> ## Documentation Index > Fetch the complete documentation index at: https://docs.bitscale.ai/llms.txt > Use this file to discover all available pages before exploring further. # AP-4: LinkedIn Research Automation > LinkedIn is the richest source of individual-level context for B2B personalization. Learn to extract, process, and activate LinkedIn data at scale without manual research. **AP-4 · AI Personalization · 100 XP · \~18 min** LinkedIn contains everything you need for genuine personalization: career history, current challenges (through posts), company changes, content preferences, and professional identity. The problem is extracting it at scale without spending 15 minutes per contact on manual research. This module covers the tools and workflows that automate LinkedIn research while preserving the depth that makes personalization genuine. *** ## What to Extract from LinkedIn Not everything on a LinkedIn profile is useful for personalization. Here's what matters: | Data Point | Source | Personalization Use | | ---------------------------------- | --------------------- | ---------------------------------------------------------- | | **Current title and company** | Profile | Basic context (already have this) | | **Career trajectory** | Experience section | "You've moved from IC to VP in 4 years" | | **Career pivot context** | Experience section | "You came to sales ops from engineering — rare background" | | **Educational background** | Education section | Rarely useful unless specific (MBA, specific bootcamp) | | **Recent posts (last 30 days)** | Activity feed | Strongest individual signal | | **Post engagement topics** | Liked/commented posts | What they pay attention to | | **LinkedIn summary / About** | Profile | Self-described priorities and beliefs | | **Skills and endorsements** | Skills section | Tool expertise, specializations | | **Recommendations given/received** | Recommendations | Character signals, professional relationships | The highest-value signals are: **recent post content** (what they're thinking about now) and **career trajectory** (what shaped their professional lens). *** ## LinkedIn Data Extraction Tools | Tool | Method | Best For | | ---------------------------- | ---------------------------------- | ------------------------------------------- | | **Proxycurl API** | Official LinkedIn data API | Profile data, current role, work history | | **PhantomBuster** | Automation (check ToS) | Activity scraping, post content | | **Evaboot** | Sales Navigator export | Bulk profile extraction | | **LinkedIn Sales Navigator** | Native | List-level profile access, alerts | | **Clay** | LinkedIn enrichment integration | Profile + recent activity in one enrichment | | **Manual + AI** | Copy profile → paste into Bitscale | High-value accounts, full depth | *** ## Building LinkedIn Research Columns in Bitscale ### Career trajectory analysis: ``` Analyze this career history and extract the key trajectory signal: Name: {{first_name}} {{last_name}} Work history (most recent first): {{work_history}} What is the most interesting or relevant aspect of their career trajectory for outreach? Focus on: - Career pivots (IC to leadership, function changes) - Speed of advancement - Industry moves - Founder background - Notable company pedigree (known companies = credibility) Return: a 1-sentence observation that would be natural to reference in an outreach email. If nothing notable, return "standard progression". ``` ### Recent content analysis: ``` Analyze these recent LinkedIn posts/comments and identify the dominant professional themes: Posts: {{recent_activity_text}} Identify: 1. Primary topic they post about most (e.g., "sales development best practices", "RevOps data quality") 2. Their apparent point of view (what do they seem to believe strongly?) 3. Any specific problems or frustrations they've expressed 4. Outreach hook: a one-sentence reference to their content that would feel genuine Return as JSON. ``` ### Profile summary extraction: ``` Extract the key professional positioning from this LinkedIn About section: About text: {{linkedin_about}} Identify: 1. How they describe their professional identity 2. What they say they're focused on now 3. Any specific methodologies, frameworks, or beliefs they mention 4. One quote or phrase that captures their professional voice (for tone-matching in copy) Return as JSON. ``` *** ## From Research to Personalization Hook Once you've extracted LinkedIn signals, you need to convert them into email hooks: ``` Given the following LinkedIn research on {{first_name}} {{last_name}}: Career observation: {{career_trajectory_observation}} Content theme: {{primary_content_topic}} Point of view: {{apparent_pov}} Profile summary signal: {{profile_summary_signal}} Write 3 potential personalization hooks — each a single opening sentence for a cold email. Requirements for each hook: - References a specific observation from the research (not generic) - Leads naturally into a business conversation - Reads as human, not automated - No compliments or flattery - Under 25 words Return 3 numbered hooks. ``` Then select the best hook programmatically: ``` Rate these 3 personalization hooks for naturalness and specificity (1-10 each): Hook 1: {{hook_1}} Hook 2: {{hook_2}} Hook 3: {{hook_3}} Select the best one. Return: {"best_hook": "hook_1/hook_2/hook_3", "selected_text": "...", "reason": "brief explanation"} ``` *** **Quick Check:** What are the two highest-value LinkedIn data points for personalization? Which extraction tool is best for bulk profile extraction? What makes a personalization hook feel genuine vs. manufactured? *** ## AP-4 Challenge: LinkedIn Research Automation (+100 XP) Automate LinkedIn research for 20 contacts using any combination of tools. **Requirements:** * Career trajectory column for all 20 contacts * Recent content theme column (with evidence) * 3 personalization hook options per contact * Best hook selection column * Final email using the selected hook as opener * A short reflection: what LinkedIn signal type produced the best hooks? Share your grid + reflection paragraph. **+100 XP on approval.** *** Personalization isn't just individual — it's also vertical. AP-5 covers how to build industry-specific messaging frameworks that scale.