Event-Triggered Enrichment: Real-Time Data Enhancement

Batch enrichment is useful, but event-triggered enrichment changes the game. Instead of enriching on a schedule, you enrich at the exact moment data is needed—when a lead submits a form, when an opportunity advances, when an account shows intent.

This guide covers how to design and implement event-triggered enrichment workflows.

Why Event-Triggered?

Batch vs Event-Triggered

Batch Enrichment:
─────────────────────────────────────────
Monday: 5,000 leads enriched overnight
Tuesday: New lead arrives at 9 AM
         → Waits until next batch (Monday)
         → 6 days before enrichment
         → Rep works with incomplete data all week

Event-Triggered:
─────────────────────────────────────────
Tuesday: New lead arrives at 9 AM
         → Enrichment triggered immediately
         → Data available in 30 seconds
         → Rep has full context instantly

Benefits

Faster speed-to-lead — Enriched data available for immediate routing
Better lead scoring — Score based on complete firmographics
Personalized response — Reps see context before first touch
Reduced data decay — Enrich when data is most current

Common Trigger Events

Lead Creation

The most common trigger:

Event: Lead created
Source: Form submission, list import, API creation

Enrichment:
- Contact info (email verification, phone)
- Company data (size, industry, tech stack)
- Social profiles (LinkedIn, Twitter)

Use Case:
- Lead routing
- Lead scoring
- Initial personalization

Contact Association

When a contact is linked to an account:

Event: Contact associated to Account
Source: Manual association, sync, automation

Enrichment:
- Account-level data (if not already enriched)
- Additional contacts at account
- Relationship mapping

Use Case:
- Account-based selling
- Multi-threaded outreach
- Org chart building

Opportunity Stage Change

When a deal progresses:

Event: Opportunity moves to "Negotiation"
Source: Rep updates stage

Enrichment:
- Executive contacts (decision makers)
- Financial data (for deal sizing)
- Competitor intelligence
- Recent news (negotiation leverage)

Use Case:
- Executive outreach
- Deal strategy
- Objection handling

Website Visit (Known Visitor)

When an identified visitor returns:

Event: Known contact visits pricing page
Source: Website analytics + CRM match

Enrichment:
- Refresh contact data (still at company?)
- Company growth indicators
- Intent signals

Use Case:
- Re-engagement timing
- Updated context for outreach

Intent Signal Detection

When buying signals appear:

Event: Account shows high intent on relevant topic
Source: Bombora, G2, LinkedIn, etc.

Enrichment:
- Find contacts at account
- Enrich contact details
- Research company context

Use Case:
- Proactive outreach
- ABM targeting
- Timely engagement

Data Quality Trigger

When data quality drops:

Event: Email bounces or phone disconnects
Source: Email platform, dialer

Enrichment:
- Find new email/phone
- Check if person changed jobs
- Update company association

Use Case:
- Data maintenance
- Job change detection
- Prevent further bounces

Architecture Patterns

Pattern 1: Direct Integration

Simplest approach—CRM triggers enrichment directly:

┌──────────┐     Webhook      ┌───────────────┐
│  CRM     │ ───────────────→ │  Enrichment   │
│          │                  │  Service      │
│          │ ←─────────────── │  (Clay, etc.) │
└──────────┘     API Push     └───────────────┘

Pros:

Simple to implement
Low latency
Easy to debug

Cons:

Tight coupling
Limited transformation
CRM-specific logic

Pattern 2: Event Bus

More scalable—events flow through a message queue:

┌──────────┐     Event      ┌───────────────┐     Event     ┌───────────────┐
│  CRM     │ ─────────────→ │  Event Bus    │ ────────────→ │  Enrichment   │
│          │                │  (Kafka, SQS) │               │  Service      │
│          │ ←───────────── │               │ ←──────────── │               │
└──────────┘    Update      └───────────────┘    Result     └───────────────┘

Pros:

Decoupled systems
Reliable delivery
Easy to add consumers

Cons:

More complex setup
Slight latency increase
Additional infrastructure

Pattern 3: Orchestration Layer

Complex workflows through dedicated orchestrator:

┌──────────┐     ┌────────────────────────────────────────────┐
│  CRM     │     │              Orchestrator                  │
│          │ ──→ │  ┌─────────┐  ┌─────────┐  ┌─────────┐   │
│          │     │  │Enrich 1 │→ │Enrich 2 │→ │Enrich 3 │   │
│          │ ←── │  └─────────┘  └─────────┘  └─────────┘   │
└──────────┘     └────────────────────────────────────────────┘
                              (Clay, n8n, Temporal)

Pros:

Complex workflow support
Retry and error handling
Visibility and monitoring

Cons:

Most complex setup
Requires orchestration tool
Higher cost

Salesforce Implementation

Flow-Based Triggers

Flow: Event-Triggered Lead Enrichment
Type: Record-Triggered Flow
Object: Lead
Trigger: After Insert

