Import from Plausible

Bring daily Plausible history into your Seean website from its full CSV ZIP export. You must own the Seean website. Imports preserve source aggregates and keep your live events, orders, tracking token and saved reports intact.

Export the correct archive

In Plausible, open the website’s Settings → Imports & Exports → Export Data and request the full CSV export. Download the resulting ZIP. Keep your own original copy. Do not use the export beside the dashboard chart: that is a different, limited report format. Plausible’s full export excludes data that it previously imported from other tools. See Plausible’s export instructions.

Note the Plausible website’s timezone at export time. The ZIP’s daily dates use that timezone but do not identify it; Seean cannot infer it from the filename.

Upload in Seean

Open your website, then Settings → Data → Import history. Keep Source on Plausible; it is selected automatically when the ZIP contains Plausible’s imported_* files.
Use Choose file to select the complete archive; its filename appears beside the button. Select its Export timezone. The default UTC must be changed if your source uses another timezone.
Choose Upload and import. Keep the page open until it says the upload is complete. Validation then continues in the background.
Wait for ready and review the numbers of days added and skipped. View rows opens the source tables on the same page; use the table dropdown and Load more rows to inspect them.

Each website keeps up to ten imports. Only one upload/import/removal can run for a website at once. A workspace can start up to three imports per hour. Demo websites cannot import. Later imports must use the timezone of the website’s earlier ready imports.

Import modes

File imports offer two optional controls above the import button:

Only source days from / through limits the import to a range of source calendar days (inclusive, in the export’s timezone). Days outside the selection are not stored - neither their totals nor their rows - and are reported as Days outside the selection; the saved original file still contains them. Leave both fields empty to import every day in the file.
Replace previously imported days lets this file’s totals and rows take over days that an earlier import owns. Off (the default) is Fill gaps only.

In every mode the importer skips a source day when Seean has any live event during that day, or when the day was still current or in the future when the upload began; live records are never overwritten and partial days are never added. In fill-gaps mode it also skips a day another saved import already owns, so a different archive containing the same days leaves the earlier accepted days intact. In replace mode the earlier import’s day is removed and this import’s day takes its place - the earlier import keeps its original file and source rows and its Days added / replaced counters shrink accordingly (a day handed back after a failed replace of an already replaced day is counted under Days added of the import that receives it). The same successfully imported ZIP is rejected if uploaded again. A ready import can therefore add zero days; its original source rows remain available for inspection.

All archive validation must finish before its accepted days become visible, and every import ends with a fresh Overview calculation (saved reports keep their earlier results). A failed import contributes no dashboard totals; if it was replacing days, those days return to the import they came from once cleanup finishes (also when you remove a replacing import before it finishes), a new Overview calculation follows, and a retry starts from exactly the earlier history. While a replacing import is still running, the days it is replacing belong to the unfinished import and are not counted. Source rows shown under View rows include skipped days as well as accepted days; they represent the original archive, not a second analytics dataset to add manually.

What is preserved

The supported archive contains imported_visitors, imported_pages and imported_sources CSV files. Filenames may also contain _YYYYMMDD_YYYYMMDD date suffixes. Recognized additional tables are entry pages, exit pages, custom events, locations, devices, browsers, operating systems and custom properties. Column headers are checked against supported Plausible export fields. Unsupported columns or formats fail validation instead of being silently dropped.

Daily visitors, visits, pageviews, bounces and total visit duration are preserved. Known breakdown columns are retained for source inspection, including source/UTM fields, geographic codes, browser/OS versions and custom-event counts. Geographic identifiers remain source codes; numeric city identifiers are not converted into city names. Missing optional tables do not imply zero activity. Custom events follow the source export’s event definitions; goals are not recreated.

Each ZIP is limited to 25 MiB compressed, 100 MiB expanded, 200,000 rows and 3,660 distinct calendar days. Dates must be valid calendar dates in 2000-2099. Numeric counts must be non-negative integers no larger than one trillion per field; text fields are limited to 4,096 characters and CSV records to 32,768 characters. Encrypted ZIPs, duplicate tables/rows, unsafe paths, corrupt checksums, unsupported columns and pageview totals that disagree between visitors and pages are rejected.

Read imported history

Accepted complete source days inside the selected date range contribute to the main KPIs, chart and available breakdowns. A note identifies imported days. Partially selected days are excluded. A source day follows its timezone, including 23/25-hour daylight-saving days; its totals belong to the day-start chart bucket rather than invented hourly activity.

Visitors adds Plausible’s daily counts to Seean visitors for the other days. It is not a deduplicated count of people across the period. Source bounces, visits and total duration join the corresponding live totals before rates are calculated. Source metric definitions can differ from Seean’s. Page, source, location, device and custom-event breakdowns use available source counts; missing optional tables remain unavailable.

Remove dashboard filters to include imported totals. Daily aggregates cannot reconstruct sessions, journeys or intersected filters. Roll-up, shared dashboards, widgets, PDF/email reports and API/MCP Overview reads include eligible history. Realtime, funnels, campaign analysis, revenue attribution and raw event/order exports still use original Seean records. Existing saved report results are preserved; new analyses use the current imports.

Errors, originals and removal

If an import fails, read its error and inspect the original export. Successfully uploaded original files remain available through Download original even after validation failure. Remove the failed import and upload a corrected full export. An upload/import that has not finished within thirty minutes is marked failed. A failed upload may have no saved file. If the upload request fails before validation starts, remove the pending import before retrying.

Remove import asks for confirmation, hides that import’s history immediately and removes its rows and saved ZIP in bounded background batches. It preserves live events, orders, tracking configuration and saved reports. Download your own copy first if you need to retain it. Removing a website also removes its imports and originals. Keep downloaded source URLs private; they grant access to the saved file.

After importing, compare representative daily pageviews and accepted/skipped dates against the original archive before retiring your previous setup. Automatic goal recreation is not available. Fathom and Umami use their own export formats; DataFast uses its API.

Something unclear? Contact support.