Most businesses that have been online for some time have lost a website at some point: a hosting account that lapsed, a redesign in which the old blog was not retained, an agency that closed and took the server with it. Years of writing disappear, and usually no one notices until someone looks for a post they remember.
Over twenty-five years of web work, we have lost several ourselves: the Raincity Studios blog from Vancouver, and the blogs that followed. For a long time, we assumed they were simply gone.
They were not. Most of the text had been preserved in the Internet Archive all along.
What does the Wayback Machine actually preserve?
More than most people expect, though not everything. The Internet Archive has been crawling the web since 1996, and its Wayback Machine has been open to the public since 2001. Entering an old address displays a calendar of every date on which a copy was captured.
Coverage is uneven. A popular blog may have been captured hundreds of times, with every post. A small business site may have five snapshots of the homepage and nothing deeper. Images are inconsistent. Anything behind a login, a search box or heavy JavaScript is usually missing. Comments sometimes survive, sometimes not.
For a standard blog from the 2000s, however, the text of most posts is often available: titles, dates, author names and body text. That is the part that matters.
It is also worth remembering that the Archive is a non-profit organization, not a guarantee. In October 2024, it was hit by a cyberattack and went offline for several days. Anything that exists only there should be copied.
We also recommend using the Save Page Now feature for the pages of your current website that you would not want to lose. It is free and takes a few seconds per page, and it can prevent considerable difficulty later.
How can old blog posts be recovered?
The process is methodical and somewhat tedious, but it works.
- List every domain and address where the content has ever been hosted: old domains, subdomains, the /blog folder of a forgotten site.
- Look them up in the Wayback Machine and identify the snapshots with the most content. Blog archive pages and category pages are especially valuable, because they link to every post.
- Collect the post addresses, then retrieve each post from its best snapshot. For more than a handful, the Archive provides ways to list every captured address of a site, and a small script saves considerable manual work.
- Remove the old theme, menus and sidebars. Keep the title, date, author and text.
- Read every post before republishing it. Some have aged well; others have not.
The final step is the one most often skipped, and it is the most important. The task is not only to restore files, but to decide what one is still prepared to publish under one's name.
What should be left out?
Not everything deserves to return. Company news from 2006, "we will be attending this conference" posts and funding round announcements were useful the week they were published and are now noise. We recommend keeping anything that explains a concept, makes an argument or shows how a problem was solved, and setting the rest aside.
Why must the original authors be credited?
Because an agency blog is the work of many people. At Raincity, colleagues wrote many of the posts. Republishing everything under a single name would be easy, and it would be wrong.
Every recovered post should therefore carry the name of its author, the date of first publication and the place where it first appeared. If the authors can be reached, inform them. Most are pleased that their earlier work is being preserved. If someone prefers that a post remain unpublished, respect that choice.
This is also a matter of trust. Readers today are wary of content that appears without context, and with reason. A clear note stating "originally published in 2008 on this blog, by this person" tells them exactly what they are reading.
Does older content still demonstrate expertise?
Yes. One might question the effort, since a 2007 post about Drupal is unlikely to rank for anything.
It may not. A body of work, however, demonstrates something a new "about" page cannot: presence at the time, engagement with problems when they were new, and a visible record of which opinions held up and which did not. In our view, this is more convincing than any list of credentials.
Translation adds further value. In a bilingual region, an archive available only in English is half an archive. We prefer to present the French and English side by side, each reviewed by a person.
A note on dates
We have spent most of this month doing exactly this for this blog. Several hundred posts from The Standard, the former Raincity Studios blog, and from the Holistic Hybrid blog are now back online here, translated, with their original authors and dates. They can be browsed on the blog.
What should be done before a website redesign?
Preserve the existing content. The simplest way to avoid needing the Wayback Machine is not to lose the content in the first place. Many blogs disappear during a website redesign, when someone decides the old posts are clutter and does not set up redirects.
Before rebuilding anything, export every page. Keep the old addresses and redirect them to the new ones. This takes an afternoon. Recovering the content later takes weeks, and some of it never returns.