Apify Linkedin Intelligence

LinkedIn intelligence for B2B go-to-market

Turn LinkedIn's three data pillars — companies (firmographics), people (profiles), and jobs (postings) — into four sales/RevOps workflows: TAM sizing, prospect list building, headcount benchmarking, and hiring-signal selling.

The key insight

LinkedIn value comes from chaining the three pillars, not from any one scraper:

• A company search gives you the accounts in a market (TAM).
• A people search inside those accounts gives you the contacts (prospects).
• A jobs feed for those accounts gives you the timing (hiring signals = budget + pain).

So the routing decision is "which pillar(s) does this question touch", then chain them. The output of one pillar (a company slug, a company name, a profile URL) is the input to the next.

Always confirm the input schema before running (apify actors info "ACTOR_ID" --input --json). LinkedIn Actor inputs differ between Actors and change often — the same intent ("scrape a company") uses profileUrls in one Actor and companyUrls/identifier in another. A wrong field name silently returns 0 results.

Prerequisites

• Apify account (sign up)
• Authentication via one of:
  • apify login (OAuth, if using the Apify CLI)
  • APIFY_TOKEN environment variable
  • Token from Apify Console → Settings → Integrations

Workflow

1. Classify the goal into one of the four workflows below (or a combination), and identify which pillar(s) it needs.
2. Resolve identifiers first. LinkedIn Actors key off exact slugs/URLs, not loose names. Discover them via SERP before scraping — apify/google-search-scraper with "<company> site:linkedin.com/company" (company slug) or "<title> <company> site:linkedin.com/in" (people). A wrong slug returns 0 rows silently.
3. Confirm the schema. Run apify actors info "ACTOR_ID" --input --json and build input against the actual field names. Note where output lands (dataset vs key-value store — see gotchas).
4. Estimate cost, then run. LinkedIn Actors are usually PAY_PER_EVENT (per profile/company/job). Multiply by your row count and confirm with the user if the estimate is significant (see references/gotchas.md).
5. Deliver. Report row count, the columns, and a link to the dataset/console. Default to CSV for lists; summarize counts/medians for TAM and benchmarking.

The four workflows

Actor routing

LinkedIn is a protected platform — never use a generic crawler (website-content-crawler, rag-web-browser) on linkedin.com. Use a dedicated Actor.

Tier = apify (Apify-maintained, prefer) or community (third-party). These are starting points — run apify actors search "linkedin" --json --limit 20 to find current alternatives, and always confirm the schema (step 3). See references/actor-index.md for the full table and field notes.

Calling Actors — Apify CLI

Three flags on every call (--json, --user-agent, 2>/dev/null):

# 0. Resolve the exact company slug first (names != slugs: "Oxylabs" -> "oxylabs-io")
apify actors call "apify/google-search-scraper" \
  -i '{"queries":"Acme Corp site:linkedin.com/company","maxPagesPerQuery":1,"resultsPerPage":5}' \
  --json \
  --user-agent apify-awesome-skills/apify-linkedin-intelligence \
  2>/dev/null

# 1. Company firmographics (TAM / headcount benchmarking) — output is in the KV store
apify actors call "dev_fusion/Linkedin-Company-Scraper" \
  -i '{"profileUrls":["https://www.linkedin.com/company/oxylabs-io/"]}' \
  --json \
  --user-agent apify-awesome-skills/apify-linkedin-intelligence \
  2>/dev/null

# 2. People search (build a prospect list by title + company + geo)
apify actors call "harvestapi/linkedin-profile-search" \
  -i '{"currentCompanies":["Acme Corp"],"jobTitles":["VP Sales","Head of Revenue"],"locations":["United States"],"maxItems":50}' \
  --json \
  --user-agent apify-awesome-skills/apify-linkedin-intelligence \
  2>/dev/null

# 3. Enrich specific profile URLs
apify actors call "harvestapi/linkedin-profile-scraper" \
  -i '{"profileUrls":["https://www.linkedin.com/in/some-handle/"]}' \
  --json \
  --user-agent apify-awesome-skills/apify-linkedin-intelligence \
  2>/dev/null

# 4. Hiring signals — jobs from a LinkedIn jobs SEARCH URL (not keywords), count >= 10
apify actors call "curious_coder/linkedin-jobs-scraper" \
  -i '{"urls":["https://www.linkedin.com/jobs/search/?keywords=Account%20Executive&location=United%20States"],"count":10,"scrapeCompany":true}' \
  --json \
  --user-agent apify-awesome-skills/apify-linkedin-intelligence \
  2>/dev/null

# Inspect input schema (DO THIS before every new Actor) / fetch dataset results
apify actors info "harvestapi/linkedin-profile-search" --input --json \
  --user-agent apify-awesome-skills/apify-linkedin-intelligence 2>/dev/null
apify datasets get-items DATASET_ID --format json \
  --user-agent apify-awesome-skills/apify-linkedin-intelligence 2>/dev/null

The input fields above are illustrative — confirm them against each Actor's real schema before running. The Apify MCP server (https://mcp.apify.com) and any MCP client (e.g. mcpc) are equivalent alternatives.

Troubleshooting

• 0 rows returned → almost always a wrong slug/name/URL or wrong input field. Re-resolve the slug via SERP (step 2), and re-check the field name against apify actors info ... --input.
• Company scraper "returned nothing" → dev_fusion/Linkedin-Company-Scraper writes to the key-value store, not the dataset. Read the KV store keys after the run.
• Jobs Actor rejects keywords → curious_coder/linkedin-jobs-scraper needs a full linkedin.com/jobs/search/?... URL, and count min is 10. URL-encode multi-word values (Bright Data → Bright%20Data).
• Headcount looks off → LinkedIn employee/TAM counts are modeled and directional, not a ground-truth census. Label them as estimates; cross-check against the company website or jobs Actor's companyEmployeesCount.
• Cost climbing on a big list → these Actors bill per result. Cap maxItems, run a small sample first to validate the schema, then scale. See references/gotchas.md.