Live Actor and maintained API: Run B2B Lead Cleaner on Apify

B2B Lead Cleaner: Domain-Level Email and MX Validation Samples

Real public Task inputs, PII-safe run outputs, and a strict JSON Schema for routing B2B lead rows into ACCEPT, REVIEW, or REJECT.

The Actor checks email syntax, provider type, DNS mail routing, company-domain consistency, duplicates, freshness, and optional ICP rules. It keeps one explained output decision per delivered source row.

Measured public examples

These results came from successful public Task runs on 2026-07-28 using Actor build 0.0.29. The dataset contract was checked against the current 0.0.30 Actor source.

Public Task	Actual run result	Input
Audit Purchased B2B Leads Before Vendor Renewal	7 delivered: 5 ACCEPT, 1 REVIEW, 1 REJECT. A supplied $70 audited-scope cost produced a reported $14 per accepted lead.	JSON
Clean a B2B Lead List Before Cold Outreach	7 delivered: 5 ACCEPT, 1 role-inbox REVIEW, 1 duplicate REJECT.	JSON
Prepare Complete B2B Leads for CRM Import	5 delivered: 5 ACCEPT with score 100 in this synthetic sample.	JSON

Why use a decision pipeline?

Need	Raw-list handling	This Actor
Routing	Manual judgment or one opaque score	Explained ACCEPT, REVIEW, or REJECT
Email evidence	Often just a non-empty string	Syntax, provider class, and DNS MX evidence
Duplicates	Rows may be silently removed	Every delivered row remains, with duplicate flags and reason codes
Source formats	Fixed column names	Common aliases, nested paths, and manual field mapping
Targeting	Separate spreadsheet filters	Optional country, city, title, industry, and account rules
Run audit	Row output only	Dataset decisions plus an OUTPUT run report

Decision pipeline

flowchart LR
    A["Dataset, email strings, or lead objects"] --> B["Field detection and normalization"]
    B --> C["Syntax, provider, and DNS MX checks"]
    C --> D["Duplicate, freshness, domain, and ICP rules"]
    D --> E{"Final decision"}
    E -->|Pass| F["ACCEPT"]
    E -->|Uncertain| G["REVIEW"]
    E -->|Hard rule failed| H["REJECT"]
    F --> I["Decision dataset and OUTPUT report"]
    G --> I
    H --> I

Loading

Public Task inputs

01. Audit purchased leads before vendor renewal

{
  "maxItems": 7,
  "qualityMode": "BALANCED",
  "rows": [
    {
      "full_name": "Maya Chen",
      "job_title": "VP Growth",
      "email": "maya.chen@apify.com",
      "company_name": "Apify",
      "company_domain": "apify.com",
      "country": "Czechia",
      "city": "Prague",
      "industry": "Software",
      "sample_record": true
    }
  ],
  "rules": {
    "requireEmail": false,
    "sourceCostUsd": 70
  },
  "targeting": {
    "countries": ["Czechia", "United States", "Germany"],
    "industries": ["Software"],
    "titleKeywords": [
      "VP",
      "Marketing Director",
      "Revenue Operations",
      "Demand Generation",
      "Sales Operations"
    ]
  }
}

The complete seven-row input is in 01_audit_purchased_b2b_leads_before_vendor_renewal_input.json.

02. Clean a messy B2B list before outreach

{
  "maxItems": 7,
  "qualityMode": "BALANCED",
  "rows": [
    {
      "full_name": "Maya Chen",
      "job_title": "VP Marketing",
      "work_email": "maya.chen@hubspot.com",
      "company_website": "https://www.hubspot.com",
      "sample_record": true
    }
  ],
  "rules": {
    "acceptScore": 90
  }
}

The complete mixed-schema input is in 02_clean_a_b2b_lead_list_before_cold_outreach_input.json.

03. Prepare complete leads for CRM import

{
  "maxItems": 5,
  "qualityMode": "BALANCED",
  "rows": [
    {
      "full_name": "Sofia Bennett (sample)",
      "job_title": "VP Sales",
      "email": "apify-task-demo-401@hubspot.com",
      "company_name": "HubSpot",
      "company_domain": "hubspot.com"
    }
  ]
}

The complete five-row input is in 03_prepare_complete_b2b_leads_for_crm_import_input.json.

Real output excerpts

These are schema-valid records from successful public Task runs. The duplicate example has identity values redacted while preserving its decision and evidence.

ACCEPT: complete business-domain record

{
  "decision": "ACCEPT",
  "qualityScore": 100,
  "reasonCodes": [],
  "emailStatus": "BUSINESS",
  "mxStatus": "VALID",
  "duplicate": false,
  "evidence": {
    "email": {
      "syntaxValid": true,
      "status": "BUSINESS",
      "mailboxExistence": "NOT_CLAIMED"
    }
  }
}

Open the complete output record

REVIEW: valid domain routing, but role inbox

{
  "decision": "REVIEW",
  "qualityScore": 85,
  "primaryReason": "Quality score is below the ACCEPT threshold.",
  "reasonCodes": [
    "QUALITY_SCORE_NEEDS_REVIEW",
    "EMAIL_ROLE_BASED"
  ],
  "emailStatus": "ROLE_BASED",
  "mxStatus": "VALID"
}

Open the complete output record

REJECT: duplicate stays visible and explained

{
  "decision": "REJECT",
  "qualityScore": 30,
  "primaryReason": "The same lead already appeared in this batch.",
  "reasonCodes": ["DUPLICATE_LEAD"],
  "duplicate": true,
  "historicalDuplicate": false
}

Open the PII-redacted output record

API quick start

curl -X POST \
  "https://api.apify.com/v2/acts/kamerozkan~b2b-lead-cleaner/runs" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "emails": [
      "person@company.example",
      "team@company.example"
    ],
    "qualityMode": "BALANCED"
  }'

After the run finishes, read decision rows from the default dataset and the run report from the default key-value store record OUTPUT.

Data contract

The complete consumer contract is in dataset_record.schema.json. Important fields include:

• decision, qualityScore, primaryReason, and reasonCodes
• normalized contact and company fields
• emailStatus, mxStatus, and optional websiteStatus
• duplicate, historicalDuplicate, and dedupeKey
• typed evidence and ordered reasons
• raw-row preservation fields and checkedAt

qualityScore is a deterministic rule-based quality score. It is not an email deliverability percentage.

Truth boundary

• VALID MX means the domain had usable public mail-routing evidence at check time.
• No SMTP mailbox probing is performed.
• The Actor does not prove that a specific mailbox exists or belongs to a person.
• Phone and LinkedIn values are normalized and preserved, not verified.
• The Actor does not guarantee current employment, consent, replies, or legal permission to contact someone.
• Process only data you are authorized to use and follow applicable privacy and marketing laws.

See DATA_NOTICE.md for sample provenance and privacy notes.

Links

• Live Actor
• Apify API endpoint
• Kamer Ozkan on Apify

License

The repository files are available under the MIT License.