TV Time shut down on July 15, 2026 — your data can still be recovered
The complete guide to saving years of tracking: export from TV Time, re-import everything into WatchTrail, and verify nothing is missing.
The export TV Time sent, before its shutdown, after a GDPR request is an archive of 53 CSV files. Only three are needed to rebuild your history. The first, tracking-prod-records-v2, has one row per watched episode with the TheTVDB episode ID and the exact watch date. The second, followed_tv_show, lists every show you followed, including the ones you never started. The third, tracking-prod-records, holds your movies, watched or to watch, with title and release year but no identifier. The fifty other files are settings, devices, notifications and comments: nothing to re-import there. On a real export, those three files held 4,673 episode rows, 174 followed shows and 322 movie rows (321 distinct titles, one duplicate). You do not have to sort them: drop the ZIP as is, WatchTrail recognises the three files and ignores the rest.
Measured on September 2, 2026 on a real account: 4,673 episodes dropped, 4,606 recovered, that is 98.6%, in 225 seconds. The 67 missing ones are explained: special episodes TheTVDB counts and TMDB does not know, and duplicates from TV Time itself, one show present twice in the export under two IDs. The 24 followed shows with no watched episode come back in the “to watch” list. Movies are matched by title and year: 316 out of 321 in the same run, 286 seconds for everything. Watch dates are preserved, so statistics and the yearly recap stay accurate. The import can be re-run: if the tab closes, run it again, no duplicate is created. A progress bar follows each show, and the final screen shows the exact count of what came in.
Every episode in the export carries a TheTVDB ID, which WatchTrail translates into a TMDB ID through an exact mapping, not a title search. The rare failures come from differences between the two databases: special episodes filed under season 0 on one side and absent on the other, or renumbered shows, like some long-running anime. Movies are a different case: the export only gives a title and a year, so matching is a search, and a namesake can slip through. In every case, the final screen lists the titles that were not found, and each one can be added in one click from the site search. On an export of 4,673 episodes and 321 movies, that list was seven episodes and five movies long.
Already have your export (the gdpr-data.zip file, or CSVs)? Keep it safe and go to step 2. Didn't request it in time? TV Time wiped all data from its servers on July 15, 2026: a GDPR request no longer returns anything. But the app kept a local copy of your library and progress on your phone, in a file named DioCache.db. It can be recovered from an iPhone backup (Finder or iTunes, unencrypted, or iMazing — the TV Time app folder, Documents/DioCache.db) or an Android backup (adb). Free tools read it and produce a file you can import here: see the dedicated question below.
On watchtrail.app: sign in by email (magic link) or Google, in 30 seconds. Free, no credit card.
Open the Import page (Profile → Import) and drop the TV Time ZIP as is — no need to open it or sort its 53 files, we do that. Separate CSVs work too. A progress bar follows each show; a 4,600-episode export takes under 4 minutes. The import is re-runnable: if it stops, run it again, no duplicates will be created.
The final screen shows the exact count: X episodes (of Y) and X movies (of Y) imported, with a detailed list of the rare unrecognized titles so you can add them manually. Your library, statuses (watching / completed), calendar and stats are immediately up to date.
The TV Time GDPR export ZIP as is (gdpr-data.zip), TV Time episode CSVs (with episode IDs), movie exports (Title/Year columns, Letterboxd format also accepted), and combined files. French and English headers are both recognized.
By their exact TheTVDB identifier (the one TV Time uses), converted to the TMDB database. This is reliable even when titles differ (localized names, renamed shows) and even when season numbering differs between the two databases (common for anime).
Yes, every imported episode keeps its original watch date — that's what makes your yearly recap and statistics accurate.
The standard TV Time export doesn't always include ratings. Whatever is present in your file gets imported; the rest can be filled in inside WatchTrail afterwards.
GDPR gives them one month to respond, extendable to three for complex requests. In practice, observed replies arrive within a few days to two weeks, by email, with a time-limited download link. Download the file as soon as it lands and keep it: it's your history, and the link expires.
An archive of about fifty CSV files, three of which really matter: « tracking-prod-records-v2 » (one row per watched episode, with the TheTVDB episode ID and the watch date), « followed_tv_show » (the shows you followed, even without a watched episode — they come back as « to watch ») and « tracking-prod-records » (your movies, watched and to-watch, with their year). The rest is settings, devices and notifications: we ignore it.
Server-side, yes: TV Time wiped all account data on July 15, 2026, and a GDPR request no longer returns anything. Phone-side, no: the app kept a local copy of your library and progress in a small SQLite database, DioCache.db. On iPhone it can be recovered from a backup (Finder or iTunes, unencrypted, or iMazing) inside the TV Time app folder, under Documents/DioCache.db; on Android from an adb backup or a rooted device. Three free, open-source tools read it and produce import files: tvtime-rescue-data (Simkl and Trakt CSVs), tvtime-backup-extractor (encrypted backups) and the "TV Time Liberator" converter (CSV and JSON). WatchTrail accepts the Liberator watched-episodes CSV, dates included, and the all-in-one format. Known limit: that cache mostly keeps the watched count per show and the episodes around your current position, so the exact date of an old episode may be missing. Failing that, search adds a show in one click and marking a whole season as watched is a single button.
Roughly one minute per 1,000 episodes, so a full history of several thousand takes a few minutes. If the tab closes or the connection drops, just run the same file again: every episode already recorded is recognised and skipped, no duplicates are created.
Your history is visible only to you unless you share it, and a private profile mode exists. More importantly, export works both ways: from your settings you can pull all of your data back out at any time, in a reusable format. What just happened to TV Time shouldn't be able to happen again without you being able to walk away with your data.
Also read: how WatchTrail compares with Trakt, Simkl and Serializd · the TV Time migration page.