Download your LinkedIn saved posts as a CSV

LinkedIn does have an export. It is just not the one people mean. Request your archive under Settings, Data Privacy, Get a copy of your data and what arrives for saved items is a CSV of post URLs and save dates. No post text, no author, no media. A list of links, some of which are already dead.

This is how to get the other kind: one row per saved post, twelve columns, everything the post actually contained.

The twelve columns

The header is fixed, in this order, so anything you build on top of it keeps working:

post_id,text,link,posted_at,labels,
image_urls,video_url,document_url,
author_name,author_id,author_username,author_link

Three details that matter once you open it in something:

• One post is always one row. Post text has its line breaks collapsed to single spaces on the way out, so a multi-paragraph post cannot split itself across rows and break your parser.
• Multi-value fields are semicolon-joined inside one quoted cell. A carousel puts every image URL in image_urls separated by ; , and your labels arrive the same way in labels.
• It is RFC-4180 quoted. Cells containing a comma, quote or newline are wrapped, and embedded quotes are doubled. Excel, Numbers, Google Sheets and every CSV library handle it without options.

Exporting it

Open Export to CSV in LinkedMash and press the export button. There is no job queue and no email to wait for: the file downloads in the same request, and the page tells you as much — Downloads instantly and is saved to your export history below.

Every export is recorded under Recent exports on that page. For CSV, the Download action regenerates the request from its saved post IDs and the current stored records. It does not retrieve immutable original bytes. If a stored post or label changed, the regenerated file can differ.

On the free plan the button reads Export CSV sample and gives you 20 posts, which is enough to check the shape before paying for anything. A paid plan exports the whole library with media, labels and author details.

Exporting a slice instead of everything

A full-library CSV is the wrong artefact for most jobs. If you want the posts about one topic, search or filter your library first and export the selection — the export follows what you have on screen rather than dumping everything.

This is the fastest way to hand a client a research list without also handing them everything else you have ever saved.

What to do with it once you have it

1. Count who you actually learn from. Pivot on author_name. The answer is usually not the people you would have named.
2. Find the topic you keep circling. Filter text for a term across a year of saves and you find the thing you have been researching without noticing.
3. Keep a real backup. A CSV survives the author deleting the post, going private, or LinkedIn changing its mind. A link does not.

When CSV is the wrong choice

• You want it to stay current: a CSV is a snapshot. Connect Google Sheets and it keeps writing on its own. The Sheets row shape differs from the fixed twelve-column CSV schema.
• You are writing code against it: take JSON instead. Nested media and author objects beat string-splitting a cell.
• You want to read it: a wall of post text in a spreadsheet cell is unreadable. Use PDF or Notion.

CSV wins on exactly one axis, and it is a good one: everybody can already open it.

The best format is the one the person you are sending it to does not have to ask about.