A reading history is deceptively complicated. One service may store a single “date read”; another allows several reading sessions. One treats a rating as a number, another as a private note. Edition identifiers, audiobook durations and rereads do not always travel cleanly.
Keep the untouched source
Export from the existing service and save that file without editing it. Make a working copy for cleanup. If the source offers several formats, choose a readable table such as CSV for inspection and keep any richer JSON archive as well.
Goodreads documents its book import and export route; other services differ, so use their current help pages rather than an old tutorial. For identifier checks, Open Library's public APIs explain how its book and edition records are structured.
Map meaning, not just columns
- Separate title and author from edition-specific identifiers.
- Keep dates in an unambiguous year-month-day format.
- Preserve original notes in a plain-text column.
- Mark rereads explicitly instead of silently overwriting the first date.
- Do not upload private notes until you know their visibility setting.
Test ten varied books
Choose a paperback, ebook, audiobook, translated edition, series volume, reread and unfinished book. Import those first. Look for duplicates, missing diacritics, changed dates and ratings that have shifted scale.
Only after the test looks right should you import the full working copy. If the destination adds unexpected public activity, delete the test and revisit privacy settings before continuing.
Maintain a master ledger
Once the move is complete, export again and compare counts. Store a simple master ledger outside the service with the fields that matter to you. It will not preserve every social interaction, but it can preserve the reading record itself.
The aim is not perfect synchronisation between platforms. It is a comprehensible history you can inspect, back up and move again.