When Google pulled the plug on Google Video back in 2011, the platform's closure sent a vast collection of user-generated content spiraling into digital obscurity — a cautionary reminder of how fragile cloud-hosted media can be when a technology giant pivots away from a product.
This week's episode of Reveal takes an unexpected detour from headline-grabbing exposés to explore a more intimate category of investigation: questions that may carry little geopolitical weight but carry enormous personal significance. The editorial team embarked on a mission to recover a short film produced by Garrison Hayes — a piece of digital creative work that vanished alongside the platform's shutdown.
What the investigation unearthed was a remarkable story of grassroots digital preservation. A dedicated coalition of volunteers mobilized to systematically rescue millions of videos along with their associated metadata, while the Internet Archive stepped in to provide the necessary storage infrastructure. The practical implication is significant: content creators who long assumed their early-era video uploads were permanently lost may, in fact, be mistaken. A recovery pathway exists — and the instructions below will walk you through it.
The recovered dataset — encompassing original metadata and functional Internet Archive retrieval URLs — resides within a large-scale plaintext file hosted at archive.org. Developers comfortable with scripting environments can parse this file using any preferred language. For non-programmers, the process is most manageable with some familiarity with text-handling tools.
- Upon visiting the archive link, you'll be presented with two download formats: a .7z compressed archive or a torrent file. For those unfamiliar with peer-to-peer torrent protocols, the .7z option functions similarly to a standard .zip archive and is the more straightforward choice. After decompression, you'll find a CSV file — however, its sheer size renders it incompatible with conventional spreadsheet applications like Excel. The most practical workaround for non-developers is to install a lightweight, free integrated development environment (IDE) such as Sublime Text, which is engineered to handle large file loads. Be prepared for a substantial initial load time given the file's scale.
- With the file open, deploy the built-in search functionality to query any identifiable keywords tied to your content — this could include the video's title, a fragment of its description, the email address linked to your original account, or even the source filename. Once you've isolated the relevant entry, locate the legacy hosting URL associated with the video. This string will contain "
googlevideo.com/videoplayback" and, in the vast majority of cases, begins with a pattern resembling "http://v8.cache7." To retrieve and view the video, append that recovered URL to the end of the Wayback Machine prefixhttps://web.archive.org/web/20041231235959id_/directly in your browser's address bar. The resulting URL will be notably lengthy — that's expected behavior.
Example:
https://web.archive.org/web/20041231235959id_/http://v1.cache5.googlevideo.com/videoplayback?id=5e903bf24b25e80d&itag=5&begin=0&ip=0.0.0.0&ipbits=0&expire=1309659155&sparams=ip,ipbits,expire,id,itag&signature=437990DFE530EA5FF22DDB2882240B5259AE749B.3C6C29148F23B4E28576C57014D57B7A0D500639&key=ck1&redirect_counter=1
An additional option for streamlining the retrieval process: create a personal copy of this purpose-built Google Sheets template, which automates the construction of the complete Wayback Machine URL. Select the full line of text corresponding to your video entry — a double- or triple-click should capture it entirely — then paste it into the first column of the spreadsheet. Crucially, ensure you capture the entire record from its opening negative sign and numeric sequence onward; this is what allows the tool to correctly parse associated metadata fields including the original email address, title, and video description.
Several important contextual notes about the archive's composition:
Approximately 92% of the archived video records include an upload timestamp. Among those dated entries, nearly all fall within a window spanning November 11, 2005 through October 26, 2010 — consistent with Google's decision to discontinue new uploads before eventually terminating hosting operations entirely. Two outlying entries carry reported upload dates of October 8, 2012 and February 4, 2011, representing anomalies within the broader dataset.
Find Your Lost Google Videos is a story from Reveal. Reveal is a registered trademark of The Center for Investigative Reporting and is a 501(c)(3) tax exempt organization.