Importing from CSV - Trackur Docs

Trackur can import jobs from a CSV file — either merging them into your current list or replacing everything you have. The format matches Trackur's own export, so moving data between accounts or restoring a backup is straightforward.

Opening the import dialog

On desktop, click the up-arrow icon in the toolbar (tooltip Import CSV). On mobile, open the toolbar's ... overflow menu and tap Import CSV — see Trackur on mobile for the full mobile layout.

A dialog titled Import Jobs from CSV opens with a dashed upload box labeled Click to select a CSV file. Click the box to open your file picker (filtered to .csv files); once you choose a file, the box shows its filename.

Click to select — no drag-and-drop

The dashed box looks like a drop zone, but dropping a file onto it does nothing. Click it and pick the file instead.

After the file is read, the dialog reports "Found 3 jobs in the file." (with your actual count) and shows two buttons: Merge with existing and Replace all. Cancel closes the dialog without importing. If the file can't be read, a red message says "Failed to parse CSV file. Please check the format."

The expected format

The file needs a header row followed by one row per job. The recognized columns are the same ones Exporting to CSV produces:

id,company,role,stage,dateApplied,todos,notes,postingUrl
,Acme Corp,Frontend Developer,Applied,2026-07-01,"[ ] Follow up with recruiter (2026-07-15); [x] Send application",Referred by Sam,https://example.com/jobs/123
,Globex,Product Manager,Interviewing,2026-06-24,,Second round scheduled,

All dates — dateApplied and any due dates inside todos — must be in YYYY-MM-DD format. An invalid date can make the whole import fail with an error toast.
The todos cell packs a job's next steps into one field: [ ] for open, [x] for done, an optional due date in parentheses, items separated by semicolons. Because semicolons separate items, a step whose own text contains a semicolon gets split into multiple steps on import.
Rows missing both company and role are silently skipped and don't count toward the "Found" number.
Older Trackur exports with nextAction and nextActionDate columns still import — they become a single open next step on the job.

Merge with existing

Merge with existing adds the file's jobs alongside your current ones. Rows whose id matches an existing job are skipped — an import never updates existing jobs. Everything else is added as a new job with a fresh ID. The toast reports exactly what happened: "Imported 2 new jobs, 3 duplicates skipped (12 total)" — or, when every row was a duplicate, "No new jobs imported — all 5 rows already exist (12 total)".

Duplicates are matched by id only

Duplicate detection uses the id column only, so a row with a blank or changed id is always added as a new job, even if its company and role match a job you already have.

Replace all

Replace all deletes every job in your account, then imports the file's rows with fresh IDs. A toast confirms "Replaced with 3 jobs".

Replace all has no undo

There is no confirmation step beyond the button itself and no way to undo. It also wipes all stage history — every imported job starts over with a fresh "Created in" entry. And it works even on a file with zero jobs: replacing with a header-only file deletes everything and imports nothing. Export a backup before you use it.

What to check after importing

Stage names are case-sensitive and unvalidated. A stage like applied (lowercase) imports fine, but the job won't appear in any column on the board because it doesn't match a real stage name. The job still exists — you can find it in Table view and fix its stage there.
Resume links are stripped. Imported jobs always arrive with no resume attached — see Attaching resumes to jobs to relink them.
Every imported job arrives active (not archived) and with empty stage history, regardless of what it was before export.