Lead Enrichment Engine

Prerequisites

• User must be logged into LinkedIn in the browser (for profile research)
• Google Sheets access for export (optional)

Connect Services (if exporting to Sheets)

await requestApproval({
  reason: "Need Sheets access to export enriched prospect list",
  type: "service_auth",
  payload: { provider: "google", service: "drive" }
})

Safety Rules (CRITICAL)

• Maximum 30 LinkedIn profile visits per session. Hard limit.
• Wait 8-15 seconds between profile visits (randomized).
• Maximum 5 company page visits per session.
• Maximum 3 LinkedIn search result pages per query.
• Wait 5-8 seconds between search page loads.
• Read-only: No connection requests, messages, likes, or comments.
• Stop immediately on any CAPTCHA, verification prompt, or rate-limit warning.
• Randomize all wait times: base_delay + random(0, extra_seconds).

Workflow A: Company Intelligence Gathering

Step 1: Accept Input

// Option 1: User provides company names directly
var companies = ["Acme Corp", "Widget Inc", "StartupXYZ"]

// Option 2: Read from Google Sheet
var sheetData = await google.sheets.getValues({
  spreadsheetId: userSheetId,
  range: "Sheet1!A2:A"  // Company name column
})
var companies = sheetData.map(row => row[0]).filter(c => c && c.trim())
console.log("Target companies: " + companies.length)

Step 2: Google Search for Company Intel

For each company, search Google for basic information:

var companyIntel = []
var page = await browser.newtab("https://www.google.com")

for (var i = 0; i < companies.length; i++) {
  var company = companies[i]
  await page.goto({ url: "https://www.google.com/search?q=" + encodeURIComponent(company + " company") })
  await page.wait({ waitTime: 3 + Math.floor(Math.random() * 3) })
  var snap = await page.snapshot()

  // Extract from Knowledge Panel and search results:
  var intel = {
    name: company,
    industry: "",        // From Knowledge Panel or meta descriptions
    hq: "",              // Headquarters location
    employeeCount: "",   // Approximate headcount
    founded: "",         // Year founded
    description: "",     // What they do (one line)
    website: "",         // Company website URL
    fundingStage: "",    // If startup: Seed, Series A/B/C etc.
    recentNews: []       // Will be filled in Step 3
  }

  // Parse snapshot for company data
  // Look for: Knowledge Panel, Wikipedia excerpt, Crunchbase, LinkedIn company page

  companyIntel.push(intel)
  console.log("Researched " + (i + 1) + "/" + companies.length + ": " + company)
}

Step 3: LinkedIn Company Page

for (var i = 0; i < companyIntel.length; i++) {
  var intel = companyIntel[i]

  // Search for company on LinkedIn
  await page.goto({
    url: "https://www.linkedin.com/search/results/companies/?keywords=" + encodeURIComponent(intel.name)
  })
  await page.wait({ waitTime: 5 + Math.floor(Math.random() * 3) })
  var snap = await page.snapshot()

  // Click on first matching company result
  // Navigate to /about/ page for detailed info

  // Extract:
  // - Company size (exact range like "51-200 employees")
  // - Industry category
  // - Specialties
  // - Headquarters
  // - Founded year
  // - Company type (Public, Privately Held, etc.)
  // - Associated LinkedIn URL (for future use)

  await page.wait({ waitTime: 8 + Math.floor(Math.random() * 5) })
}

Step 4: Google News Check

for (var i = 0; i < companyIntel.length; i++) {
  var intel = companyIntel[i]

  // Search Google News (past 30 days)
  await page.goto({
    url: "https://www.google.com/search?q=" + encodeURIComponent(intel.name) + "&tbm=nws&tbs=qdr:m"
  })
  await page.wait({ waitTime: 3 + Math.floor(Math.random() * 2) })
  var snap = await page.snapshot()

  // Extract top 3 news headlines:
  // - Fundraising announcements
  // - Product launches
  // - Executive hires/departures
  // - Partnerships
  // - Awards/recognition

  // These become personalization hooks for outreach
  intel.recentNews = [
    // { headline: "Acme raises $30M Series B", date: "Apr 8, 2026", source: "TechCrunch" }
  ]
}

