Lead Enrichment Actor

Business contact enrichment: discovery → extraction → verification → CRM-ready output

Turn a list of business URLs or a natural search query into structured contact records. The Lead Enrichment Actor discovers candidate business pages, extracts contact details from each page, optionally verifies emails via DNS, filters low-confidence rows, and emits clean records with email, emailStatus, phone, social profiles, and company metadata.

What it does

• Discovery — use explicit startUrls, or run DuckDuckGo-backed search discovery from a searchQuery
• Direct enrichment — fetch candidate pages via HTTP and extract company name, email, phone, address, and social links
• Subpage expansion — follow high-value internal subpaths such as /contact, /about, /team, /locations, plus data-attribute/meta/JSON-LD contact signals
• Homepage-first ranking — prefer business homepages and high-value subpages over directory/listing pages before extraction
• Directory handling — detect directory/listing pages and extract individual business candidates from them, bounded by a per-directory-domain cap
• Validation — drop non-business rows, blacklisted domains, and low-confidence records before emitting
• Telemetry — emit discovery, filtering, and contact-quality stats for every run
• Output — raw dataset rows, plus optional CSV / HubSpot / Salesforce export files

How it works

Input
 ├─ startUrls provided? ── Yes ──► Normalize URLs
 │                               │
 └─ No                           ▼
     searchQuery ─► DuckDuckGo discovery
                         │
                         ▼
            Filter DDG redirectors / blacklisted domains
                         │
                         ▼
              Homepage-first candidate ranking
              ├─ Boost root /contact /about /team /locations
              ├─ Deprioritize /search /find /listings /directory
              └─ Collapse to best-scoring URL per domain
                         │
                         ▼
              Direct enrichment path
              ├─ Fetch page via HTTP
              ├─ Extract company name, website, domain, address
              ├─ Follow high-value subpages up to depth limit
              ├─ Merge main + subpage HTML
              ├─ Extract email, phone, social links
              └─ Populate emailStatus and completenessScore
                         │
                         ▼
                    Confidence gate
                    ├─ Low confidence / aggregator / social / no contact?
                    │     └─► Drop row
                    └─ Pass
                            │
                            ▼
                        Emit row
                            │
                            ▼
                    outputFormat csv/hubspot/salesforce?
                    ├─ Yes ──► Write mapped CSV
                    └─ No ──► Finish

Directory/listing pages:
  ├─ Detect directory/listing page
  ├─ Extract business cards inline from listing rows
  ├─ Enrich cards with deeper contact/social extraction
  ├─ Keep only qualified business names
  └─ Emit candidate records directly

Aggregator/Directory domain handling:
  ├─ Hard-blacklist known social and directory hosts
  ├─ Per-directory-domain cap prevents one host from dominating
  ├─ Same-domain locality subpages allowed up to the cap
  └─ Low-confidence aggregator rows are dropped

Usage examples

Example 1 — enrich a single business website

{
  "startUrls": ["https://www.retaildogma.com"],
  "maxRequestsPerCrawl": 20
}

Example 2 — batch append to CRM

{
  "startUrls": [
    "https://www.retaildogma.com",
    "https://www.prisync.com",
    "https://www.shopify.com"
  ],
  "maxRequestsPerCrawl": 50,
  "outputFormat": "hubspot"
}

Example 3 — search-to-lead with email verification

{
  "searchQuery": "dentists in Sydney",
  "maxRequestsPerCrawl": 30,
  "verifyEmails": true,
  "outputFormat": "salesforce",
  "proxyConfiguration": {
    "useApifyProxy": true,
    "groups": ["RESIDENTIAL"]
  }
}

Example 4 — raw output with deduplication across runs

{
  "searchQuery": "plumbers in Austin TX",
  "maxRequestsPerCrawl": 25,
  "dedupeAcrossRuns": true,
  "outputFormat": "raw"
}

Example 5 — search-to-lead with proxy

{
  "searchQuery": "dentists in Sydney",
  "maxRequestsPerCrawl": 20,
  "verifyEmails": true,
  "outputFormat": "salesforce",
  "proxyConfiguration": {
    "useApifyProxy": true,
    "groups": ["RESIDENTIAL"]
  }
}

Commands

Run via Apify CLI

# Run with a search query
apify call ev2BXp2HWlPCmQFWg \
  -i '{"searchQuery":"dentists in Sydney","maxRequestsPerCrawl":20}'

# Run with explicit start URLs
apify call ev2BXp2HWlPCmQFWg \
  -i '{"startUrls":["https://www.retaildogma.com"],"maxRequestsPerCrawl":20}'

# Run with email verification and Salesforce export
apify call ev2BXp2HWlPCmQFWg \
  -i '{"searchQuery":"dentists in Sydney","verifyEmails":true,"outputFormat":"salesforce"}'

Run via Python

from apify_client import ApifyClient

client = ApifyClient("<YOUR_API_TOKEN>")

run = client.actor("kind_cordial/lead-enrichment-lite").call(
    run_input={
        "searchQuery": "dentists in Sydney",
        "maxRequestsPerCrawl": 20,
        "verifyEmails": False,
        "outputFormat": "raw",
    }
)

dataset_id = run["defaultDatasetId"]
for item in client.dataset(dataset_id).iterate_items():
    print(item)

Build and run locally

# Install dependencies
pip install -r requirements.txt

# Run actor locally
python -m src

# Or with explicit input JSON
echo '{"searchQuery":"dentists in Sydney","maxRequestsPerCrawl":20}' | python -m src

Output fields

Field	Description
sourceUrl	Discovered or submitted URL
companyName	Detected business or brand name
website	Canonical website URL
domain	Bare domain
email	Extracted email address(es)
emailStatus	found / not_found / DNS verification result
phone	Extracted phone number(s)
address	Postal address snippet when available
linkedin	LinkedIn profile/company URL(s)
facebook	Facebook page URL(s)
instagram	Instagram profile URL(s)
twitter	X / Twitter URL(s)
completenessScore	Record completeness score

When outputFormat is csv, hubspot, or salesforce, an additional field-mapped CSV is saved to the key-value store as contacts.csv, contacts_hubspot.csv, or contacts_salesforce.csv.

Behavior summary

• Default mode is fast HTTP extraction.
• Search discovery ranks business homepages and high-value subpages ahead of directory/listing pages, then collapses to the best-scoring URL per domain.
• Directory/listing pages are used as sources to extract candidate individual business records; they are not emitted as final rows.
• Known aggregator and social domains are blacklisted from direct enrichment.
• A per-directory-domain cap prevents any single listing host from dominating the run.
• Failed fetches, low-confidence rows, and blacklisted domains are filtered before emission; output count reflects processed qualified businesses, not raw input count.