Flow: CSV import and export
You can load objects into Flow from a CSV table and export them back — for example, to move deals from another CRM or to hand a task list over to Excel. The "Import" and "Export" items are in the "Import and export" menu (an icon on the toolbar above the type's table, on the "Table" tab).
Import
Import creates objects of one type — the type whose table is open. You need permission to create objects of this type, and to update found objects you also need permission to change them.
Which files work
CSV up to 10 MB and up to 100,000 rows, no more than 200 columns, with cells up to 100 KB.
UTF-8 encoding (including with a BOM, as Excel saves it), UTF-16 with a BOM, or Windows-1251 — detected automatically.
The delimiter is a comma, a semicolon or a tab; Flow picks whichever appears most often in the first row.
Split a large file into parts.
Steps
Open the "Import and export" menu → "Import" and click "Choose file". Flow parses the file right away and shows the number of data rows.
Check or uncheck "First row is headers".
Map columns to fields. The table shows "File column", a "Sample" from the first rows, and "Field". Columns whose header matches a field's label or key are mapped automatically; a column named "Name" or "Title" (or the Russian equivalents) is mapped to the object's title. Leave unneeded columns as "— don't import —". The same field can't be chosen for two columns.
Choose what to do "If the object already exists" (see below).
Click "Import N". The import runs in the background: the window shows progress, and you can close it — the objects will appear in the table.
Relation fields aren't filled in by import.
If the object already exists
Option | What happens |
|---|---|
Create all rows as new objects | Each row becomes a new object |
Skip rows whose object is found | A row with an already existing key value is skipped |
Update the found object with filled cells | The found object is updated; if none is found, a new one is created |
For the last two options, choose "Find object by field": the title or a mapped field of the "Text", "Number", "Integer", "Select", "Member" or "Date" type. Only non-archived objects you can see are searched. If several objects are found for a value, the row gets an error. Repeated keys within the file are taken into account: the second row finds the object created by the first.
On large types, the key field should be indexed — see Flow: object types and fields. Without an index, a search that is too expensive can stop the import.
How cells are read
Numbers: spaces are removed, and a comma is the decimal separator (
1 250,5).Dates:
DD.MM.YYYYorYYYY-MM-DD.Yes/no:
yes/no(the Russian equivalents work too).Select: the option's key or its label.
Multi-select, tags and other lists: items separated by
;or,.Checklist: one item per line;
[x] textis a checked item,[ ] textor just text is an unchecked one.Money: the amount and currency, for example
1 250,50 RUB.
On creation, an empty cell doesn't set the field (the default value is used), and on update it doesn't change it. You can't clear a field with an import — this is so that an empty column doesn't wipe out data.
Results and errors
Each row is checked the same way as when you create an object manually: required fields, allowed values, the type's rules and your permissions. A row with an error doesn't affect the others.
When the import finishes, the window shows the counters "Created", "Updated", "Unchanged", "Skipped", "With errors" and a "Rows with errors" list: the file row number (counting the header row) and the reason. Fix those rows in the table and upload them as a separate file.
Only the person who uploaded a file can see it; it is kept for 7 days.
Export
"Export" downloads a CSV file:
if a saved view is selected and you haven't changed it — the view's objects with its columns;
otherwise — the current list: your filters (including "Assignee"), sorting and table columns.
Search isn't applied to the export — narrow the list with filters. The subtask counter and "Spent" columns aren't exported.
File format:
UTF-8 with a BOM (Excel opens it without asking for an encoding), with a comma as the delimiter;
the first row has field labels, and select values are written as option labels. System columns ("Title", "Status", "Assignee", dates) are labeled in your interface language; field and option names are as they were set in the workspace;
the file includes only objects and fields you are allowed to read;
text that starts with
=,+,-or@gets an apostrophe at the beginning so the spreadsheet editor doesn't treat it as a formula.
Export is limited to 100,000 rows. If there are more objects, Flow reports "The first … rows were exported — that's the export limit" — narrow the filter and export the rest separately.
Through the API you can export a file with field keys in the header — such a file is convenient to load back with an import. See Flow API: views, export, import and background operations.
Bulk actions
In the table, check objects with the checkboxes — a "Selected: N" toolbar appears with actions: "Status", "Assignee", the deadline field (for example, "Deadline"), "Field" and "Archive", plus "Clear selection". Changes run in the background with your permissions; at the end Flow shows how many objects were processed and how many failed — with the reason (for example, permissions, transition conditions, or relations that forbid archiving). Learn more in Flow: table, board, calendar and filters.
What's next
Flow: table, board, calendar and filters — filters and views for export.
Flow API: views, export, import and background operations — import and bulk operations from your own program.
Пікірлер