# clay_automation

> > **Purpose:** Build Clay enrichment tables, signal detection systems, and API integrations for GTM operations. Implement waterfall enrichment logic, credit-conscious workflows, and CRM sync patterns.

- Author: josephdeville
- Repository: josephdeville/GTMClaudSkills
- Version: 20260108150942
- Stars: 2
- Forks: 0
- Last Updated: 2026-02-07
- Source: https://github.com/josephdeville/GTMClaudSkills
- Web: https://mule.run/skillshub/@@josephdeville/GTMClaudSkills~clay_automation:20260108150942

---

# Clay Automation Engineering Skill

> **Purpose:** Build Clay enrichment tables, signal detection systems, and API integrations for GTM operations. Implement waterfall enrichment logic, credit-conscious workflows, and CRM sync patterns.

## When to Use This Skill

Use this skill when:
- Building Clay enrichment tables and workflows
- Implementing signal detection (tech stack, intent, pain points)
- Creating API integrations with HTTP enrichment
- Syncing Clay data to HubSpot/Salesforce
- Managing enrichment credits and optimization

---

## Core Architecture

### Waterfall Enrichment Logic

```
Primary Provider → Fallback Provider → Manual Flag
       ↓                   ↓                ↓
   (if found)        (if not found)    (needs review)
```

**Key Principles:**
- Deduplication FIRST, before any enrichment
- Conditional runs to avoid re-enriching cached data
- Weekly refresh cycles for credit budgeting
- Test batches (10-50 rows) before full execution

---

## Signal Table Structure

Build signal tables in three layers:

### Layer 1: Surface-Level Signals (Explicit)
- Job postings mentioning specific challenges
- G2 reviews highlighting issues
- LinkedIn posts from decision-makers about problems

### Layer 2: Subsurface Signals (Implicit)
- Tech stack gaps (missing complementary tools)
- Org changes (new hires in relevant departments)
- Funding events (expansion mandates)

### Layer 3: Signal Validation
- Use Claygent to verify signals are current
- Weight by recency and source quality
- Create composite scores from multiple indicators

---

## Table Types

| Table Type | Use Case | Refresh |
|------------|----------|---------|
| **Tech Stack** | Competitive displacement, integrations | Monthly |
| **Pain Point Mapping** | Industry-specific challenges | Quarterly |
| **Intent Signal** | High-intent prospect identification | Weekly |
| **Territory Intelligence** | Coverage and assignment | Monthly |

---

## Claygent Prompt Pattern

```
Search for recent LinkedIn posts or job listings from [Company]
that indicate they're experiencing [Specific Pain Point].

Look for:
- Job postings mentioning [keywords]
- Executive posts discussing [challenge]
- Growth announcements requiring [solution]

Return JSON:
{
  "signal_detected": true/false,
  "evidence": "brief quote or description",
  "confidence": "high/medium/low",
  "date": "when signal was observed"
}
```

---

## HubSpot Sync Patterns

### Before Creating Records
- Check for existing records (use lookup tables)
- Map columns to properties with consistent naming
- Include metadata (source, date, confidence)

### Sync Configuration
- "Update if exists, skip if not" for enrichment
- Timestamp tracking for last enrichment date
- Segmentation lists based on signal combinations

---

## Common Workflows

### Lead Enrichment
1. Import leads
2. Deduplicate against existing
3. Enrich company data (Clearbit, Apollo)
4. Enrich contact data (email, LinkedIn)
5. Run signal detection
6. Calculate composite lead score
7. Route to CRM

### Competitive Displacement
1. Identify accounts using competitor products
2. Enrich with decision-maker contacts
3. Detect change signals (funding, leadership, reviews)
4. Map pain points to differentiators
5. Generate personalized angles
6. Sync to CRM

---

## Credit Management

| Practice | Why |
|----------|-----|
| Test batches first | Avoid wasting credits on broken logic |
| Conditional enrichment | Don't re-enrich existing data |
| Monitor per provider | Identify expensive workflows |
| Prefer native integrations | Often cheaper than HTTP API |

---

## Quality Assurance Checklist

Before production:
- [ ] Verify accuracy on 10+ sample rows
- [ ] Test error handling (API failures)
- [ ] Confirm deduplication catches edge cases
- [ ] Validate signal scoring distributions
- [ ] Check CRM sync creates/updates correctly
- [ ] Document column purposes and sources

---

## Troubleshooting

| Issue | Solution |
|-------|----------|
| **Enrichment failures** | Check API keys, rate limits, input format |
| **Credit burn** | Add conditional logic, audit unnecessary calls |
| **Data quality** | Validation columns, multiple sources, manual review |
| **Sync errors** | Verify field mapping, check for required fields |

---

## Integration Points

| System | Pattern |
|--------|---------|
| **HubSpot** | Native integration for bulk ops |
| **Make/n8n** | Webhooks for complex orchestration |
| **Firecrawl** | HTTP enrichment for deep scraping |
| **Apollo/Clearbit** | Native providers for firmographics |