Hyper-Personalized LinkedIn
Outreach with Claude Code
Find leads, score them with AI, and send value-driven messages that start real conversations. A 3-part system for non-technical people.
Find Leads
Sales Navigator + Apify scraping
Score Leads
Claude Code ranks by fit
Send Value
Personalized outreach with free deliverables
Part 1: Find & Collect Leads
Use LinkedIn Sales Navigator to build a targeted lead list, then scrape it with Apify.
1Set Up LinkedIn Sales Navigator
- You need a LinkedIn Sales Navigator subscription. It's the only reliable way to do advanced lead searches on LinkedIn.
- Create a Lead Search with your ideal client profile (ICP). For example:
- Location: United States
- Title: Founder, CEO, Owner
- Industry: Marketing
- Refine until you have a focused list of 100–500 leads that match your ICP.
2Connect with Your Leads
- Start sending connection requests to your leads from Sales Navigator.
- Pro tip: Use Claude Cowork to automate sending personalized connection requests — we have a full guide on setting that up.
- Keep connection requests under 20–30 per day to stay within LinkedIn's safe zone.
- Wait for people to accept. You'll typically get a 20–40% acceptance rate with personalized requests.
3Filter & Export Your Connections
- Once you have accepted connections, go back to Sales Navigator and create a new search.
- Filter by 1st-degree connections + your ICP criteria (location, title, industry). This gives you only the people who accepted your connection request and match your target.
- Copy the URL of this Sales Navigator search — you'll need it for the next step.
4Scrape Leads with Apify
- Go to apify.com and create a free account. You get $5/month in free credits.
- Open the LinkedIn Sales Navigator Scraper actor.
- You'll need your LinkedIn cookies from Sales Navigator. Here's how to get them:
- Open Sales Navigator in Chrome
- Press
F12to open Developer Tools - Go to the Application tab → Cookies → linkedin.com
- Copy the
li_atcookie value
- Paste your Sales Navigator search URL and cookie into the Apify actor, then run it.
- Download the results as a CSV file — this is your lead list with names, titles, companies, and LinkedIn URLs.
Part 2: Score Your Leads with AI
Use Claude Code to automatically analyze and rank your leads so you focus on the highest-value prospects first.
5Install Claude Code
- Download Visual Studio Code (free) if you don't have it.
- Open the terminal in VS Code: press
Ctrl + `(orCmd + `on Mac). - Install Claude Code:
npm install -g @anthropic-ai/claude-code - If you don't have Node.js, download it first from nodejs.org (choose LTS).
- Type
claudein the terminal and follow the prompts to log in with your Anthropic account.
6Connect Apify to Claude Code
Claude Code needs access to Apify to scrape detailed profile data for scoring. Set up the MCP connection:
- In your Apify account, go to Settings → Integrations and copy your Personal API Token.
- Create a new project folder for your outreach system. Open it in VS Code.
- Inside the project folder, create a file called
.mcp.jsonwith this content:{ "mcpServers": { "apify": { "command": "npx", "args": [ "-y", "@apify/actors-mcp-server", "--actors", "curious_coder/linkedin-profile-scraper" ], "env": { "APIFY_TOKEN": "your_apify_token_here" } } } } - Replace
your_apify_token_herewith your actual Apify API token. - Start Claude Code by typing
claudein the terminal. It will automatically detect the Apify MCP connection.
7Add Your Anthropic API Key
Claude Code uses your Anthropic account to process and score leads. If you don't have an API key yet:
- Go to console.anthropic.com
- Create an account and go to API Keys.
- Generate a new key and add some credits ($5–10 is enough to score hundreds of leads).
- When you run
claudein the terminal for the first time, it will ask for your API key. Paste it in.
8Score Your Leads
- Drop the CSV file you downloaded from Apify into your project folder.
- Tell Claude Code to score the leads. Example prompt:
Read the leads CSV file. For each lead, use the Apify LinkedIn profile scraper to get their full profile data. Then score each lead 1-10 on these criteria: - Buying power (founder/CEO = high, manager = medium) - LinkedIn activity (posts frequently = high, rarely = low) - Seniority level (C-suite = high, mid-level = lower) Save the results to a new CSV with all original data plus scores and a total score. Sort by total score descending. - Claude Code will go through each lead, scrape their full LinkedIn profile, analyze the data, and produce a ranked list.
- Customize the scoring criteria for your business. The example above is for selling a B2B service to founders — adjust the metrics based on who your ideal client is.
Part 3: Generate Value-Driven Outreach
Instead of sending generic cold messages, lead with free value. Give people something useful — that's how you start real conversations.
9Set Up Kie.ai for Image Generation
To send truly personalized value, you can generate improved LinkedIn visuals (banner, profile picture, headline) for each lead. This is the "lead with value" approach.
- Go to kie.ai and create an account.
- Kie.ai uses the Nana Banana Pro model for high-quality image generation. Get your API key from your account settings.
- Create a
.envfile in your project folder:KIE_API_KEY=your_kie_api_key_here
10Generate Personalized Outreach for Each Lead
- With your scored leads and tools connected, ask Claude Code to generate personalized outreach. Example prompt:
Take the top 20 leads from the scored CSV. For each person: 1. Read their LinkedIn profile data 2. Generate an improved LinkedIn headline for them 3. Use Kie.ai to create an improved LinkedIn banner that matches their personal brand 4. Write a short, friendly LinkedIn message that says: "Hey [name], I put together an improved headline and banner for your LinkedIn profile — totally free, no strings attached. Here they are: [attach]. Hope you find it useful!" Save each person's assets in a separate folder. - Claude Code will generate custom deliverables for each lead — an improved headline, a professional banner, and a personalized message.
- The key principle: you're leading with genuine free value. You're not pitching — you're helping. That's what starts conversations.
11Send Your Messages
- Review the generated messages and assets for each lead. Make sure they look good and feel personal.
- Send the messages manually on LinkedIn, or use Claude Cowork to help you send them at scale (with your approval on each message).
- Keep it natural: send 10–20 messages per day. Personalization means nothing if you blast 200 people at once.
- When someone responds positively, continue the conversation naturally. The free value opened the door — now build the relationship.
12Adapt This System to Your Business
The example above generates LinkedIn profile improvements — that works if you sell LinkedIn services. But the framework works for any business:
- SEO agency: generate a free mini SEO audit for each lead's website
- Design studio: create a quick mockup improving their landing page
- Sales training: record a personalized video reviewing their outreach
- Marketing consultant: write 3 sample social media posts for their brand
- SaaS company: build a custom demo or ROI calculation for their use case
The principle is always the same: lead with value, not a pitch. Claude Code can generate whatever free deliverable makes sense for your offer.
More Free Resources
Claude Code for Content
Set up AI-powered content creation with Claude Code
Claude Code for Marketing
Set up AI-powered marketing workflows with Claude Code
Viral Instagram Reels
Create scroll-stopping Reels with Claude Code and AI
Claude Code for SEO
SEO-optimize your website using Claude Code and AI
Authentic LinkedIn System
Free n8n workflow for authentic personal brand content
Viral LinkedIn Content Machine
Free n8n workflow for automated viral LinkedIn posts
LinkedIn AI Blueprint
4 free AI tools for LinkedIn growth
Claude Cowork for Outreach
Automate LinkedIn lead gen and cold outreach with AI
Free Profile Analysis
Get a personalized AI analysis of your LinkedIn profile
Top LinkedIn Profiles
Study what makes high-performing LinkedIn profiles work