Most recruiters treat LinkedIn like a search bar. Type a title, scroll, forget, repeat next week. Meanwhile, the good stuff was already sitting there the whole time.

Every person on your team has spent years building a network on LinkedIn without even trying. Engineers, founders, hiring managers, people who already said yes to connecting with them. That trust is free, and still nobody’s using it.

We’ve spent years telling clients to network strategically, and it still took us this long to use our own.

The setup: one shared database, built from your team’s LinkedIn exports

Credit where it’s due: this workflow started with something Dani Unsworth, Chief of Staff at Semaloop, shared publicly. We tested it with a client; it worked, so we had to share our version.

Step 1: Export your LinkedIn archive
Settings & Privacy → Data Privacy → Get a copy of your data → Download larger data archive. It claims 24 hours, but it usually shows up in one or two.

Step 2: Ask the whole team to do the same
One person’s export is a nice list. Four or ten people’s exports is an actual weapon.

Step 3: Dump everything into Claude Cowork
Upload every file, ask it to merge them into one clean sheet. It handles the formatting and the duplicates so you don’t have to open a single spreadsheet manually.

Step 4: Tell it what you’re matching against
Recruiting? Paste in your open roles. Sales? Paste in your target list. Claude doesn’t need a perfect brief, just a link.

Step 5: Let it tag every contact
Role, seniority, department, and a match rating: strong match, match, maybe, not really. Build one prompt template and reuse it every time so results stay consistent.

Step 6: A human still checks the work
This part’s on you. Review what Claude flagged, run it past whoever owns that connection, then decide who gets a message.

Step 7: Every new hire adds to the pile
Same export, same five minutes, every time someone joins. The database keeps growing on its own.

Why this works

A search gives you results that vanish the second you close the tab. An export gives you something permanent, something AI can scan for connections nobody remembers making.

That’s the real shift. Most of these connections aren’t lost because they weren’t valuable; they’re lost because nobody on your team was ever going to manually scroll through four thousand names looking for the right one. AI does that in minutes.

If your company runs a referral bonus program, each match becomes a paid opportunity. If someone on your team has been quietly connected to a great candidate for years, they get to be the one who reintroduces them, and they get paid for it.

The part nobody likes to admit

The tagging won’t be perfect out of the box. Somebody with a vague job title is going to get miscategorized, and you’ll have to fix it by hand. So it’s mostly automated. The last 20 percent is still just you, a coffee, and some questionable job titles.