GitHub
All Skills
Outreach & Personalizationv1.0.4

Outreach Personalization Research

Research a prospect deeply and generate actionable personalization angles with draft opening lines for cold outreach.

Outreach Personalization Research

Research a prospect deeply and generate actionable personalization angles with draft opening lines for cold outreach.

Instructions

When a user wants to personalize outreach to a specific prospect, gather rich context and synthesize it into concrete angles and openers.

Steps

  1. Identify the prospect. Extract any available identifiers from the user's request:

    • LinkedIn URL
    • Email address
    • Name + company If the user provides minimal info, ask for at least a name and company or a LinkedIn URL.

  2. Enrich the person by calling mcp__claude_ai_Amplemarket__enrich_person with all available identifiers. Set reveal_email to true for contact details. Credit note: Enabling reveal_email and reveal_phone_numbers consumes additional Amplemarket credits per reveal. Key data points to extract:

    • Current title and role
    • Career history and previous companies
    • Education and schools
    • Skills and endorsements
    • Location

  3. Enrich the company by calling mcp__claude_ai_Amplemarket__enrich_company with the company domain or linkedin_url. Key data points to extract:

    • Industry and sub-industry
    • Company size and growth trajectory
    • Tech stack and tools used
    • Company description and value proposition
    • Headquarters and locations
    • Recent funding or milestones

  4. Search for additional context by calling mcp__claude_ai_Amplemarket__search_people with company_domains: [prospect's company domain], person_seniorities matching the prospect's level, page_size: 5 to understand the team structure around them.

  5. Identify personalization angles. Analyze the enriched data to find 3-5 angles from these categories:

    • Role-based: Pain points specific to their title and department (e.g., a VP of Sales likely cares about pipeline velocity)
    • Company stage: Tailor to their company size and growth phase (startup vs. enterprise challenges)
    • Career trajectory: Reference a recent role change, promotion, or career milestone
    • Tech stack: Mention specific tools they use and how your solution integrates or compares
    • Industry trends: Connect to challenges or opportunities in their specific industry
    • Shared context: Same alma mater, previous company, mutual connections, shared geography
    • Company news: Recent funding, product launches, hiring sprees, or expansions
    • Team structure: Reference their team size or recent hires in their department

  6. Draft opening lines. For each angle, write a concise, natural-sounding opening line (1-2 sentences) that:

    • Demonstrates specific research (not generic)
    • Connects to a plausible pain point or interest
    • Feels human, not templated
    • Avoids being sycophantic or over-the-top

  7. Format the output as a Personalization Brief:

    Prospect Summary (one-paragraph context)

    Personalization Angles (3-5 angles, each with):

    • Angle name and category
    • Supporting data point from enrichment
    • Draft opening line
    • Strength rating (Strong / Medium / Soft)

    Recommended Best Angle with reasoning

    Contact Details for executing the outreach

Important Notes

  • Quality over quantity: 3 strong angles are better than 5 weak ones.
  • Never fabricate data points. Only use information returned from the enrichment APIs.
  • If enrichment data is sparse, be transparent about it and focus angles on the data that is available.
  • Adapt tone guidance if the user specifies their outreach style (formal, casual, humorous).

Examples

Example 1: LinkedIn URL Personalization

User prompt: "Help me personalize outreach to linkedin.com/in/sarah-chen-vp-marketing"

What the skill does:

  1. Calls mcp__claude_ai_Amplemarket__enrich_person with linkedin_url: "https://linkedin.com/in/sarah-chen-vp-marketing", reveal_email: true.
  2. Calls mcp__claude_ai_Amplemarket__enrich_company with domain extracted from enrichment.
  3. Calls mcp__claude_ai_Amplemarket__search_people with the company domain and seniority filter to understand team structure.
  4. Synthesizes angles and drafts openers.

Example output:

PERSONALIZATION BRIEF: Sarah Chen

Prospect Summary Sarah Chen is the VP of Marketing at CloudMetrics (cloudmetrics.io), a Series B cloud analytics company with 180 employees based in Austin, TX. She joined 8 months ago from HubSpot where she was a Senior Director of Demand Gen. She holds an MBA from Wharton. CloudMetrics recently raised a $45M Series B and appears to be scaling their go-to-market team.

