--- title: "Import your analytics — Tracwell docs" description: "Import analytics history from Seline, Umami, Databuddy or Vercel into Tracwell. Review coverage, dates and supported data before importing." canonical: "https://tracwell.app/docs/imports" markdown: "https://tracwell.app/docs/imports.md" --- # Import your analytics Keep your traffic history when you switch to Tracwell. Import from Seline, Umami, Databuddy or Vercel, then continue collecting new visits with Tracwell. ## Before you start Open Settings → Import data and choose a provider. Select the destination website explicitly: you must be its owner or an administrator. Check that the source website matches the destination before confirming. Install Tracwell to collect new visits. An import adds historical reports; it does not connect the old provider for ongoing synchronization. ## Import from Seline Seline imports are in limited rollout and appear only in enabled workspaces. 1. 1 Copy your project token In Seline, open your project’s Settings → Integrations and copy its API token. 2. 2 Choose your dates In Tracwell, choose Seline and the destination website. Paste the token and select completed UTC days, up to 366 days and 10,000 pageviews per import. 3. 3 Review and save Preview the history, check the source project and dates, then save. Tracwell reads the source again when saving and waits for storage confirmation. Daily pageview and visitor summaries include available page, referrer, country, device, browser and OS breakdowns. Custom events, charges and raw visitor histories are not imported. Tracwell does not save your token. Preview and save each make requests to Seline that count toward its API allowance. Larger ranges must be split into non-overlapping periods. If a single day exceeds 10,000 pageviews, it cannot currently be imported. Incomplete results or full-range totals that disagree with Seline’s summary are rejected. [Seline API documentation](https://seline.com/api/overview) ## Import from Umami 1. 1 Export your data In Umami Cloud, open your profile menu → Settings → Data → Export. Choose your website and date range, then request the export. Download it from the link Umami emails you. 2. 2 Prepare the ZIP Place website\_event.csv directly inside the ZIP, without a containing folder. You can keep event\_data.csv and session\_data.csv alongside it. If the download is gzip, extract it and ZIP the CSV files first. 3. 3 Choose the source website In Tracwell, choose Umami and your destination website. Enter the Umami website ID from that website’s settings, then select your ZIP. 4. 4 Upload and review Let Tracwell prepare the file, then review the detected dates, times and notices before confirming the import. Only supported website\_event.csv exports are accepted. Dashboard summary exports are not supported. Self-hosted compatibility depends on the CSV columns, not just the Umami version. Pageview history and available page, referrer, country, browser, device and OS breakdowns are imported. Custom events and properties from event\_data.csv or session\_data.csv are not imported. Audience details missing from website\_event.csv cannot be recovered by adding those files. [Umami export documentation](https://docs.umami.is/docs/cloud/export-data) ## Import from Databuddy 1. 1 Export as CSV Open your website in Databuddy → Settings → Export. Choose CSV, enable a custom date range, select the period you need, and click Export Data. 2. 2 Keep the original ZIP Upload the downloaded ZIP unchanged. It must contain events.csv and metadata.json directly inside it. Tracwell detects the source website ID from the export. 3. 3 Review the history Choose Databuddy and the destination website in Tracwell. Upload the ZIP, review the detected time range and notices, then confirm the import. Only screen\_view events contribute to imported pageviews. Available page, referrer, country, device, browser and OS breakdowns are included. Custom events, page\_exit events, properties, errors and web vitals are not imported. Exports containing one million or more events are rejected because Databuddy may have capped the file. Export a smaller date range from Databuddy instead. Always review the actual timestamps: selecting an end date does not prove the file contains that entire day. [Databuddy dashboard documentation](https://www.databuddy.cc/docs/dashboard) ## Import from Vercel 1. 1 Create an access token In Vercel, switch to your Personal Account → Settings → Tokens → Create. Give the token a name such as Tracwell import, select the account or team that owns your project, then create and copy the token. 2. 2 Find your project IDs Find the project ID and, if applicable, team ID in Vercel settings. 3. 3 Choose the period Select Vercel and the destination website in Tracwell. Enter the token, project ID and optional team ID. Choose completed UTC days, up to 366 days per import. 4. 4 Preview and import Preview the available history, check the destination and dates, then save. Wait for storage confirmation before treating the import as complete. [Create a token in Vercel](https://vercel.com/account/tokens) , then paste it into the Vercel access token field in Tracwell. Vercel uses an API connection rather than a ZIP upload. Tracwell does not save your access token. Only history available through Vercel’s API can be retrieved; importing cannot recover data that Vercel no longer provides. Imports contain production traffic summaries. Vercel’s returned top-value limits apply to breakdowns; custom events are not imported. ## Review dates and times For Umami and Databuddy ZIPs, Tracwell detects the earliest and latest event timestamps after upload. Review the displayed timezone and confirm the start and cutoff. You can adjust the interval before importing; preparation must finish again after a change. The start is included and the cutoff is excluded. To switch tracking partway through a day, choose the time Tracwell started collecting as the cutoff. Native Tracwell events at or after that cutoff remain visible. Events before the imported start remain visible too. Detected timestamps show what is present in the file, not proof that the export is complete. Gaps may be missing data or periods without traffic. Seline, Vercel and older daily imports cover whole UTC days; new ZIP imports retain hourly summaries. ## Existing history and Undo Non-overlapping imports can remain active together. An import that overlaps the UTC dates of an active import is blocked, even when their times within the same day differ. Existing imported history stays available. Within an active import’s coverage, dashboard reports use imported summaries in place of native traffic to avoid double counting. Native events are not deleted. Outside that coverage, native traffic remains visible. Undo all imports removes all active imported history from the dashboard and restores native reporting. Saved imports remain in history; use Include in dashboard to reactivate a compatible saved import that does not overlap active coverage. ## What imported history includes Imports populate traffic trends and supported breakdowns in Overview and Acquisition. They do not recreate raw events, session journeys, funnels, custom-event reports or revenue attribution. Visitor totals are estimates across sources and time buckets, not a shared identity history. ZIP processing continues in the background after upload. You can close the dialog and return to Import history to review progress. Your existing dashboard stays available while an import is prepared. If a ZIP is rejected, check its file structure and the reported error. The upload limit is 256 MiB compressed, with up to 1 GiB per expanded file and 2 GiB expanded in total.