Gov Influence
Menu

Identity review

How entity matching works

Federal systems may spell or structure the same entity differently. Similar names can also identify different organizations, so Gov Influence resolves identity before calculating or linking records.

Reviewed Sep 5, 2026 · Examples cover Q1–Q2 2026 lobbying records and 119th Congress legislative records

Five identity decisions

  • Exact alias: a complete approved source-record variant can map to the canonical entity.
  • Former name: dated source wording can document a prior name without merging every record that uses it.
  • Affiliate: a related but separately identified entity is not automatic evidence of sameness.
  • Subsidiary: ownership language does not automatically merge separately filed clients.
  • False positive: a similar complete name can remain excluded when it identifies another entity or fails the allowlist.

Worked examples: Meta, Amazon, and Boeing

Meta totals accept the exact aliases “META PLATFORMS, INC.” and “META PLATFORMS INC.” The reviewed former-name record preserves Facebook wording but does not add subsidiaries or standalone Facebook client records. Amazon totals use the exact “AMAZON.COM SERVICES LLC” alias; “AMAZON WEB SERVICES, INC.” and “AMAZON.COM, INC.” remain separate source names. Boeing accepts “BOEING,” “BOEING COMPANY,” and “THE BOEING COMPANY,” while “BOEING EMPLOYEES' CREDIT UNION” is excluded as a false positive.

How to read this on a profile

  • Compare the canonical name with included source-record names.
  • Open the alias-review disclosure to see exclusions.
  • Treat former-name evidence as dated context, not permission to merge affiliates or subsidiaries.
  • Check the record count after the canonical amendment rule.
  • For cross-source links, read the match basis and public limitation.

Visitor checklist

  • Is the complete source name an approved alias?
  • Is there a stable official identifier?
  • Could the name describe an affiliate, subsidiary, committee, or unrelated entity?
  • Is former-name evidence dated and sourced?
  • Was the match reviewed for publication?
  • Would the conclusion change if the match were wrong?

What is intentionally hidden

Internal candidates that did not pass review are not exposed. Published profiles show only the reproducible approved aliases, included source names, and safe exclusion examples needed to explain the boundary.

Profiles and documentation