Entry Conditions:
- Email is not blank
- Enrichment_Status__c is blank or null

Steps:

1. Update Records (immediate)
   Enrichment_Status__c = "Pending"

2. Action: HTTP Callout
   Endpoint: [Clay webhook URL]
   Method: POST
   Headers: Content-Type: application/json
   Body:
   {
     "lead_id": "{!$Record.Id}",
     "email": "{!$Record.Email}",
     "company": "{!$Record.Company}",
     "event_type": "lead_created",
     "timestamp": "{!$Flow.CurrentDateTime}"
   }

Platform Event Approach

For more complex scenarios, use Platform Events:

// Publish event when Lead is created
trigger LeadTrigger on Lead (after insert) {
    List<Enrichment_Request__e> events = new List<Enrichment_Request__e>();

    for (Lead l : Trigger.new) {
        if (String.isNotBlank(l.Email)) {
            events.add(new Enrichment_Request__e(
                Record_Id__c = l.Id,
                Object_Type__c = 'Lead',
                Email__c = l.Email,
                Company__c = l.Company,
                Event_Type__c = 'created'
            ));
        }
    }

    if (!events.isEmpty()) {
        EventBus.publish(events);
    }
}

Subscribe externally via CometD or Pub/Sub API.

Change Data Capture

Use Salesforce CDC for external systems:

1. Enable CDC on Lead object
2. Subscribe to /data/LeadChangeEvent
3. External service receives all creates/updates
4. Filter for enrichment-worthy events
5. Trigger enrichment

HubSpot Implementation

Workflow-Based Triggers

Workflow: Event-Triggered Enrichment
Type: Contact-based

Enrollment Trigger:
- Contact is created
- AND Email is known
- AND Enrichment Status is not "Complete"

Actions:

1. Set property value
   Enrichment Status = "Pending"

2. Trigger webhook
   URL: [Clay webhook URL]
   Method: POST
   Body:
   {
     "contact_id": "{{contact.vid}}",
     "email": "{{contact.email}}",
     "company": "{{contact.company}}",
     "event_type": "contact_created"
   }

3. Delay: 60 seconds (wait for enrichment)

4. Branch: If Enrichment Status = "Complete"
   Yes → Continue workflow
   No → Send internal notification

Operations Hub Custom Code

For complex logic:

// Custom code action in Operations Hub
exports.main = async (event, callback) => {
  const axios = require('axios');

  const contactEmail = event.inputFields['email'];
  const contactId = event.inputFields['hs_object_id'];

  // Only enrich business emails
  const personalDomains = ['gmail.com', 'yahoo.com', 'hotmail.com'];
  const domain = contactEmail.split('@')[1];

  if (personalDomains.includes(domain)) {
    callback({
      outputFields: {
        should_enrich: 'false',
        reason: 'personal_email'
      }
    });
    return;
  }

  // Trigger enrichment
  try {
    await axios.post('https://your-enrichment-endpoint.com/enrich', {
      email: contactEmail,
      contact_id: contactId
    });

    callback({
      outputFields: {
        should_enrich: 'true',
        enrichment_triggered: 'true'
      }
    });
  } catch (error) {
    callback({
      outputFields: {
        should_enrich: 'true',
        enrichment_triggered: 'false',
        error: error.message
      }
    });
  }
};

Enrichment Service Design

Clay Webhook Handler

Clay can receive webhooks and trigger enrichments:

Clay Table Setup:

1. Create incoming webhook
   URL: https://api.clay.com/webhooks/[table-id]

2. Configure input columns:
   - email (from webhook payload)
   - company_domain (from webhook payload)
   - source_id (Salesforce/HubSpot ID)
   - event_type

3. Add enrichment columns:
   - [Your waterfall setup]

4. Add push action:
   - Push to Salesforce/HubSpot
   - Include source_id for matching

Custom Enrichment Service

If building your own:

// Express.js enrichment service
const express = require('express');
const app = express();

app.post('/enrich', async (req, res) => {
  const { email, company, source_id, source_system } = req.body;

  // 1. Check cache (avoid re-enriching recent data)
  const cached = await cache.get(email);
  if (cached && cached.age < 24 * 60 * 60 * 1000) {
    await pushToCRM(source_system, source_id, cached.data);
    return res.json({ status: 'cached', data: cached.data });
  }

  // 2. Waterfall enrichment
  let enrichedData = {};

  // Try Apollo first
  const apolloData = await enrichWithApollo(email);
  if (apolloData) {
    enrichedData = { ...enrichedData, ...apolloData };
  }

  // Fill gaps with Clearbit
  if (!enrichedData.title || !enrichedData.phone) {
    const clearbitData = await enrichWithClearbit(email);
    enrichedData = { ...enrichedData, ...clearbitData };
  }

  // Verify email
  enrichedData.email_valid = await verifyEmail(email);

