Quick start from your past posts (LinkedIn export)

If you've posted before, Sayweek doesn't need weeks to learn: your old posts show right away how long you write, how you open and what your voice sounds like. The posts become the brand's voice samples, and with learning switched on we also write style rules from them. See also: How does Sayweek learn your voice?

Uploading your LinkedIn export

1. 1On LinkedIn: Me → Settings & Privacy → Data privacy → Get a copy of your data. Pick your posts (Shares / Posts), or ask for the full archive.
2. 2LinkedIn e-mails you a download link – the selected files usually arrive within minutes, the full archive within 24 hours.
3. 3In Sayweek, open Settings → Brands & channels and scroll to the Quick start from your past posts card.
4. 4Pick the brand, drop the ZIP (no need to unzip it) or Shares.csv on the LinkedIn export tab, check the preview, then click the import button.

Pasting posts

On the Paste tab you can paste several posts at once. Separate them with a line of --- or two empty lines. Even 5–20 posts say a lot.

What we do with the posts

• We drop duplicates – posts uploaded again or nearly identical ones too – and shares shorter than 40 characters.
• One import keeps at most 300 posts (the newest), 500 per brand in all.
• We compute statistics: typical length, emojis, hashtags, most common opening, posting rhythm.
• If learning is switched on for the brand, Claude reads a few typical posts and writes at most 8 style rules from them.
• New drafts take examples from these posts as well – different ones for each draft.

What we learned – and what you can delete

Under What we learned from your posts you see the statistics, the rules and the imported posts. You can delete any rule or post one by one, and Undo this import removes the whole import together with the rules learned from it. A deleted rule is never learned again.