How to export and download your LinkedIn messages

To download your own LinkedIn messages, request the Messages category through LinkedIn’s desktop account-data settings. Keep the downloaded records as your original copy, then inspect a working copy before sorting it or moving information elsewhere.

That second step matters if you want to reconnect with someone or hand a conversation to a colleague. A file of messages can contain several records involving the same person, messages you sent and conversations with several participants. Counting rows or copying the first profile-link column into a CRM can produce the wrong list.

Request your message data from LinkedIn

LinkedIn’s data-download instructions give this desktop route:

1. Open Me → Settings & Privacy → Data privacy.
2. Under How LinkedIn uses your data, open Download your data.
3. Select Messages among the available data categories and choose Request archive.
4. Follow LinkedIn’s download email and save the files on your own computer.

Check the options shown in your account. LinkedIn says available categories depend on your account activity, and this download feature is unavailable on mobile.

The current Help page puts Messages in its “Available within 48 hours” table, even though its general introduction mentions shorter times. Treat that as LinkedIn’s guidance, not a guaranteed arrival time. It also says the download remains available for 72 hours. That is a download-link window; it does not tell you how many years of messages a file contains.

Check who each message record describes

LinkedIn describes the Messages category as information about sent, received and archived messages, including dates, subject, content and a member’s public profile URL. It does not publish an exact message-file schema on that Help page.

If your download contains messages.csv, open a copy and read the actual headers. A CSV is a table, so it may need filtering to show one conversation. Keep a copy in its original order. Before sorting by time, confirm how your spreadsheet reads the dates and whether a time zone is shown.

Start with a short conversation you recognize. Find a message you sent and a reply you received, then compare the identity fields against the message text. This tells you what the columns mean in your file before you process a larger set.

The free LinkedIn message reader opens one UTF-8 CSV in a separate local reader tab. Confirm its columns, search for literal text across the fields and inspect records in their source order. It can show records sharing a mapped conversation ID; it does not establish a complete thread or turn message rows into prospects.

• A column named SENDER PROFILE URL: on a message you sent, this can point to your own account. It is not automatically the person you want to save.
• A recipient field containing several people: treat it as a group conversation. Establish which participant the planned note or follow-up concerns.
• Several rows for the same exchange: preserve them when reading the conversation. They do not represent several new prospects.
• A name without a usable profile link: leave the identity unresolved until you verify it. Do not construct a LinkedIn URL from the name.

Use those checks where they fit the fields in your file. If it lacks the identifiers you need, stop the person-list conversion there. Keep the original message records while you resolve the participants.

Keep missing, archived and deleted conversations separate

A conversation missing from your main inbox may be in Archived. LinkedIn’s archive instructions and restore instructions describe how to find and restore it. Moving a conversation into that folder and requesting a data download are separate operations.

Deletion has a different consequence. LinkedIn says a deleted conversation cannot be restored or accessed. Do not request an archive on the assumption that it will recover a conversation you deleted.

For an expected record missing from a finished download, check that you requested Messages and opened the intended archive. Compare against an example you can still access in LinkedIn. Its data-download Help page points to a Data Access Request Form for data you cannot find.

The description of the Messages category does not establish that every attachment, image, read receipt or delivery state will be represented. Check the particular material you need before treating the download as sufficient for your purpose. If you need to preserve something that is absent, retain the accessible original separately rather than describing the text export as a complete copy.

Move useful context into a separate person record

For a small handoff, a verified profile link and a factual note may be enough. Preserve the message date, what the person actually asked for and any next step that was agreed. Distinguish that historical exchange from what you know now. A reply from two years ago does not establish that a follow-up is still relevant.

In an invented example, an old message says, “Send the workshop outline in November.” A useful note records that request and its date. It should not silently become “interested prospect” or permission to start an automated campaign.

Keep the downloaded archive separately from the file you plan to share or import. Share only the context needed for the handoff with someone entitled to receive it. There is no reason to upload an entire private inbox to prepare a small prospect list.

In OutreachGenie, the Import archive control in the content area imports LinkedIn posts from Shares.csv. It does not import message transcripts. The prospect CSV/XLSX importer accepts person fields and requires a LinkedIn profile URL; importing that separate file does not recreate conversations, reply states or campaign history.

If you have verified the people you want to retain, use the prospect CSV import workflow for their person records and add reviewed context manually as notes. The sales handoff guide explains what a colleague needs to understand before continuing a conversation. If you need first-degree connection records instead of message content, request the separate connections export.

Keep unresolved identities out of the import. The useful endpoint is a person you can identify and a note you can explain from the original exchange.