Deciding to move your data out of Google is easy. The move itself is a project, because “my Google data” is really four different collections with four different exit routes - and treating them as one big download is how people lose photo metadata and discover, months later, that a spreadsheet did not survive the trip.
First, inventory what you actually have
Fifteen minutes of looking before anything moves:
Uploaded files in Drive - PDFs, images, zips, anything that was a file before it met Google. These move cleanly; they are just bytes.
Google-native documents - Docs, Sheets, Slides. These are not files and cannot leave as themselves; any exit converts them, with losses. This is the category that needs a decision, not a transfer.
Google Photos - usually the largest collection by size, with its own export quirks around metadata.
Everything else - Gmail, Calendar, Contacts. Out of scope here, though Takeout (below) covers them too.
In Drive, filtering by file type gives you the proportions in a minute. The ratio of native documents to real files determines how much of this project is mechanical and how much is judgement.
The two export routes, and when each fits
Server-side import, for Drive files. The destination pulls your files directly from Google, datacentre to datacentre, without any of it flowing through your machine. We support this: connect your Google account, choose folders, and the import runs server-side while your laptop stays closed. For any real quantity of data this is the difference between an afternoon and a fortnight of babysitting uploads - our Drive migration guide walks the full sequence, including the cleanup worth doing before anything moves.
Google Takeout, for everything else. Takeout produces downloadable archives of any Google product, and it is the only route for Photos and the native-document exports. Expect it to arrive as multiple multi-gigabyte zips, sometimes days later for large accounts. If your home connection makes re-uploading those archives painful, note that a download link can be pulled straight into cloud storage without touching your machine.
The Photos problem, specifically
Takeout is the only complete export for Google Photos, and it has a famous quirk: the photos arrive with much of their metadata - capture dates, locations, album membership - in JSON sidecar files next to the images rather than inside them. Files whose EXIF timestamps were edited in Google Photos are particularly affected: the corrections live in the sidecars.
Practical consequences. Budget time to rejoin metadata with one of the open-source tools built for exactly this (search “takeout metadata merger” - the ecosystem is healthy because the problem is universal). Do the merge before importing the library anywhere, while the sidecars still sit beside their images. And spot-check dates on a sample afterwards - a library that sorts by upload date instead of capture date means the merge did not take.
Going forward, automatic photo backup from the phone replaces the Photos pipeline: originals with their metadata intact, straight from the camera roll.
Decide about the native documents honestly
Exporting converts Docs to .docx, Sheets to .xlsx, Slides to .pdf or .pptx - and conversion is lossy: comments, revision history and complex formulas may not survive. Our migration guide covers this in detail, but the decision compresses to: export and spot-check the archives and finished work; keep genuinely live collaborative documents in place until you have chosen what replaces the collaboration, not just the storage. A partial exit here is a correct outcome, not a failure.
Verify, run in parallel, then cancel
The exit is done when the new location has been the working copy for a few weeks, not when the transfer finishes. Compare counts and total sizes, open the files you would most hate to lose, check the deep folders arrived intact. Rebuild only the share links still in use - with expiry dates this time - and let the old ones lapse.
Keep the Google account alive but read-only in spirit through the parallel period; gaps announce themselves when you go looking for something specific. Then downgrade the storage plan, and the project is genuinely finished.
The sequence on one line
Inventory, clean up, import Drive server-side, Takeout for Photos with the metadata merge, decide the native documents deliberately, verify by opening things, run parallel, cancel. None of it is difficult; all of it goes better in that order.
Frequently asked questions
How do I move my files out of Google Drive?
The efficient route is a server-side import: connect your Google account at the destination and files copy directly between datacentres. Google Takeout is the fallback for everything an import cannot reach, such as Photos.
Does Google Takeout keep photo metadata?
Partially. Takeout exports much of the metadata, including edited dates and album membership, in JSON sidecar files next to the images, so plan to merge it back with a dedicated tool before importing the library anywhere.
What happens to Google Docs when I leave Google?
They must be converted, usually to docx, xlsx or PDF, and conversion is lossy: comments, revision history and complex formulas may not survive. Export finished work, and keep live collaborative documents in place until you have replaced the collaboration itself.
