Forty case studies in the CMS, drafted, nobody typing

Forty items created, drafted first

The create tool takes a draft flag, so the forty land unpublished with their reference fields resolved and a person finishes them in the editor. For a migration, or a backlog of content that exists everywhere except the CMS, that is the whole job.

Corrections and removals happen in the Webflow editor, so point a companion at getting content in and keep the ongoing copy edits where the designer and the preview are.

How do the fields actually get filled?

By slug, not by the friendly label in the designer, so the companion lists the collection first and reads what the fields are really called. Name and slug are required. Reference fields take an id, so it lists the referenced collection to find the right one.

The interesting half is upstream. The case studies live in documents, the metrics in a spreadsheet, the client logo on their own site. A companion reads all of that in the accounts and tabs you are already signed into, assembles each item, and stages the set. That gathering is the same work as any marketing content operation.

• List sites, then collections, then items; ids flow forward from each step.
• Read existing items in full, including draft and archived state, to avoid duplicates.
• Create items as drafts when a human should finish them in the editor.
• Field values keyed by slug, with reference fields resolved to real ids first.

Why does publishing wait for you?

Because publishing pushes every pending draft change live, not only the items a companion created, but whatever else sits unpublished in the project, including a designer’s half-finished section. The publish tool is marked destructive in its own definition and confirms before it runs.

So the shape is two steps with a person between them: create the items as drafts, review them in Webflow, publish when the site is in a state you would show a customer.

Turn documents and sheets into CMS items

Forty documents to read, a spreadsheet to reconcile, image credits to check on somebody else’s site, then rows in a collection, drafted, with the reference fields resolved. That is a data extraction job with a CMS at the end of it.

This surface is sites, collections, items and publishing: the content database rather than the site build. Compare it with the Sanity tools when your CMS is headless and the audit is a query.

Frequently asked questions