Workflow B: Decision Maker Discovery

Step 1: Search LinkedIn for Target Roles

var prospects = []
var targetRoles = ["VP Marketing", "Head of Growth", "CMO"] // User-configurable

for (var i = 0; i < companyIntel.length; i++) {
  var intel = companyIntel[i]

  for (var role of targetRoles) {
    var searchQuery = role + " " + intel.name
    await page.goto({
      url: "https://www.linkedin.com/search/results/people/?keywords=" + encodeURIComponent(searchQuery) + "&origin=GLOBAL_SEARCH_HEADER"
    })
    await page.wait({ waitTime: 5 + Math.floor(Math.random() * 3) })
    var snap = await page.snapshot()

    // Parse search results (10 per page, only use first page)
    // For each result extract:
    // - Full name
    // - Headline (title + company)
    // - Location
    // - Profile URL
    // - Connection degree (1st, 2nd, 3rd)
    // - Number of mutual connections

    // Pick the best match (title contains target role + company matches)
    break // One search per company is enough for safety
  }

  // Rate limiting between companies
  await page.wait({ waitTime: 5 + Math.floor(Math.random() * 5) })
}

Step 2: Profile Enrichment (Top Prospects)

For the best-matched prospect at each company, visit their full profile:

for (var i = 0; i < prospects.length; i++) {
  var prospect = prospects[i]

  await page.goto({ url: prospect.profileUrl })
  await page.wait({ waitTime: 8 + Math.floor(Math.random() * 7) }) // 8-15 second delay
  var snap = await page.snapshot()

  // Extract enriched data:
  prospect.enriched = {
    fullTitle: "",           // Current title at current company
    tenure: "",              // How long in current role
    previousCompanies: [],   // Last 2-3 positions
    education: "",           // Degree + school
    skills: [],              // Top endorsed skills
    about: "",               // Summary/About section (first 200 chars)
    recentPosts: [],         // Last 3 posts (topic + date)
    isActive: false,         // Posted in last 30 days?
    mutualConnections: [],   // Names of shared connections (if visible)
  }

  console.log("Enriched " + (i + 1) + "/" + prospects.length + ": " + prospect.name)
}

Step 3: Check Recent Activity (Optional)

// For prospects where you need to know if they're active posters:
await page.goto({ url: prospect.profileUrl + "/recent-activity/all/" })
await page.wait({ waitTime: 5 })
var activitySnap = await page.snapshot()

// Check: When was their last post? What topics?
// If posted in last 30 days: isActive = true
// If last 7 days: great candidate for warm-up engagement before outreach

Workflow C: Prospect Scoring

Score each prospect (0-100) based on fit:

for (var prospect of prospects) {
  var score = 0
  var breakdown = []

  // Role match (0-30)
  if (exactTitleMatch(prospect.enriched.fullTitle, targetRoles)) {
    score += 30; breakdown.push("Role match: +30")
  } else if (partialTitleMatch(prospect.enriched.fullTitle, targetRoles)) {
    score += 15; breakdown.push("Partial role match: +15")
  }

  // LinkedIn activity (0-20)
  if (prospect.enriched.isActive) {
    score += 20; breakdown.push("Active on LinkedIn: +20")
  }

  // Connection degree (0-25)
  if (prospect.connectionDegree === "1st") {
    score += 25; breakdown.push("1st degree connection: +25")
  } else if (prospect.connectionDegree === "2nd") {
    score += 15; breakdown.push("2nd degree connection: +15")
  } else {
    score += 5; breakdown.push("3rd degree connection: +5")
  }

  // Company signals (0-10)
  if (prospect.companyIntel.recentNews.length > 0) {
    score += 10; breakdown.push("Recent company news (outreach hook): +10")
  }

  // Tenure stability (0-10)
  if (tenureMonths(prospect.enriched.tenure) >= 12) {
    score += 10; breakdown.push("Stable tenure >1yr (decision power): +10")
  }

  // Mutual connections (0-5)
  if (prospect.enriched.mutualConnections.length > 0) {
    score += 5; breakdown.push("Has mutual connections: +5")
  }

  prospect.score = score
  prospect.scoreBreakdown = breakdown
}

// Sort by score descending
prospects.sort((a, b) => b.score - a.score)

Workflow D: Dossier Generation & Export

Generate Outreach Hooks

For each prospect, generate 2-3 personalized outreach hooks:

for (var prospect of prospects) {
  prospect.outreachHooks = []

  // Hook 1: News-based (if available)
  if (prospect.companyIntel.recentNews.length > 0) {
    var news = prospect.companyIntel.recentNews[0]
    prospect.outreachHooks.push({
      type: "news",
      hook: "Congrats on [" + news.headline + "]. [Relevant observation about what this means for their role]."
    })
  }