Personalization Angles

1. New Role, New Strategy (Career Trajectory) - Strong Data: Sarah joined CloudMetrics 8 months ago as VP Marketing, coming from HubSpot. Opening line: "Sarah, transitioning from a marketing powerhouse like HubSpot to lead marketing at a Series B company is a big shift. I imagine you are rebuilding the demand gen playbook from scratch rather than inheriting one."

2. Post-Funding Growth (Company News) - Strong Data: CloudMetrics raised $45M Series B; headcount grew from 120 to 180 in 6 months. Opening line: "With CloudMetrics' recent Series B and the team nearly doubling, I'd guess the pressure to scale pipeline in lockstep with headcount growth is very real right now."

3. Tech Stack Gap (Technology) - Medium Data: CloudMetrics uses Salesforce, Marketo, and Google Analytics but no ABM or intent data platform detected. Opening line: "I noticed CloudMetrics is running Marketo and Salesforce but doesn't seem to have an intent data layer yet. Curious if that's on your roadmap as you scale outbound."

4. Team Building (Team Structure) - Medium Data: Search shows 3 marketing hires in the last 3 months, all junior/mid-level. Opening line: "Looks like you have been building out the marketing team quickly at CloudMetrics. When you are scaling a team that fast, having the right tooling in place before the team outgrows it can save a lot of rework."

5. Wharton Connection (Shared Context) - Soft Data: Sarah holds an MBA from Wharton. Opening line: "Fellow Wharton alum here. Saw you are leading marketing at CloudMetrics and wanted to reach out." (Only use if sender is also a Wharton alum.)

Recommended Best Angle: Angle 1 (New Role) or Angle 2 (Post-Funding Growth). Both are timely, specific, and connect to real pain points. Combine them for maximum impact.

Contact Details

Example 2: Name + Company Personalization

User prompt: "Find personalization angles for James Park, CTO at Notion"

What the skill does:

  1. Calls mcp__claude_ai_Amplemarket__enrich_person with name: "James Park", company_name: "Notion", reveal_email: true.
  2. Calls mcp__claude_ai_Amplemarket__enrich_company with domain: "notion.so".
  3. Searches for engineering leadership context at Notion.
  4. Returns personalization brief with CTO-specific angles (technical challenges, engineering team growth, infrastructure decisions).

Example 3: Personalization with Style Guidance

User prompt: "Write an opener for maria@finova.io - keep it casual and short"

What the skill does:

  1. Enriches Maria via email.
  2. Enriches finova.io.
  3. Generates angles with casual, concise opening lines (1 sentence max, conversational tone).

Troubleshooting

ProblemSolution
Very sparse enrichment dataFallback chain: 1) Try additional identifiers (LinkedIn URL, company domain) to improve match quality. 2) Focus angles on company-level data (industry, size, funding, hiring signals) when person data is thin. 3) Search for the prospect's peers at the same company to infer team structure and role context. 4) Be transparent: "Limited data on this person. Angles are based primarily on company signals."
No email foundSuggest LinkedIn InMail or connection request. Provide the draft opener adapted for LinkedIn format.
Angles feel genericEnsure you are pulling specific data points. If enrichment is thin, recommend the user share any additional context they have about the prospect.
User wants more anglesSearch for the prospect's content (posts, articles) or expand the company enrichment to find additional data points.
Wrong person enrichedVerify with the user by confirming title and company. Fallback: use company_domain instead of company_name for disambiguation, or ask for LinkedIn URL for exact match.
Person enrichment succeeds but company enrichment failsFallback chain: 1) Try enrich_company with domain instead of name. 2) Try LinkedIn company URL. 3) Use person-level data for role-based and career trajectory angles, and note: "Company data unavailable. Angles focus on the prospect's career and role."
Multiple people match the same name + companyPresent all matches with titles and LinkedIn URLs. Ask the user to confirm before generating angles. Never generate personalization for the wrong person.
Company exists but search_companies returns 0 resultsFallback chain: 1) Try enrich_company with domain directly. 2) Try the LinkedIn company URL. 3) Try parent company domain. Some companies are enrichable by domain but not searchable by name.