Company Enrichment API — Turn a Domain Into Firmographics, Tech Stack & Verified Emails

What it does: give it a company domain (e.g. stripe.com) and it returns a full structured profile — industry + NAICS/SIC, employee band, revenue band, categorized tech stack, LinkedIn/Crunchbase, role emails, and decision-makers — read live from the company's own website. Who it's for: SDRs, RevOps, founders, and data teams who need fresh B2B company data without buying a Clearbit, Apollo, Clay, or ZoomInfo seat. What it costs: $0.02 per enriched company ($20 per 1,000) plus $0.01 per run. No subscription, no API key, no seats — and records below your confidence threshold are not charged.

What company enrichment returns here

Company enrichment means taking an identifier you already have — a domain — and filling in everything else about that company. One domain in, one structured record out, grouped like this:

Group	Fields you get back
Identity	companyName, description, logo, foundedYear, companyType, country, address
Firmographics	industry, naicsCode, sicCode, employeesEstimate (band + source), revenueBand
Technographics	technographics (categorized: Frontend, Analytics, Payments, CRM, Hosting…) and the flat techStack
Contacts	emails[] with type (role vs personal), emailPattern, phones, decisionMakers[] with name, title and email
Deliverability	domainHasMx, mxProvider, per-email confidence, emailInferred
Social	linkedinUrl, crunchbaseUrl, plus socialLinks (Twitter/X, Facebook, Instagram, YouTube)
Quality	confidenceScore (0–1) on the record, a confidence on every email and decision-maker, and pagesCrawled

Anything not published on the site comes back null — never a guess. See the full example record below.

🎥 Video walkthrough

Why live enrichment beats a static database

Database vendors sell you records they collected months ago; B2B contact data decays at roughly 2% per month as people change jobs and companies pivot. This Actor doesn't have a database to go stale — every run crawls the company's live site (homepage, about, team, leadership, contact pages) and extracts what is published right now, with:

• a 0–1 confidence score on every email, every decision-maker, and the record overall;
• a live DNS MX check on the domain (domainHasMx, mxProvider — e.g. "Google Workspace") so you know the mailbox infrastructure exists before you send;
• an honest emailInferred flag — a pattern email (first.last@) is only constructed when two or more real emails confirm the pattern, and it is always labeled as inferred;
• null instead of a guess when a signal isn't on the site.

Company Enrichment vs Clearbit vs Apollo vs Clay vs ZoomInfo

An honest comparison — these are strong platforms; the difference is the model you pay for and where the data comes from:

	This Actor	Clearbit (HubSpot Breeze)	Apollo	Clay	ZoomInfo
Data source	Live crawl of the company's site	Static database + ML	Static database	Aggregates other providers' credits	Static database
Pricing model	Pay per result: $20/1k	Credit packs via HubSpot	Per-seat plans (from ~$49/user/mo)	Monthly credit plans (from ~$134/mo)	Annual contract, typically five figures
Subscription required	No	Yes	Yes (beyond free tier)	Yes	Yes
Seat licenses	None	Per HubSpot seat	Per user	Per workspace	Per user
Confidence score per field	Yes, 0–1	Partial	Partial	Depends on source	No
Live MX verification	Yes	Partial	Partial	Via credits	Partial
Pay $0 for weak records	Yes (minConfidence)	No — credits burn either way	No	No — credits burn either way	No
Minimum commitment	$0.03 (one domain)	Credit pack	Monthly plan	Monthly plan	Annual contract

(Competitor pricing is list pricing at the time of writing; check each vendor for current terms.)

When to still use them: if you need a 200M-contact search engine to discover companies you don't know yet, a database product fits. If you already have the domains and need them enriched accurately today, that's this Actor.

How to enrich a list of company domains

1. Paste domains (or full URLs), one per line.
2. Optionally tune maxPagesPerDomain (crawl depth, default 5) and minConfidence (the bar below which you are not charged, default 0.3).
3. Run — export JSON, CSV, or Excel, or push to your stack via Apify integrations.

