Skip to main content

Research Methodology

Research Methodology

Approach

This is financial OSINT applied to political movements. The goal is mapping networks, not just listing donations.

Core Questions (per target)

  1. Who gives them money? (Donors, patrons, crowdfunding backers)
  2. Where does it come from? (Domestic/foreign, individual/corporate/state)
  3. How does it move? (Direct donation, company pass-through, charity, crypto, cash)
  4. What do they spend it on? (Media, legal fees, organising, property, personal enrichment)
  5. Who else does the funder support? (Network mapping — shared donors reveal alliances)

Research Phases

Phase 1: Public Records Sweep

  • Electoral Commission: all donations >£7,500 to registered parties
  • Companies House: directorships, shareholdings, PSC (Persons with Significant Control) for donor companies
  • Charity Commission: registered charities with political adjacent activity
  • Land Registry: property ownership for key figures and orgs
  • OpenCorporates: international company connections

Phase 2: Informal Economy

  • Crowdfunding platforms: GoFundMe, GiveSendGo, Kickstarter (archived campaigns)
  • Patreon: creator pages, patron counts, tier structures
  • Telegram: paid channels, donation bots, crypto wallets posted in chats
  • YouTube/X monetisation: SuperChats, subscriber counts, sponsor credits
  • Merchandise stores: Shopify/Teespring/etc. — domain WHOIS, company registration

Phase 3: International Flows

  • US 501(c)(3) and 501(c)(4) filings for orgs with UK activity
  • IRS Form 990 for American donors funding UK entities
  • Russian foundation reports (where available pre-sanctions)
  • EU populist party financial disclosures (Europarl requirements)
  • Crypto wallet analysis (public addresses posted by groups)

Phase 4: Network Mapping

  • Build entity graph: Person → Company → Donation → Recipient
  • Identify shared directors, shared addresses, shared accountants
  • Map temporal patterns (donation clusters around events — riots, elections, court cases)
  • Cross-reference with known intelligence front organisations

Verification Standards

Confidence Levels

Level Definition Evidence Required
CONFIRMED Documented in official/public record Electoral Commission filing, court document, company filing
REPORTED Published by credible investigative source Named journalist at major outlet, academic paper, NGO report
ALLEGED Claimed by involved party or single source Self-reported, leaked document, anonymous source
SPECULATIVE Pattern-based inference Temporal correlation, shared address, circumstantial

Rules

  • Never publish SPECULATIVE as fact. Always label clearly.
  • Archive every URL within 24 hours of citing it.
  • Cross-reference: a fact needs 2+ independent sources to be CONFIRMED.
  • Distinguish between "X donated to Y" and "X's company donated to Y" — legal personality matters.
  • Track name changes: far-right groups rebrand frequently.

Tools & Techniques

OSINT Toolkit

  • Maltego / yEd: Network graph visualisation
  • OpenRefine: Bulk data cleaning (Electoral Commission CSVs)
  • Wayback Machine / archive.today: URL preservation
  • Aleph (OCCRP): Cross-border entity search
  • OpenSanctions: Sanctions/PEP database checks
  • ICIJ Offshore Leaks: Panama Papers, Pandora Papers connections

Search Patterns

  • Site-specific: site:electoralcommission.org.uk "Reform UK"
  • