How Does Archiveteam Decide Which Websites or Data to Archive?

Updated

Archiveteam does not use a fixed scoring rubric. Its own description says it is a "loose collective" that reacts to "shutdowns, shutoffs, mergers, and plain old deletions," and its main page frames the mission as saving digital heritage before it is lost. In practice, selection is driven by perceived risk of loss, community tips, and the availability of a crawl method that volunteers can actually run. The site is an "offloading point and information depot" for projects "related to saving websites or data that is in danger of being lost."

The core criterion: risk of disappearance

The clearest signal is imminent or likely loss. Archiveteam says it has caught wind of shutdowns, shutoffs, mergers, and deletions since 2009. That means a site does not need to be famous to qualify; it needs to be at risk. A small forum with a shutdown notice can be a stronger candidate than a large platform with no announced changes.

How candidates surface

Two public channels are named on the main page:

  • Deathwatch — a place to flag anything shutting down.
  • IRC — an alternative way to tell the team directly.

Anyone can submit a lead through these routes. The main page also asks, "Anything shutting down? Put it on the Deathwatch or tell us on IRC!" This is the front door for new projects.

How projects are classified

Once a candidate is accepted, it is sorted by urgency and scope rather than by topic. The main page uses these buckets:

Category What it means Examples from the page
Short-term, urgent Immediate, time-critical work None currently listed
Medium-term Active but less time-pressured Meta Ad Library, US Government, Radio Free Asia, Radio Free Europe/Radio Liberty, Voice of America
Long-term Ongoing crawls with no clear end Distributed recursive crawls, Telegram, Twitch, URLTeam, URLs
Recently finished Completed work Listed under "Recently finished projects"
Upcoming Announced but not yet started Listed under "Upcoming projects"

The medium-term list shows that news value and public-interest significance matter. Radio Free Asia, Radio Free Europe/Radio Liberty, and Voice of America are described as media organizations "at risk of closure," which is both a risk signal and a cultural/political one. The US Government project is similarly framed as archiving government material.

Technical feasibility shapes the list

A site also has to be crawlable in a way the team can distribute. The long-term list includes "Distributed recursive crawls" for "large websites that have no clear structure that can easily be split into work items." That entry is essentially a description of a technical problem: if a site cannot be divided into work items, it needs a different crawling approach. So selection is not only "is this worth saving?" but also "can we actually save it with the tools and volunteers we have?"

What this means if you want to nominate something

If you have a site in mind, the practical path is:

  1. Check whether it is already listed under an active, finished, or upcoming project.
  2. If it is shutting down or being deleted, add it to the Deathwatch or raise it on IRC.
  3. Expect the team to weigh urgency, public interest, and whether the crawl can be split into work items.

There is no published formula beyond that. The main page presents the process as reactive and community-driven, not as a formal review board.