data-enrichment-integration

Savant verdict: Review advised

Security analysis or source-hub review identified patterns requiring attention.

Live evaluation

Not evaluated yet. Workspaces can request a live evaluation.

Security assessment

Scan pending.

Structure

• No license declaredConfirm you may reuse this skill before importing it into your repository.
• The source hub's security checks flagged this skillSnyk: warn (Risk: MEDIUM · 1 issue)

SKILL.md

---
name: data-enrichment-integration
description: When the user wants to build or improve a sales bot's ability to pull in firmographic or contact data mid-conversation. Also use when the user mentions "data enrichment," "lead enrichment," "pulling company data," "contact data lookup," or "real-time data."
---

# Data Enrichment Integration for Sales Bots

You are an expert in integrating data enrichment into automated sales systems. Your goal is to help design bots that pull in firmographic, technographic, and contact data to enhance conversations and personalization.

## Initial Assessment

Before providing guidance, understand:

1. **Context**
   - What data do you currently have on leads?
   - What data would improve your conversations?
   - What enrichment providers do you use or consider?

2. **Current State**
   - How is enrichment happening today?
   - What data gaps exist?
   - What's the quality of your current data?

3. **Goals**
   - What would better enrichment enable?
   - What decisions depend on enriched data?

---

## Core Principles

### 1. Enrich at the Right Time
- Too early wastes money on unqualified leads
- Too late misses personalization opportunity
- Balance cost with value

### 2. Use Data Wisely
- Having data ≠ sharing data
- Personalize without being creepy
- Quality over quantity

### 3. Keep Data Fresh
- Data decays quickly
- Re-enrich periodically
- Validate before using

### 4. Fallback Gracefully
- Enrichment can fail
- Have conversation paths for missing data
- Don't break the flow

---

## Types of Enrichment Data

### Firmographic Data

**Company information:**
- Company name (official)
- Industry/sector
- Employee count
- Revenue range
- Location(s)
- Founded date
- Company type (public, private, startup)

**Uses:**
- Qualification
- Routing
- Personalization
- Segmentation

### Contact Data

**Person information:**
- Full name
- Job title
- Department
- Seniority level
- Email (verified)
- Phone (direct)
- LinkedIn profile
- Employment history

**Uses:**
- Personalization
- Multi-threading
- Authority assessment
- Outreach targeting

### Technographic Data

**Technology stack:**
- Software used
- Platforms
- Integrations
- Tech categories

**Uses:**
- Competitive insight
- Integration selling
- Technical fit
- Use case targeting

### Intent Data

**Buying signals:**
- Topic research
- Competitor visits
- Content downloads
- Review site activity

**Uses:**
- Timing outreach
- Prioritization
- Message relevance
- Trigger-based engagement

---

## Enrichment Providers

### Major Providers

| Provider | Strengths | Best For |
|----------|-----------|----------|
| Clearbit | Firmographic, real-time | B2B SaaS |
| ZoomInfo | Contact data depth | Outbound prospecting |
| Apollo | Contact + enrichment | Volume outreach |
| Lusha | Direct dials | Phone outreach |
| 6sense | Intent data | ABM |
| Bombora | Intent signals | Enterprise |
| BuiltWith | Technographics | Tech targeting |

### API Considerations

**Evaluate providers on:**
- Data accuracy
- Coverage (% of lookups successful)
- Freshness
- API reliability
- Pricing model
- Rate limits

---

## When to Enrich

### Inbound Lead Enrichment

```
Trigger: Form submission or chat initiation
Timing: Immediate

Flow:
1. Lead submits email
2. Enrich on email (async)
3. Use data in conversation
4. Store for future use
```

**What to enrich:**
- Company info
- Contact info
- Role and seniority

### Outbound Pre-Enrichment

```
Trigger: Building prospect list
Timing: Before outreach

Flow:
1. Import contact list
2. Batch enrich
3. Score and segment
4. Personalize outreach
```

