CRM Data Enrichment Agent – Contact & Lead Scoring

What does CRM Data Enrichment Agent do?

CRM Data Enrichment Agent enriches CRM contacts with publicly available data and optional third-party APIs. Pass a list of contacts (JSON, CSV, or dataset), and the Actor normalizes fields, validates emails, deduplicates contacts, enriches company data, scores lead quality, and can push results into your automation stack.

What can this Actor do?

• Multiple input options: JSON array, CSV text, or Apify dataset ID.
• Website enrichment: homepage metadata, emails, social links, phone/address, and basic role hints.
• Optional API providers: Clearbit and Hunter.
• Smart deduplication with configurable keys.
• Quality scoring with configurable weights.
• Workflow integrations: n8n, OpenClaw, and generic webhooks.
• Intelligence layer: adaptive enrichment, domain cache, AI fallback, progressive pipeline, learning memory, and budget guardrails.

What data can CRM Data Enrichment Agent extract?

Data point	Description
Normalized email/domain	Cleaned contact identity signals
Website metadata	URL, title, meta description
Contact signals	Found emails, phone, social links
Optional API enrichment	Clearbit and Hunter outputs
Scoring	0-100 score with per-signal breakdown
Deduplication	Duplicate status + duplicateOf mapping
Confidence/explainability	Confidence score + reason traces (when enabled)

How do I use it?

1. Open the Actor and choose input source (json, csv, or dataset).
2. Provide contacts and field mapping.
3. Enable website enrichment (default), optional providers, and scoring.
4. Optionally configure integrations and intelligence settings.
5. Run the Actor and review Output/Dataset tabs.

Input

Main input groups:

• contactsSource, contactsJson, contactsCsv, datasetId
• Field mapping: emailField, firstNameField, lastNameField, companyField
• enrichment: website fallback, Clearbit, Hunter
• crawling: concurrency/rate/timeouts/retries/proxy options
• scoring: enabled + weights
• integrations: n8n/OpenClaw/webhooks + route webhooks
• intelligence: adaptive/cache/AI/progressive/learning/budget
• output: partial push cadence and debug artifacts

Output

You can download the dataset in JSON/CSV/Excel from the run Dataset tab. The Output tab links to:

• default dataset items
• key-value store OUTPUT (summary + full results)

Each dataset item is one enriched contact:

{
  "contactId": "abc123",
  "input": { "email": "john@example.com", "firstName": "John", "company": "Example Inc" },
  "normalized": {
    "email": "john@example.com",
    "emailValid": true,
    "domain": "example.com",
    "companyNameNormalized": "Example Inc",
    "firstName": "John",
    "lastName": null
  },
  "enrichment": {
    "website": {
      "url": "https://example.com",
      "title": "Example Inc - Home",
      "foundEmails": ["info@example.com"],
      "socialLinks": { "linkedin": "https://linkedin.com/company/example" }
    },
    "clearbit": null,
    "hunter": null
  },
  "metrics": { "enrichmentSourcesUsed": ["website"], "fetchTimesMs": {}, "errors": [] },
  "dedupe": { "isDuplicate": false, "duplicateOf": null, "dedupeHash": "..." },
  "score": { "value": 75, "breakdown": { "emailValid": 25, "hasWebsite": 20, "hasPhone": 0, "hasSocial": 15, "companySignals": 15 } }
}

New intelligence and efficiency features

1. Adaptive enrichment policy

• Skips paid providers when confidence is already high.

2. Domain-level cache across runs

• Reuses website/Clearbit/Hunter enrichment by root domain.

3. AI fallback

• Modes: off, heuristic, webhook.

4. Progressive pipeline

• Restricts paid API usage to top confidence band.

5. Confidence + explainability

• Adds metrics.confidenceScore, confidenceBreakdown, and explainability traces.

6. Domain learning memory

• Tracks success/failure history for allowlist/blocklist behavior.

7. Role/title inference signal

• AI fallback can infer likely job title when extraction is weak.

8. Routing rules

• Dedicated webhooks for high-score and failed contacts.

9. Smart retry strategy

• Exponential backoff + jitter for integration delivery.

10. Budget guardrails

• Max cap on external paid API calls per run.

Use cases

1. Outbound lead prioritization

• Rank contacts before SDR assignment by score + confidence.

2. CRM data refresh

• Periodic enrichment to refresh website/social/phone signals.

3. n8n/OpenClaw orchestration

• Trigger downstream workflows directly from integration payloads.

4. Cost-controlled enrichment at scale

• Use adaptive + progressive + budget caps to control spend.

5. Lead routing automation

• Send high-score contacts to sales and failed contacts to research queue.

Marketing and intent signals

Use these as GTM signals:

• Contactability: valid email, phone, social presence.
• Company maturity: richer metadata, address presence, about/contact pages.
• Quality/intent proxy: stronger company signals and lower failure patterns.
• Reliability: confidence score, explainability notes, domain-learning state.

Suggested workflow:

• Route score >= threshold and strong confidence to outbound sequence.
• Route failed/partial leads to nurture or verification workflows.

Tips and advanced options

• Start with website-only enrichment to keep costs low.
• Keep concurrency/rate conservative for stable crawling.
• Enable caching and learning for recurring runs on similar domains.
• Enable adaptive/progressive mode before turning on paid APIs at scale.
• Use route webhooks for high-score and failed queues.

Compliance and support

Use responsibly and in compliance with GDPR/CCPA, third-party API terms, and target-site policies.

If this Actor helps your workflow, please leave a 5-star rating on the Actor page.