> ## 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-1: Personalization Principles > Before you automate personalization, you need to understand what actually makes outreach feel personal — and what's just performing personalization theater. **AP-1 · AI Personalization · 100 XP · \~17 min** Most "personalized" outreach isn't personalized — it's personalization theater. Swapping in a company name and a generic compliment about their recent LinkedIn post is not personalization. A recipient knows immediately whether you've actually engaged with their context or just populated a template. Genuine personalization is demonstrating understanding, not demonstrating effort. *** ## The Personalization Spectrum | Level | What It Looks Like | Perceived as | Effort | Scale | | ----------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------ | ------------------ | ------- | ------------------- | | **Level 0** | "Hi \[First Name], I wanted to reach out..." | Spam | Zero | Infinite | | **Level 1** | "Hi \[Name], I saw you work at \[Company]..." | Slightly less spam | Minimal | Infinite | | **Level 2** | "Hi \[Name], I saw \[Company] just raised a Series B..." | Mildly relevant | Low | High | | **Level 3** | "Hi \[Name], your Series B + the 8 SDR roles you're hiring suggests you're building the outbound motion now — that's exactly when \[specific problem] shows up..." | Genuine | Medium | Medium (without AI) | | **Level 4** | Level 3 + connected to their specific LinkedIn content, job history, company context | Remarkable | High | Low (without AI) | The goal of AI-assisted personalization is to produce Level 3–4 output at Level 1–2 effort. *** ## What Makes Personalization Feel Genuine Three elements separate genuine personalization from theater: **1. Specificity of observation** Generic: "I saw you're in sales leadership." Specific: "I saw you moved to RevOps after 6 years in field sales — that crossover usually comes with a specific frustration with visibility into rep activity." The specific version demonstrates you actually read something and connected it to an insight. **2. Connection to their current moment** Generic: "Companies like yours often struggle with outbound efficiency." Specific: "You're 3 months into a new VP of Sales role at a company that just doubled its SDR team — the infrastructure scaling problem usually hits right now." The specific version shows temporal awareness — you're reacting to their current situation, not their general category. **3. Relevance without presumption** Generic: "I know you must be very busy, but..." Specific: "If building the data enrichment layer for the new team is on your list for this quarter, I can show you what others at your stage have done in 20 minutes." The specific version makes a specific assumption about their priority without presuming you know their situation perfectly. *** ## The Three Personalization Layers Think of personalization as three nested layers: **Layer 1: Company-level context** What is this company doing right now? Recent funding, hiring patterns, product launches, market position. *Source: Signal Intelligence track signals* **Layer 2: Role-level context** What does someone in this role typically care about? What are the challenges, metrics, and pressures of this job title at this company stage? *Source: Persona research + job description analysis* **Layer 3: Individual-level context** What has this specific person written, published, commented on, or signaled through their LinkedIn activity? *Source: LinkedIn profile + content + activity* The most powerful personalization uses all three layers. The strongest single layer is Layer 1 (company-level) because it's the most actionable and the most unique to them. *** ## Personalization That Backfires Some "personalization" actively hurts response rates. Avoid: * **Over-researched stalker energy**: "I saw your post from 2019, your conference talk from 2021, and I know you went to Stanford..." — feels invasive, not thoughtful * **Flattery that feels manufactured**: "Your thought leadership on sales development has really been impressive" — nobody believes this * **Wrong-person personalization**: Referencing their role but getting the context wrong — worse than no personalization * **Old news as current news**: Referencing something that happened 4 months ago as if it just happened — signals stale research The test: would a smart person who received this email think "they actually understand my situation" or "this is clever automation"? If it's the latter, it needs work. *** **Quick Check:** What's the difference between Level 2 and Level 3 personalization? What are the three elements of genuine personalization? What makes personalization backfire? *** ## AP-1 Challenge: Personalization Audit (+100 XP) Take 10 cold emails you've received (or find examples online) and audit them. **Requirements:** * Classify each email on the personalization spectrum (Level 0–4) * Identify which of the three layers (company, role, individual) are present * For the 3 lowest-level emails, rewrite them to Level 3 using real or plausible signals * A paragraph explaining what distinguishes the best from the worst Submit your audit + rewrites. **+100 XP on approval.** *** The quality of your AI-generated copy is entirely determined by the quality of your prompts. AP-2 teaches prompt engineering specifically for outreach.