  // Hook 2: Post-based (if they're active)
  if (prospect.enriched.recentPosts.length > 0) {
    var post = prospect.enriched.recentPosts[0]
    prospect.outreachHooks.push({
      type: "post",
      hook: "Your recent post about [" + post.topic + "] resonated - especially [specific point]. [Bridge to your value prop]."
    })
  }

  // Hook 3: Mutual connection
  if (prospect.enriched.mutualConnections.length > 0) {
    prospect.outreachHooks.push({
      type: "mutual",
      hook: "We're both connected to [" + prospect.enriched.mutualConnections[0] + "]. [Reason for reaching out]."
    })
  }

  // Hook 4: Company growth signal
  if (prospect.companyIntel.employeeCount && parseInt(prospect.companyIntel.employeeCount) > 50) {
    prospect.outreachHooks.push({
      type: "growth",
      hook: "I noticed [Company] is growing the [department] team. [Observation about scaling challenges at this stage]."
    })
  }
}

Present Dossier Report

## Prospect Dossiers (5 companies researched)

### 1. Jane Doe - VP Marketing @ Acme Corp (Score: 85/100)
| Field | Details |
|-------|---------|
| Title | VP Marketing (2 yrs 3 mo) |
| Company | Acme Corp - SaaS, 250 employees, Series B, SF |
| LinkedIn | linkedin.com/in/janedoe |
| Connection | 2nd degree (3 mutual) |
| Activity | Active - posted 3 days ago about "content ops at scale" |
| Previous | Dir. Marketing @ PrevCo, Marketing Manager @ OtherCo |

**Company News:** "Acme raises $30M Series B" (TechCrunch, Apr 8)
**Best Outreach Hook:** "Congrats on the Series B! Scaling content ops is always
the fun challenge at that stage. Your recent post about it really resonated..."

**Score Breakdown:** Role +30, Active +20, 2nd degree +15, News hook +10, Stable tenure +10

Export to Google Sheet

var sheet = await google.sheets.createSpreadsheet({
  title: "Enriched Prospects - " + new Date().toISOString().slice(0, 10),
  sheets: ["Prospects", "Company Intel"]
})

// Prospects sheet
var prospectHeaders = [[
  "Score", "Name", "Title", "Company", "LinkedIn URL",
  "Connection Degree", "Mutual Connections", "Active on LinkedIn",
  "Recent Post Topic", "Tenure", "Best Outreach Hook", "Notes"
]]
var prospectRows = prospects.map(p => [
  p.score,
  p.name,
  p.enriched.fullTitle,
  p.companyIntel.name,
  p.profileUrl,
  p.connectionDegree,
  p.enriched.mutualConnections.join(", "),
  p.enriched.isActive ? "Yes" : "No",
  p.enriched.recentPosts.length > 0 ? p.enriched.recentPosts[0].topic : "",
  p.enriched.tenure,
  p.outreachHooks.length > 0 ? p.outreachHooks[0].hook : "",
  ""
])

await google.sheets.updateValues({
  spreadsheetId: sheet.id,
  range: "'Prospects'!A1",
  values: prospectHeaders.concat(prospectRows)
})

// Company Intel sheet
var companyHeaders = [[
  "Company", "Industry", "Size", "HQ", "Founded",
  "Funding Stage", "Website", "Recent News", "LinkedIn URL"
]]
var companyRows = companyIntel.map(c => [
  c.name, c.industry, c.employeeCount, c.hq, c.founded,
  c.fundingStage, c.website,
  c.recentNews.map(n => n.headline).join("; "),
  c.linkedinUrl || ""
])

await google.sheets.updateValues({
  spreadsheetId: sheet.id,
  range: "'Company Intel'!A1",
  values: companyHeaders.concat(companyRows)
})

console.log("Exported: https://docs.google.com/spreadsheets/d/" + sheet.id)

De-duplication

• Track all profile URLs visited across searches
• If the same person appears in multiple company searches, keep only the highest-scored instance
• Flag prospects who appeared in multiple queries (sign of strong ICP fit)

Error Handling

• LinkedIn CAPTCHA: Stop immediately, report to user how many prospects were enriched before stopping
• Profile not found / 404: Skip, note in output, move to next
• Google rate limit: Switch to a different search engine or wait 60 seconds
• Company not found on LinkedIn: Use Google results only, note limited data
• Sheet export fails: Present dossier in-chat as fallback