LinkedIn Outreach at Scale — 50 Personalized Connections Per Week

The Problem

I was doing LinkedIn outreach the way everyone does: manually. Browse a profile. Read their bio. Craft a connection request that doesn't sound like a bot. Send. Repeat.

Ten connections a day if I was disciplined. Usually more like five. Each one taking 5-10 minutes of reading, writing, sending. Two hours a day for a channel that pays off weeks later when someone actually accepts and replies.

After hosting Mr.Chief Paris meetups, I had a specific problem: 93 registrations on Luma, but half of them I'd never connected with on LinkedIn. These were warm leads — people who showed up to my event — and I was losing them because I couldn't manually send 50 personalized connection requests.

The math was simple. Manual: 10/day × 5 days = 50/week = 10 hours. Automated: 50/week = 0 hours. Same output, zero input.

The Solution

An Mr.Chief pipeline: export Luma attendees → enrich with LinkedIn profiles → generate personalized connection requests → queue a follow-up sequence → track results. The agent handles research, writing, and scheduling. I handle the actual conversations once someone replies.

The Process

The pipeline has four stages:

Stage 1: Lead extraction from Luma

# Export event attendees
# Agent fetches the attendee list from the Luma event page
# Extracts: name, email, company, title, LinkedIn URL (if provided)

Stage 2: LinkedIn enrichment

task: |
  For each attendee without a LinkedIn URL:
  1. Search LinkedIn for "[name] [company]"
  2. Verify the match (title, location, profile photo context)
  3. Store the enriched profile: name, title, company, headline, recent posts

  For each attendee WITH a LinkedIn URL:
  1. Pull their profile data: headline, about section, recent activity
  2. Note any shared connections, mutual interests, or conversation hooks

Stage 3: Personalized message generation

task: |
  For each enriched lead, generate a 3-step sequence:

  STEP 1 — Connection Request (max 300 chars):
  Template: "Hey [first_name] — saw you at Mr.Chief Paris [or: registered for].
  [One specific detail about their profile/work]. Would love to connect."

  STEP 2 — First Message (3 days after acceptance):
  "Thanks for connecting, [first_name]. [Reference to their work/company].
  We're building [brief pitch relevant to their role]. Curious if
  [specific question about their experience with the problem we solve]?"

  STEP 3 — Follow-up (7 days after Step 2 if no reply):
  "[Light touch referencing something recent — their post, company news,
  or industry development]. Either way, glad to have you in the network."

  Personalization rules:
  - NEVER use generic openers ("I'd love to pick your brain")
  - ALWAYS reference something specific: their headline, a post, their company's product
  - Match formality to their profile (casual for startup founders, professional for corporate)
  - If they posted about a topic recently, open with that

Stage 4: Execution and tracking

task: |
  Queue the outreach sequence:
  - Send connection requests in batches of 10/day (LinkedIn safety limits)
  - Track acceptance: log date, status
  - On acceptance: schedule Step 2 message for +3 days
  - On no reply to Step 2: schedule Step 3 for +7 days
  - On reply at any stage: flag for manual conversation

  Weekly report:
  - Connections sent: X
  - Accepted: Y (Z%)
  - Conversations started: W
  - Meetings booked: V

Example of a generated connection request:

"Hey Sarah — saw you registered for Mr.Chief Paris. Your work on
agent orchestration at Mistral looks fascinating, especially the
multi-model routing stuff. Would love to connect."

Versus what I'd get from a generic tool:

"Hi Sarah, I noticed we share similar interests. I'd love to connect
and explore synergies. Let's chat!"

Night and day.

The Results

The acceptance rate is higher with automation because the personalization is actually better. When I was doing it manually at volume, messages got lazy by connection #30. The agent gives the same quality to #50 as #1.

Try It Yourself

1. Start with a warm list — event attendees, webinar registrants, people who engaged with your content
2. Enrich profiles before writing messages (you can't personalize without context)
3. Set strict daily limits (10-15 connection requests/day to stay within LinkedIn's guidelines)
4. Write the personalization rules yourself — the agent executes your playbook, not a generic one
5. Always flag replies for manual handling — automation gets the door open, you walk through it

The line between "automated outreach" and "spam" is personalization quality. If the recipient can tell it's automated, you failed. If they think you actually read their profile — because the agent did — you win.

50 connections a week, each one personalized to their actual profile, their actual work, their actual interests. The agent reads every profile. I just handle the conversations that matter.