  // 3. Cache result
  await cache.set(email, { data: enrichedData, timestamp: Date.now() });

  // 4. Push back to CRM
  await pushToCRM(source_system, source_id, enrichedData);

  res.json({ status: 'enriched', data: enrichedData });
});

async function pushToCRM(system, id, data) {
  if (system === 'salesforce') {
    await salesforce.update('Lead', id, mapToSalesforceFields(data));
  } else if (system === 'hubspot') {
    await hubspot.updateContact(id, mapToHubSpotProperties(data));
  }
}

Timing Considerations

Latency Requirements

Lead Routing Scenario:
─────────────────────────────────────────
Requirement: Route lead to correct rep
Enrichment needed: Company size, industry
Acceptable latency: <5 seconds

Why: Round-robin vs enterprise rep routing
     must happen before notification

Solution:
- Use fastest enrichment provider
- Cache company data
- Parallel enrichment calls

Sales Outreach Scenario:
─────────────────────────────────────────
Requirement: Enrich before rep calls
Enrichment needed: Full profile, research
Acceptable latency: <60 seconds

Why: Rep won't call for several minutes
     anyway

Solution:
- Full waterfall enrichment
- Include AI-generated research
- No need to optimize for speed

Async vs Sync

Synchronous (wait for result):

Good for:
- Lead routing that depends on enrichment
- Real-time lead scoring
- Immediate personalization

Challenge:
- User waits for response
- Timeout risk
- Must be fast (<5s)

Asynchronous (fire and forget):

Good for:
- Background data enhancement
- Deep research
- Non-blocking workflows

Challenge:
- Need callback mechanism
- More complex error handling
- State management

Recommended Approach

Event Type              | Sync/Async | Latency Target
─────────────────────────────────────────────────────
Lead creation (routing) | Sync       | <3 seconds
Lead creation (full)    | Async      | <60 seconds
Opportunity advance     | Async      | <2 minutes
Intent signal           | Async      | <5 minutes
Data quality trigger    | Async      | <1 hour

Error Handling

Retry Logic

async function enrichWithRetry(email, maxRetries = 3) {
  for (let attempt = 1; attempt <= maxRetries; attempt++) {
    try {
      const result = await enrich(email);
      return result;
    } catch (error) {
      if (attempt === maxRetries) {
        // Final failure - log and alert
        await logFailure(email, error);
        await alertOps(email, error);
        throw error;
      }

      // Exponential backoff
      const delay = Math.pow(2, attempt) * 1000;
      await sleep(delay);
    }
  }
}

Dead Letter Queue

For failed enrichments:

Workflow:
1. Enrichment fails after all retries
2. Record added to "Failed Enrichment" queue
3. Daily job processes queue:
   - Retry with different provider
   - Manual review if still fails
4. Alert if queue grows too large

Partial Success Handling

Scenario: Apollo returns title, but no phone

Options:
1. Partial update (update what we got)
2. Wait and retry (try another provider)
3. Queue for follow-up (get phone later)

Recommendation:
- Always do partial update
- Mark which fields still need data
- Background job fills gaps later

Monitoring

Key Metrics

Event-Triggered Enrichment Dashboard:

Volume:
- Events received: X/hour
- Enrichments triggered: Y/hour
- Success rate: Z%

Latency:
- P50 latency: Xms
- P95 latency: Yms
- P99 latency: Zms

Quality:
- Fields populated: X%
- Enrichment coverage: Y%
- Error rate: Z%

Alerting

Alert Conditions:

1. High error rate
   If error_rate > 5% for 15 minutes
   → Page on-call

2. High latency
   If P95_latency > 10 seconds
   → Slack notification

3. Queue backup
   If dead_letter_queue_size > 100
   → Slack notification

4. Provider down
   If provider_success_rate < 90%
   → Switch to backup provider

Clay + CRM Integration Patterns — Integration architectures
Waterfall Enrichment — Multi-source enrichment
Modern Outbound Data Stack — System architecture
Salesforce Data Quality Scorecard — Measuring impact
The B2B Data Decay Problem — Why real-time matters

Why Event-Triggered?

Batch vs Event-Triggered

Benefits

Common Trigger Events

Lead Creation

Contact Association

Opportunity Stage Change

Website Visit (Known Visitor)

Intent Signal Detection

Data Quality Trigger

Architecture Patterns

Pattern 1: Direct Integration

Pattern 2: Event Bus

Pattern 3: Orchestration Layer

Salesforce Implementation

Flow-Based Triggers

Platform Event Approach

Change Data Capture

HubSpot Implementation

Workflow-Based Triggers

Operations Hub Custom Code

Enrichment Service Design

Clay Webhook Handler

Custom Enrichment Service

Timing Considerations

Latency Requirements

Async vs Sync

Recommended Approach

Error Handling

Retry Logic

Dead Letter Queue

Partial Success Handling

Monitoring

Key Metrics

Alerting

Related Guides