**What to enrich:**
- Complete firmographic profile
- Verified contact info
- Technographic fit
- Intent signals

### Mid-Conversation Enrichment

```
Trigger: Learn new information in conversation
Timing: During conversation

Flow:
1. User mentions company name
2. Real-time lookup
3. Adjust conversation with context
4. Store enriched data
```

**What to enrich:**
- Company details mentioned
- Verify information shared
- Find additional contacts

---

## Implementation

### Basic Enrichment Flow

```
function enrichLead(email) {
  // Check cache first
  cached = getFromCache(email)
  if (cached && !isStale(cached)) {
    return cached
  }

  // Call enrichment API
  try {
    enriched = enrichmentProvider.lookup(email)

    // Validate and clean
    validated = validateEnrichment(enriched)

    // Store in cache and CRM
    storeEnrichment(email, validated)

    return validated
  } catch (error) {
    logError(error)
    return null  // Graceful fallback
  }
}
```

### Real-Time Enrichment

```
function enrichDuringConversation(message, context) {
  // Extract signals from message
  company_mentioned = extractCompany(message.content)
  email_provided = extractEmail(message.content)

  // Enrich asynchronously
  if (company_mentioned && !context.company_data) {
    enrichCompanyAsync(company_mentioned, context)
  }

  if (email_provided && !context.contact_enriched) {
    enrichContactAsync(email_provided, context)
  }

  // Continue conversation (don't block)
  return generateResponse(message, context)
}

async function enrichCompanyAsync(company, context) {
  data = await enrichmentProvider.lookupCompany(company)
  context.company_data = data
  // Enriched data available for next turn
}
```

### Batch Enrichment

```
function batchEnrich(leads, options) {
  results = []

  for (batch in chunk(leads, options.batch_size)) {
    // Respect rate limits
    await rateLimiter.waitForSlot()

    // Parallel enrichment within batch
    batch_results = await Promise.all(
      batch.map(lead => enrichLead(lead.email))
    )

    results.push(...batch_results)

    // Progress tracking
    reportProgress(results.length, leads.length)
  }

  return results
}
```

---

## Using Enriched Data

### In Qualification

```
function qualifyWithEnrichment(lead) {
  score = 0

  // Employee count scoring
  if (lead.enriched.employee_count >= 500) score += 3
  else if (lead.enriched.employee_count >= 50) score += 2
  else if (lead.enriched.employee_count >= 10) score += 1

  // Industry fit
  if (IDEAL_INDUSTRIES.includes(lead.enriched.industry)) score += 2

  // Title/seniority
  if (lead.enriched.seniority == "c_suite") score += 3
  else if (lead.enriched.seniority == "director") score += 2
  else if (lead.enriched.seniority == "manager") score += 1

  // Technology fit
  if (usesCompetitor(lead.enriched.technologies)) score += 2
  if (usesComplementaryTech(lead.enriched.technologies)) score += 1

  return score
}
```

### In Personalization

**Use data naturally:**
"I see you're based in Austin—great tech scene there!"
"With a team of [employee count], scaling [process] is probably a focus."

**Don't overdo it:**
Bad: "I see you've been at [Company] for 3 years, previously worked at [Old Company], studied at [University]..."
Good: "I noticed you're focused on [role/department]—[relevant point]."

### In Routing

```
function routeLead(lead) {
  // Enterprise routing
  if (lead.enriched.employee_count > 1000) {
    return assignToEnterpriseTeam(lead)
  }

  // Geographic routing
  if (lead.enriched.country != "US") {
    return assignToInternationalTeam(lead)
  }

  // Industry routing
  if (lead.enriched.industry in SPECIALIZED_INDUSTRIES) {
    return assignToIndustrySpecialist(lead)
  }

  // Standard routing
  return standardRoundRobin(lead)
}
```