{ "domains": ["stripe.com", "notion.so"], "maxPagesPerDomain": 5, "minConfidence": 0.5 }

Real output shape (one row per domain):

{
  "domain": "stripe.com",
  "companyName": "Stripe",
  "industry": "Financial Services & Fintech",
  "naicsCode": "522320",
  "employeesEstimate": { "band": "1001-5000", "source": "site-copy" },
  "technographics": { "Frontend": ["Next.js"], "Payments": ["Stripe"] },
  "linkedinUrl": "https://www.linkedin.com/company/stripe/",
  "emailPattern": "first.last",
  "emails": [{ "email": "press@stripe.com", "type": "role", "onDomain": true, "confidence": 0.7 }],
  "decisionMakers": [{ "name": "Jane Smith", "title": "VP Sales", "email": "jane.smith@stripe.com", "emailInferred": false, "confidence": 0.82 }],
  "domainHasMx": true,
  "mxProvider": "Google Workspace",
  "confidenceScore": 0.85
}

Every field's provenance is documented: companyName/description from Schema.org & OpenGraph, industry classified from site content, employee band from JSON-LD → "team of N" copy → named team count, founded year from schema → copy → © range, logo from schema → og:image → favicon.

Finding the CEO, founders and other decision makers on a domain

decisionMakers returns named people found on the domain — typically from an about, team or leadership page — with the role text as published.

How often it fills: roughly one domain in five. A company that publishes a leadership page gets you names; most sites do not have one, and the Actor returns an empty array rather than a guess. Treat this column as a bonus that sharpens a subset of the list, not as the reason to run the enrichment.

When it is empty, linkedinUrl is the reliable fallback: it comes back for essentially every domain that has a company page, and it is one click from the people you were looking for.

Working out a company's email pattern before you guess an address

emailPattern is reserved for domains that expose enough addresses to infer a shape such as {first}.{last}@.

In practice it is almost always empty, because inferring a pattern needs several personal addresses on the same domain and most sites publish one role address or none. Do not build a workflow that depends on this column.

What does work: the emails array returns the addresses actually published on the site — about two domains in five carry at least one. A run over your list, filtered to rows where emails is non-empty, is a real send list. Rows where it is empty are form-only companies, which is a different play, not a failed lookup.

Checking deliverability before a domain enters a cold email campaign

Every record includes a live DNS lookup of the domain's MX records, done at run time, not read from a cache:

• domainHasMx — true when the domain publishes mail exchangers at all. false means nothing can receive mail there: a parked domain, a redirect, or a brand that routes mail through a different domain. Those are the rows to drop before you import a list, and they are why part of a purchased list bounces on day one.
• mxProvider — who runs the mailbox, classified from the MX hostnames: Google Workspace, Microsoft 365, Zoho, Proofpoint, Mimecast, GoDaddy, Amazon SES, Transactional ESP, or Other. Proofpoint and Mimecast mean a security gateway is filtering inbound mail, which is useful to know before you decide on send volume and copy.

This is an infrastructure check on the domain, not an SMTP probe of each mailbox — it tells you mail can be received there and by whom, without touching the recipient's server.

Employee count and revenue band from a website alone

Both employeesEstimate and revenueBand are best-effort and depend on the company publishing the figure.

Measured on real domains: employeesEstimate fills for about one in five, revenueBand almost never. A website rarely states its own headcount or revenue, and this Actor reads the website rather than buying a third-party database, so when the site is silent the field is empty. That is the honest trade of live enrichment: everything returned is sourced from the company right now, and nothing is invented to fill a column.

If firmographic size is the field your scoring model runs on, size it from what does come back reliably instead: industry plus naicsCode, the length of technographics, and whether the domain runs enterprise infrastructure.

Finding a company's industry and NAICS code from its domain

industry, naicsCode and sicCode come back for every domain that resolves — these are the columns that survive a full list, so they are the ones to build segmentation on.

The pairing is what makes them useful. industry is the readable label you put in a CRM field; naicsCode and sicCode are the standard codes your RevOps team already filters on, and they let you match against government and industry datasets that are keyed by code rather than by name.

For a list that arrived as bare domains with no context, running this and grouping by naicsCode is usually the fastest way to see what you actually have before deciding who to contact.

