How Does Google Takeout Work?
Google Takeout collects your data from each Google product server-side, packages it into downloadable archives, and emails you a link. When you request an export, Google queries every product you selected, serializes the results into formats like JSON and MBOX, compresses them into ZIP or TGZ files, and generates a private, time-limited download link delivered to your inbox or a cloud drive.
That is the short version. This page opens the hood: what happens step by step when you click Create export, which formats you get and why, why the job runs asynchronously instead of instantly, and how the raw output differs from a readable report. If you just want the click-by-click steps, see our Google Takeout instructions. If you want to know what Takeout even is, start with what is Google Takeout.
What happens when you request an export
Takeout is not a download button that streams your data on the spot. It is a job you submit to Google's servers, which then run through several stages before anything reaches you:
- Query each product. Google reads the data store behind every product you selected, from Location History to Gmail, pulling the records tied to your account.
- Serialize the results. Each product's data is written out in an export format: JSON for most activity, MBOX for mail, original files for Photos and Drive.
- Bundle into archives. The serialized files are compressed into ZIP or TGZ archives so they download as one package instead of thousands of loose files.
- Split by size limit. If the total exceeds the size cap you chose, Google splits it into multiple numbered archives.
- Generate a download link. Google creates a private, time-limited link and emails it to you, or drops the archive into the cloud drive you picked.
At no point is anything removed from your account. Takeout produces a copy. The originals stay exactly where they were.
Why the export runs asynchronously
Small exports could in theory be built instantly, but a full Google account is anything but small. Gathering years of activity across dozens of products, serializing it, and compressing gigabytes of Photos and Gmail is heavy work. Google runs it as a background job on its own servers so it can process at scale without tying up your browser.
That is why you get an email later instead of a file right away. A lightweight export of just Search or Location History can finish in a few minutes. A full account with Photos, Drive, and Gmail can take hours. For a breakdown of what drives the wait, see how long Google Takeout takes.
What formats you get, and why
Takeout does not use one universal format. Each product exports in whatever format best preserves its data, which is why an unzipped archive looks like a grab bag of file types:
- JSON for most activity data: Location History, Search and Web Activity, YouTube history, Chrome. It stores structured records with timestamps and fields, which is precise for machines but unreadable for people.
- MBOX for Gmail, the standard mailbox format that packs thousands of messages into a single file.
- Original files for Google Photos and Drive, so a JPEG stays a JPEG and a document keeps its native type, each paired with a JSON sidecar for metadata.
- ICS for Calendar events and vCard or CSV for Contacts, the portable standards those tools use.
- ZIP or TGZ as the outer wrapper. ZIP opens natively on Windows and Mac. TGZ compresses tighter and suits very large exports.
The logic is portability: each format is chosen so the data can be re-imported elsewhere. The tradeoff is readability. To see how the raw JSON is structured, read our Google Takeout JSON viewer guide.
Sample data · your real report will look like this
Location History
Free previewSearch Activity
Free previewYouTube Activity
Free previewChrome Browsing
Android Activity
Gmail
A snapshot in time, not a live sync
A Takeout archive captures your data at the moment Google builds it. It is a photograph, not a video feed. Anything you do after the export was generated will not appear in that file. If you want your archive to reflect newer activity, you request a fresh export.
For people who want ongoing copies without remembering to do it, Takeout offers scheduled exports. Pick the recurring option and Google will build a new export every 2 months for a full year, six snapshots in total, each delivered the same way as a one-off export. It is the closest Takeout gets to automatic backups, though each snapshot is still a point-in-time copy rather than a live sync.
How the size limit and splitting work
Before Google builds the archive, you choose a per-file size limit, commonly 1 GB, 2 GB, 4 GB, 10 GB, or 50 GB. If the whole export fits under that cap, you get a single archive. If it does not, Google splits the output into several numbered files that you download separately and can recombine after unzipping.
Smaller limits mean more files but each downloads more reliably on a shaky connection. Larger limits mean fewer files to juggle. For guidance on picking a limit and what a typical export weighs, see Google Takeout file size.
Already have your export? Read it now.
Drop your ZIP. Free preview. $19 $9 for the full report.
The delivery step: link, email, or cloud drive
Once the archive is built, Google needs to hand it to you safely. By default it emails a download link that stays valid for about a week and requires you to be signed in to your account to use. Because the file holds sensitive data, the link is private and expires on purpose. If it lapses, you regenerate the export.
You can skip email entirely and send the output straight to Google Drive, Dropbox, OneDrive, or Box. That is handy for very large exports, where downloading and re-uploading by hand would be slow. The mechanism is the same either way: Google builds the archive server-side, then places it wherever you told it to.
Raw output vs. a readable report
Everything above describes how Google produces the archive. What it does not solve is reading it. Unzip a real export and you face hundreds of folders of JSON, MBOX, and HTML, with timestamps as Unix milliseconds and coordinates as raw integers. The mechanism is excellent at portability and terrible at legibility.
That is the job TakeoutReader does. It takes the raw output the Takeout mechanism produces and parses it in your browser into plain English: how many searches, which years were busiest, top sites, a location timeline, and more. You get the meaning of the data without decoding the machine format by hand. You can start from the export at takeout.google.com and finish with a report in about a minute.
Frequently asked questions
How does Google Takeout work in simple terms?
When you request an export, Google queries each Google product you selected, serializes the results into export formats like JSON and MBOX, bundles everything into ZIP or TGZ archives, and emails you a time-limited link to download the files. Nothing is deleted from your account. It is a copy, built on Google's servers.
Why does Google Takeout take so long?
Takeout runs asynchronously on Google's servers. For a large account, gathering data across many products, serializing it, and compressing it into archives can take minutes to hours. Small exports finish in minutes. Gmail, Drive, and Photos are the slowest because of their sheer size. See our guide on how long Google Takeout takes for typical timings.
What file formats does Google Takeout use?
It depends on the product. Most activity data (Location History, Search, YouTube, Chrome) comes as JSON. Gmail is exported as MBOX. Google Photos and Drive files keep their original formats, with JSON sidecar files for metadata. Calendar uses ICS, Contacts uses vCard or CSV. The whole set is bundled into ZIP or TGZ archives.
Is Google Takeout a live sync or a one-time snapshot?
It is a snapshot in time. The archive reflects your data at the moment Google built it, not a live feed. If you want ongoing copies, Takeout lets you schedule recurring exports every 2 months for a year, which creates six snapshots automatically.
Why is the download link time-limited?
Because the archive contains sensitive personal data, Google keeps the download link valid for only about a week and requires you to be signed in to your account to use it. After it expires you can regenerate the export. You can also route the output straight to Google Drive, Dropbox, OneDrive, or Box instead of email.
Why are the exported files so hard to read?
The formats are built for machines, not people. JSON stores timestamps as Unix milliseconds and coordinates as raw integers, spread across nested folders. That is the gap TakeoutReader fills: it parses the raw output in your browser and turns it into a plain-English report with counts, dates, and top entries.
Turn your Google Takeout into a readable report
Drop your Google Takeout ZIP. Free preview in 60 seconds. The full report unlocks for $19 $9 one-time. No subscription. No account.
100% private. Processed in your browser. Nothing uploaded. TakeoutReader is not affiliated with Google.