Mass Edit

Some jobs don't fit the page editor. Rewriting every meta description in the space, filling in the German title for 200 products, finding all the entries where the teaser is still empty — opening each page one by one is the wrong tool.

Mass Edit turns that work into a spreadsheet: you pick the fields you care about, and every entry that has them shows up as rows in an editable grid, with one column per language.

You'll find it in the space sidebar under Mass Edit. Anyone who can view content can open it; editing and saving needs content editing rights, and Save & publish needs publishing rights.

The three-step setup

The grid stays empty until you tell it what to load.

1. Pick fields. The field selector lists every text-carrying field in your space, gathered from all your content types — including fields that only ever appear inside nested blocks. Fields are listed by their label with the technical key next to it, so title in a hero block and title on a page are the same column.
2. Pick languages. All space languages are selected by default; drop the ones you don't need to keep the grid narrow. The source language is always shown first and marked Source.
3. Narrow it down with the search filter (optional, but the difference between 40 rows and 4,000).

Which fields show up

Fields appear when their type can hold translatable content: text, textarea, markdown, rich text, number, date, meta, and table. Assets, references, options, booleans and link fields aren't editable here — they're structure, not prose.

Two flavours of field end up in the list:

• Translatable fields — editable in every language column.
• Fields marked source only — supported types that your content type does not mark as translatable. They're useful to see and edit, but only in the source column; the language columns show a lock, because a non-translatable field always delivers the canonical value anyway (Internationalization).

Some fields expand into several rows, because that's how they're actually translated:

Field type	What you get in the grid
Meta	One row each for title, description, OG title, OG description
Table	One row per text cell
Rich text	One row holding the formatted text as HTML — tags stay, so keep them intact

Filtering

The filter bar works like everywhere else in the app — pick a field, an operator, a value:

• Name, slug, full slug — contains, starts with, ends with, equals
• Content type and status (published / draft)
• Created at / updated at — before, after
• The value of any selected field — contains, starts with, ends with, equals, plus is empty and is not empty

That last one is the workhorse: select meta, filter Meta title is empty, and you have your entire to-do list on one screen.

Editing in the grid

Rows are grouped per entry: the first row of each entry carries its name and full slug, and every row names the field it belongs to (nested block fields show their path, so you can tell the hero headline from the teaser headline).

• Edited cells get an amber outline so you can see your own trail across a long grid.
• Escape reverts the cell you're in back to its stored value.
• ↑ / ↓ jump to the same column in the row above or below once the cursor reaches the start or end of the text — you can tab down a whole column without touching the mouse.
• Cells grow with their content and can be dragged taller.

Your edits survive paging. Move to page 4, edit there, come back — everything is still pending, and one save writes all of it. The counter in the Save button always shows the true number of open changes, not just the ones on screen. Discard throws them all away.

Narrowing the selection while you have unsaved edits is the one case the app asks about: dropping a field or a language hides its cells but does not drop the pending changes, so you're offered the choice to discard first.

Saving

• Save stores everything as drafts — nothing goes live, exactly like saving in the editor.
• Save & publish (in the button's menu) saves and publishes the affected entries in one go.

Three details worth knowing:

• Missing translations are created for you. Typing into a language column of an entry that has no version in that language yet creates one.
• Clearing a cell means clearing the value. An emptied cell is saved as empty, not skipped — that's how you undo a wrong translation from here.
• Large saves are sent in batches. Each entry and language writes its own version, so the grid splits a big save into chunks and shows saved / total as it goes. Entries that fail keep their amber outline and stay editable; everything that landed is cleared. Nothing is lost to a timeout halfway through.

Every write goes through the normal content pipeline: versions are created, history and audit logs record who changed what, and validation applies as usual. If some entries fail, the result tells you how many saved and how many errored.

Translating with AI

With an AI configuration set up for the space (Settings → AI), Translate with AI fills the language columns for you. Use the button's menu to pick a different configuration than the default.

What it does:

• Works on the whole current selection, not just the visible page — fields, languages, and filters define the job.
• Translates cells that are empty, or whose source you just edited in this session (the existing translation is stale). Cells you've already translated are left alone.
• Skips source only fields.
• Streams results into the grid as they arrive, in batches, with a live applied / total counter.

Translations land as pending edits — nothing is written until you press Save, so you get to review and fix first. If the selection is too large for one run, you'll be told to narrow the filters and repeat.

Export and import

The header buttons hand the same selection to the translation exchange formats — XLIFF, CSV, Excel, JSON, or YAML:

• Export respects your fields, languages, and filters: what you exported is what you were looking at, including rows whose cells are still empty (that's usually the point — the translator needs to see the gaps).
• Import applies a returned file back, as drafts for review or published immediately, optionally creating language versions that don't exist yet. A summary shows exactly what changed.

A file exported from Mass Edit round-trips: it has one column per language including the source, so editing the source column and importing works, and a cell you blank out is imported as a deliberate clear. That's the grid's own shape, and it's why the export you take from here differs slightly from the one in the content editor, which ships a read-only source column for translators.

One exception: XLIFF models the source as read-only by design, so an edited source will not come back through it. Use CSV, Excel, JSON or YAML when you want to edit source text outside the app.

Otherwise this is the same import/export used from the content editor, so an agency workflow started here can be finished there and vice versa.

When to use what

Situation	Best tool
Writing or restructuring one page	Content editor
Translating one page carefully, in context	Editor in localization mode
The same field across many pages — SEO sweeps, missing translations	Mass Edit
Handing work to an external agency	Export from Mass Edit, import when it comes back
Moving or restructuring many pages	Canvas

Good to know

• Only canonical entries are listed — one row group per piece of content, with its languages as columns. You never edit a translation "as its own page" here.
• Entries whose content type doesn't have any of the selected fields simply aren't listed.
• Rows are ordered stably, so paging never skips or repeats an entry.
• Everything you do here is scriptable too — the grid runs on the mass-edit/fields, mass-edit/rows and contents/export endpoints of the Management API. A PATCH mass-edit/rows call takes at most 100 entries; send more in batches.