Skip to main content

​
New

​
Company Dataset

We’re launching a dedicated company dataset. Each company record includes firmographic data (industry, size, domain, LinkedIn profile) and is linked to person records via company_legion_id. This is our first release of company data — coverage and depth will expand in upcoming builds. Person records now include company linkage fields:
  • company_legion_id — Links to the corresponding company record
  • company_linkedin_url, company_linkedin_id — Employer LinkedIn identifiers
  • experience.organization.legion_id — Each job in a person’s history links to its company record
Workforce analytics available on company records:
  • legion_employee_count — headcount derived from person experience data
  • legion_new_hire_count — new hires across 1-month, 3-month, 6-month, and 12-month windows
  • legion_attrition_count — employees who left across the same time windows
  • legion_employee_growth_rate — net growth rate across 1-month, 3-month, 6-month, and 12-month windows
  • legion_turnover_rate — attrition as a percentage of starting headcount across the same time windows
  • legion_seniority_distribution — current employee breakdown by seniority level
  • legion_job_function_distribution — current employee breakdown by job function
  • legion_expense_category_distribution — current employee breakdown by expense category
  • legion_tenure_distribution — current employee breakdown by tenure bucket
  • legion_education_distribution — current employee breakdown by highest degree level
  • legion_seniority_growth_rate — growth rate broken out by seniority level
  • legion_job_function_growth_rate — growth rate broken out by job function
  • legion_expense_category_growth_rate — growth rate broken out by expense category
  • legion_employee_count_by_month — monthly headcount time series with gross hires and departures (up to 60 months)
  • legion_average_tenure — average tenure in months of current employees

​
Address Validation (CASS/NCOA)

We’ve begun processing US mailing addresses through USPS CASS certification and NCOA (National Change of Address). This is an ongoing effort — coverage will increase over the next several builds as we work through the full corpus. For addresses processed so far:
  • Validated to the delivery point level with ZIP+4
  • NCOA forwarding moves applied, keeping addresses current for up to 48 months after a move
  • Addresses that fail delivery point validation are removed
  • Location confidence scoring reflects validation status

​
New Person Fields

  • linkedin_id — Numeric LinkedIn member ID
  • is_platform_worker — Identifies gig economy and platform-based workers (rideshare, delivery, freelance marketplace)
  • postal_code_4 — ZIP+4 extension for CASS-validated US addresses
  • current_jobs_last_updated — Date when a person’s current job information last changed
  • current_jobs_last_confirmed — Date when a source last observed the current job
  • current_location_last_updated / current_location_last_confirmed — Same pair for location data

​
Bug Fixes

​
Seniority Classification

  • Individual contributor roles where “manager” is part of the role name (product manager, project manager, account manager, etc.) were incorrectly classified as seniority_level = "manager". Filtering for managers now returns people with direct reports and budget authority. IC managers get seniority from explicit modifiers — “Senior Product Manager” → senior, “Associate Product Manager” → junior, “Product Manager” → null.
  • HR business partners, talent partners, care partners, and similar non-equity roles were incorrectly classified as seniority_level = "partner". Partner now reliably identifies law firm, consulting, and PE/VC partners.
  • Sole proprietors now classified as owner. Chairman/chairwoman/chairperson now classified as c_level.

​
Job Function Classification

  • Company industry is now used as a disambiguation signal for ambiguous titles.
  • Corrected classification for service advisors, call center roles, insurance producers, and other edge cases.

​
Top-Level Job Selection

  • When a person holds multiple current positions, volunteer and honorary roles (committee member, chapter president, trustee, deacon) were sometimes selected as the primary job. These are now deprioritized so job_title and company_name reflect the person’s primary professional role.

​
Address Formatting

  • Invalid secondary address lines (e.g., fabricated apartment numbers on single-family homes) are stripped.