a substantial chunk of the public web continuously and stores snapshots of pages with the dates they were captured.

To use it, paste the dead URL into the search box at web.archive.org. You will get a calendar view showing every date the page was captured. Pick a date close to when you remember reading the article. The snapshot opens at the URL web.archive.org/web/<date>/<original-url>.

Two practical notes. First, snapshots are not perfect. Sometimes the HTML is captured but the images or stylesheets are missing, so the page looks broken. The text content is almost always intact regardless. Second, some sites use robots.txt rules that ask the Internet Archive not to crawl them. The Archive respects those rules. In those cases, the calendar will be empty and you move to the next tool.

The Wayback Machine also has a browser extension that automatically captures the page you are reading. We have it installed because the cost is near zero and the future-self benefit is real. Every time a page later goes dead, the extension's prior capture is what saves the link.

Tool 2: archive.today

archive.today (also available as archive.ph, archive.is, and several other mirror domains) is a separate archive service operated independently from the Internet Archive. It often has captures of pages that the Wayback Machine does not, and vice versa.

The interface is the same. Paste the URL. Hit search. You get a list of capture dates. Click one to open the snapshot.

Two cases where archive.today shines. First, paywalled articles. archive.today captures the page as it appears to a logged-out reader, which often includes the full text on sites that show paywalls only after some scroll behavior. Second, sites that block the Wayback Machine via robots.txt sometimes have archive.today captures because the two services obey different rules.

If you find an article on a site you read frequently that is interesting and the site is small, save it to archive.today yourself by pasting the URL into the "save this page" box. The capture is permanent and shareable as a single URL.

Tool 3: the Memento Project federated search

The Memento Project at Los Alamos National Laboratory aggregates results across many web archives. The interface accepts a URL and a target date, then searches dozens of archive services and returns the closest snapshots from each.

This is the tool we reach for when the obvious archives have not captured the page. Memento queries archives we have never heard of, including academic and national-library archives that are not indexed by search engines and that we would not find on our own. For older content (pre-2015 in particular), it often returns hits the consumer-grade tools miss.

The interface is utilitarian. The results page lists each archive that has a capture of the URL near your date and links you out to view it. Click through to the one that looks closest.

Tool 4: the last-resort moves

If the three archives above all come up empty, you have a few options of decreasing reliability.

Search the article title. Sometimes the article was syndicated. A piece originally published on a small blog might also have appeared on Medium, on a content aggregator, or in a newsletter archive. Quote the title in Google or DuckDuckGo and see what comes up. The original site is dead but the syndicated copy may not be.

Search a distinctive phrase from the article. If you remember a sentence, paste it in quotation marks into a search engine. This finds verbatim copies, including ones on Substack newsletters, on Reddit comment threads where someone quoted the piece, or on Wikipedia citations.

Email the author. Many writers keep their own archive of their published work, even when the site that published it is dead. If the author is findable (most are, via their personal site or a current employer), a polite email asking whether they have the piece works more often than you would think.

The print archive. For long-form journalism that may have appeared in a magazine, the digital site may be dead but the print copy is in a library somewhere. WorldCat (the union catalogue) can tell you which libraries hold the issue. This case is uncommon in practice and often the most satisfying when it pays off.

What we have learned to do differently

When you find an article worth keeping, save the URL into your own archive at the moment you read it. We use a combination of saved Wayback Machine captures (via the browser extension) and a personal text file listing articles we will probably want to find again. A markdown file with title, URL, and a one-line note is enough.

The reason this matters: you cannot rely on the original publisher to keep the article alive. The publisher's incentive to keep their archive working drops to near zero the moment the article stops earning traffic. Yours does not. If the piece matters to you, treat the URL as a thing you are personally responsible for preserving.

Most dead links can be recovered in under a minute with the first two tools. The last two are worth remembering for the cases where the first two fail.