Research a prospect with Perplexity, then connect on LinkedIn

The short answer

Perplexity is built for research with sources. Connected to Finchreach, it can take that research one step further: find the right person at the company on LinkedIn, check whether you share connections, and draft a connection request with a note that mentions something real. The request waits in your approval queue until you send it.

Finchreach is a remote MCP server at https://mcp.finchreach.com/mcp. MCP is a standard plug that lets AI apps use outside tools. Perplexity supports custom remote MCP connectors with an API key option, so you can add Finchreach from its connectors screen.

Why research before sending a LinkedIn connection request?

A blank or generic connection request is easy to ignore. A note that shows you understand the person’s company and timing gets read. The trouble is that doing that research by hand takes fifteen minutes per prospect, so most people skip it.

Perplexity handles the outside research: recent funding, product launches, hiring, public talks. Finchreach handles the LinkedIn side: linkedin_search with the people vertical finds the person and can filter by network distance (1st, 2nd or 3rd+), the companies vertical finds the company page, and linkedin_send_invitation queues a request with a note of up to 300 characters.

What does a prospect research session look like?

Start broad with the company, the way you would in Perplexity anyway:

Ask your assistant“Research Northwind Freight. What do they sell, who are their customers, and what changed in the last six months (funding, launches, leadership, hiring)? Cite sources.”

Then bring in LinkedIn:

Ask your assistant“Search LinkedIn people for the head of operations or VP of logistics at Northwind Freight. Prefer 2nd-degree connections. Show me up to five candidates with headline and network distance.”

Pick the person, then ask for the note. Be explicit about tone and length:

Ask your assistant“Draft a connection note to the second person, under 300 characters. Mention their new Dallas hub from the research, say I work on route planning software, and don’t pitch. Show me the note, then queue the invitation.”

The invitation shows up at /approvals. If you’ve changed your mind, reject it and ask for a new draft; if you approve it, it goes out from your account.

How do you connect Perplexity to LinkedIn?

1. Create an account at finchreach.com/sign-up, connect your LinkedIn account under Accounts, and create an API key under API keys.
2. In Perplexity, open Plugins in the sidebar (or Customize → Connectors) → Add custom → Add MCP connector → Remote.
3. Enter a name, the URL https://mcp.finchreach.com/mcp, choose Authentication API Key and paste your key, pick Transport Streamable HTTP, tick the acknowledgement and click Add.
4. Enable the connector, then ask Perplexity to list your LinkedIn accounts to confirm it works.
5. The Perplexity guide has screenshots and troubleshooting; the docs cover API keys and approvals.

As of October 2026, Perplexity’s help center doesn’t say which header the API key is sent in. Finchreach accepts both Authorization: Bearer <key> and X-API-Key: <key>, which covers the common cases; if the connector fails, try pasting the key with and without the Bearer prefix. On the Mac app you can also add Finchreach as a local connector that runs npx -y mcp-remote https://mcp.finchreach.com/mcp --header "Authorization: Bearer <key>". Third-party sources say custom connectors need Pro, Max or Enterprise; Enterprise admins manage them in Enterprise settings.

What makes a connection note worth accepting?

• One specific, recent fact about them or their company, taken from the research.
• One line on who you are, in plain words.
• No ask in the first message. Save the conversation for after they accept.
• Short. LinkedIn caps notes at 300 characters, and shorter often reads better.
• Your voice. Ask Perplexity to match a note you already wrote and liked.

Because Perplexity shows its sources, check the fact you are about to quote. A wrong detail in a first message is worse than no detail.

What happens after the prospect accepts?

Accepted invitations show up in your notifications, which the assistant can read with linkedin_get_notifications. That is the moment for a first real message, and the research you already did makes it easy to write one that continues the thread instead of starting a pitch.

Ask your assistant“Check my notifications for accepted invitations from this week. For each person we researched, draft a short first message that follows up on the note I sent, with one question about their current priorities. Show me the drafts before queueing them.”

Keep the research notes from each session in the chat or a document, so a follow-up a week later doesn’t need the work done twice.

How many prospects should you research per day?

This workflow is meant for a short, high-quality list, not volume. Research a handful of accounts properly, send a few personal requests, and approve them yourself. LinkedIn limits how many invitations a personal account can send, and a burst of requests that get ignored hurts more than it helps. Finchreach doesn’t run campaigns or sequences; it executes the actions you approve.

Is automated LinkedIn outreach allowed?

Automating a personal LinkedIn account carries risk under LinkedIn’s User Agreement. Finchreach reduces it with human approval of every write and a human pace, but no tool can make it zero.

If you need many seats or many sending accounts, dedicated outreach platforms exist for that; read our comparisons for an honest view of where Finchreach fits.

Frequently asked questions

Can Perplexity message people who are already connections? Yes. linkedin_send_dm writes to existing conversations or new recipients, and every message waits for approval.

Can Perplexity read my LinkedIn inbox? It can, through the read tools, when a chat calls for it. Reads don’t need approval; anything that writes does.

Can it export a lead list? Finchreach returns search results to the chat; there is no CRM export. Copy what you need, or connect a workflow tool such as n8n.

Where do I start? Sign up, then read the Perplexity integration page.