AI Lead Enrichment — Firmographics, ICP Scoring & Outreach

Turn company, domain, and contact lists into sales-ready lead enrichment data. This AI lead enrichment actor uses an LLM with live web search to research firmographics, infer role and seniority, score ideal-customer-profile fit, cite sources, and draft a personalized outreach opener.

Export results, run via API, schedule and monitor runs, or integrate with your CRM and outbound tools.

---

What you get per lead

Field	Example
company_name, domain	Apify, apify.com
industry, company_size, hq_location, founded_year	Web scraping & automation, 51-200, Prague, 2015
company_description	Full-stack web scraping and automation platform.
person_name, role_title, seniority	Jane Doe, VP of Sales, VP
summary	Mid-size scraping platform; strong fit for proxy cost savings.
icp_fit_score (0-100) + icp_fit_reasons	85 — matches industry, headcount in range
personalized_opener	"Saw Apify scaled its proxy fleet—we cut similar costs 40% for data teams."
citations, confidence	source URLs, high/medium/low

The actor preserves the original lead context and returns structured fields for CRM imports, lead routing, sales research, and agent workflows. Failed enrichments remain visible with an error value so you can retry them selectively.

---

Input

Field	Type	Description
leads	array	Objects with any of company, domain, fullName, role, location, notes.
icpDescription	string	Your ideal customer profile — used to score each lead's fit.
offerContext	string	What you're selling — used to write the personalized opener.
model	string	OpenRouter model slug (default openai/gpt-4o-mini).
useWebSearch	boolean	Ground enrichment in live web results (recommended).
webSearchResults	integer	Web results per lead (1–10).
maxLeads	integer	Cap on leads per run (cost control).

Example input

{
  "leads": [
    { "company": "Apify", "domain": "apify.com" },
    { "fullName": "Jane Doe", "company": "Acme", "role": "VP Sales" }
  ],
  "icpDescription": "B2B SaaS, 50-1000 employees, data/AI teams",
  "offerContext": "Managed proxy + data pipeline that cuts scraping costs 40%",
  "model": "openai/gpt-4o-mini",
  "useWebSearch": true
}

---

API key (required)

Enrichment runs through OpenRouter — set a single environment variable on the actor (Console → Settings → Environment variables, mark as Secret):

OPENROUTER_API_KEY = sk-or-...

Pick any model via model. You pay OpenRouter directly for model + web-search usage; the actor's PPE events cover the enrichment layer.

---

Pricing (Pay Per Event)

Event	When
actor-start	Once per run
lead-enriched	Each lead successfully enriched

Leads that fail (missing key, model error) are not charged — but still appear in the dataset with an error so you know which ones to retry.

The number of chargeable enrichment events is capped by the smaller of your input list and maxLeads. OpenRouter model and web-search charges are billed separately by OpenRouter, so choose the model and search depth that fit your accuracy and cost requirements.

Run AI lead enrichment with the API

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('USERNAME/ai-lead-enrichment').call({
  leads: [{ company: 'Apify', domain: 'apify.com' }],
  icpDescription: 'B2B SaaS companies with 50-1,000 employees',
  offerContext: 'Data automation platform for revenue teams',
  useWebSearch: true,
  maxLeads: 100,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Connect the output to a CRM, spreadsheet, webhook, data warehouse, or AI agent through Apify's standard integrations. For agents, the actor provides one semantic record per submitted lead.

---

Use cases

• Outbound sales — enrich a prospect list and get tailored openers in one run.
• CRM hygiene — fill missing firmographics on existing records.
• Lead scoring / routing — rank inbound leads by ICP fit before they hit a rep.
• Agent pipelines — call via the Apify MCP server to enrich leads on demand.

Who should use this lead enrichment actor?

• SDR and outbound teams researching accounts before contact
• RevOps teams standardizing incomplete CRM records
• Growth teams prioritizing lists by ICP fit
• Agencies preparing personalized prospecting campaigns
• AI agents that need structured company context before taking the next step

Do not use it as an email verifier, phone-number finder, identity-resolution service, or authoritative company registry. It enriches the information supplied and may infer fields from public web context; it does not promise verified contact details.

---

Tips

• More input context (domain + role + notes) → better, higher-confidence enrichment.
• Write a specific icpDescription — vague profiles produce vague scores.
• For large lists, raise maxLeads and schedule runs; keep webSearchResults low to control cost.

Data quality and limitations

LLM-generated firmographics and scoring are probabilistic. Review citations, confidence, and error before importing records into production systems. A domain usually produces more reliable company matching than a name alone. ICP scores are relative to your written icpDescription; they are not universal lead-quality scores.

For consistent comparisons, reuse the same ICP description and model across a batch. Human-review high-value leads before outreach, particularly when the source evidence is weak or conflicting.

FAQ

What information is required for each lead?

Provide at least a company, domain, full name, role, location, or useful notes. Company plus domain is the strongest starting point for firmographic enrichment.

Does it find verified emails or phone numbers?

No. This actor focuses on research, firmographics, role context, ICP scoring, citations, and personalized outreach copy.

Can I use the results in a CRM?

Yes. Export the dataset or retrieve it through the API, then map the stable output fields to your CRM. Review inferred values before overwriting trusted records.

Responsible sales use

Use the actor for legitimate B2B research and comply with privacy, marketing, anti-spam, and data-protection laws that apply to your region. Do not use generated personalization to misrepresent a relationship or fabricate facts. Keep the OpenRouter key in a secret environment variable.