Company technology and stack detection by domain

Two columns describe what a company runs: techStack is the readable list (["HubSpot", "Cloudflare", "React"]) and technographics is the structured view of the same detection.

Both fill for essentially every live domain, which makes them the most dependable filter in the output. Concretely:

• selling a migration — filter for the platform you migrate away from
• selling an integration — filter for the tool it plugs into
• qualifying by budget — a domain running several paid SaaS products is spending money, which is a better signal than a headcount guess

Because the detection runs against the site live, it reflects what the company uses today rather than what a database recorded months ago.

Checking a domain can receive email before importing it

domainHasMx and mxProvider answer a question most enrichment skips: can this domain accept mail at all?

domainHasMx is a live DNS check. A domain with no MX record cannot receive email, so every address you might send there will bounce — and a bounce on a cold domain costs sender reputation, not just one lead. mxProvider names who handles the mail (Google Workspace, Microsoft 365, and so on), which is itself a segmentation signal.

Filtering an import on domainHasMx before it reaches your sequencer is a cheap way to protect the sending domain you spent months warming.

Reading the confidence score

Every row carries confidenceScore, which reflects how much the Actor could actually verify on that domain rather than how complete the row looks.

Use it as a sort, not a filter: work the high-confidence rows first, and treat the low ones as leads that need a human glance. A low score usually means the site was thin — a single-page site, a parked domain, or a redirect — and no enrichment tool will do better on it.

What teams use this for

• Outbound prospecting — domain in, decision-maker + MX-verified email out.
• CRM append — backfill industry, NAICS, size, and tech stack on existing accounts.
• ICP & TAM building — segment by size, industry, country, or technographics ("everyone on Shopify + Klaviyo").
• Technographic & competitive intel — the categorized tool list a company actually runs.
• Investor and recruiter research — size, founding year, and who to contact, in one row.

Frequently asked questions

Is this a Clearbit alternative without a subscription? Yes — that's the core trade. You get firmographics, technographics, and contacts per domain, billed per result ($20/1k) instead of through credit packs or an annual platform contract. You lose Clearbit's discovery search; you keep (and improve on) enrichment freshness.

How accurate are the emails? Each email ships with a 0–1 confidence score, an onDomain flag, and a live MX check on the domain. Constructed pattern emails are explicitly flagged emailInferred: true — you always know confirmed vs. inferred.

Can it replace Apollo for enrichment? For enrichment of domains you already have, yes, without the per-user plan. For searching a contact database you don't have inputs for, Apollo-style tools still make sense.

Do I need an API key or account with a data vendor? No. It runs entirely on Apify. You can also call it via the Apify API from your own code and get JSON back.

What does a 1,000-domain run cost? $0.01 start + $0.02 per record that clears your minConfidence. If 700 of 1,000 resolve with substantive data: ≈ $14.01. The 300 weak records cost nothing.

Why are some fields null? Because the signal wasn't published on the site. Returning null instead of a fabricated value is deliberate — it's what makes the non-null fields trustworthy.

Does it find decision-makers at large enterprises? Best results are companies that publish team/leadership pages (SMB, mid-market, agencies, professional services). Enterprises that hide their org behind a login still return firmographics, tech stack, and role emails.

Can I use the output in HubSpot / Salesforce / Sheets? Yes — export CSV/Excel directly, or wire the dataset to your CRM with Apify's integrations (Zapier, Make, webhooks).

Is it GDPR-safe? It reads only publicly published business information from the company's own website. You are responsible for lawful use in your jurisdiction.

How is this different from your CRM Enrichment Actor? This one returns the full nested profile (all emails, all decision-makers, categorized technographics). CRM Enrichment returns one flat row per company shaped for direct CRM import.

Related Actors

• CRM Enrichment — flat, import-ready company rows.
• People Enrichment — name + company → title, seniority & verified email.
• Lead Enrichment & Scoring — 0–100 ICP-fit score per lead.
• All my Actors — Yahoo Finance, SEC EDGAR, App Store & Google Play, Deep Research MCP.

---

Built by Renzo Madueño. Found a bug or want a field added? Open an Issue — I read and respond to every one.