---

## Data Quality Management

### Validation Rules

```
function validateEnrichment(data) {
  validated = {}

  // Company name - clean and standardize
  if (data.company_name) {
    validated.company_name = cleanCompanyName(data.company_name)
  }

  // Employee count - range validation
  if (data.employee_count && data.employee_count > 0) {
    validated.employee_count = data.employee_count
  }

  // Email - format validation
  if (data.email && isValidEmail(data.email)) {
    validated.email = data.email.toLowerCase()
  }

  // Phone - format and validate
  if (data.phone && isValidPhone(data.phone)) {
    validated.phone = formatPhone(data.phone)
  }

  return validated
}
```

### Freshness Management

```
function isStale(enrichment_record) {
  days_old = daysSince(enrichment_record.enriched_at)

  // Different freshness rules by data type
  if (enrichment_record.type == "contact") {
    return days_old > 30  // Contact data stales faster
  }

  if (enrichment_record.type == "company") {
    return days_old > 90  // Company data more stable
  }

  if (enrichment_record.type == "intent") {
    return days_old > 7  // Intent data very time-sensitive
  }

  return days_old > 60  // Default
}
```

### Handling Missing Data

```
function getCompanySize(lead) {
  if (lead.enriched.employee_count) {
    return lead.enriched.employee_count
  }

  // Fallback: ask in conversation
  return askInConversation("company_size")
}

function personalizeMessage(template, lead) {
  // Use enriched data where available
  company = lead.enriched.company_name || lead.stated_company || "your company"
  industry = lead.enriched.industry || null

  message = template.replace("{company}", company)

  // Only include industry reference if we have it
  if (industry) {
    message = message.replace("{industry_reference}",
      `in the ${industry} space`)
  } else {
    message = message.replace("{industry_reference}", "")
  }

  return message
}
```

---

## Cost Management

### Enrichment Economics

**Cost factors:**
- Per-lookup pricing
- Volume discounts
- Multiple providers
- Cache hit rates

**Optimization:**
- Enrich on qualification (not every lead)
- Cache aggressively
- Use cheaper providers for basic data
- Premium providers for contact data

### Tiered Enrichment

```
function enrichByValue(lead) {
  lead_score = quickScore(lead)

  if (lead_score >= 80) {
    // High potential: full enrichment
    return fullEnrichment(lead)
  }

  if (lead_score >= 50) {
    // Medium potential: basic enrichment
    return basicEnrichment(lead)
  }

  // Low potential: minimal enrichment
  return minimalEnrichment(lead)
}

function fullEnrichment(lead) {
  // Firmographic + contact + technographic + intent
  return multiProviderEnrich(lead, ["clearbit", "zoominfo", "bombora"])
}

function basicEnrichment(lead) {
  // Firmographic + contact only
  return singleProviderEnrich(lead, "clearbit")
}

function minimalEnrichment(lead) {
  // Just verify email
  return emailVerify(lead.email)
}
```

---

## Common Mistakes

### 1. Over-Enriching
**Problem:** Enriching every lead, wasting money
**Fix:** Qualify first, enrich qualified leads

### 2. Stale Data
**Problem:** Using old enrichment data
**Fix:** Check freshness, re-enrich periodically

### 3. Creepy Personalization
**Problem:** Using too much data visibly
**Fix:** Subtle use, don't reveal everything you know

### 4. Single Provider Dependency
**Problem:** One provider fails, everything breaks
**Fix:** Fallback providers, graceful degradation

### 5. Blocking on Enrichment
**Problem:** Conversation waits for enrichment
**Fix:** Async enrichment, proceed without if needed

---

## Questions to Ask

If you need more context:
1. What data do you currently have on leads?
2. What enrichment providers do you use?
3. What's your lead volume?
4. What decisions depend on enriched data?
5. What's your budget for enrichment?

---

## Related Skills

- **lead-qualification-logic**: Using enrichment for scoring
- **personalization-at-scale**: Using enrichment for personalization
- **conversation-memory**: Storing enriched data
- **multi-channel-coordination**: Using data across channels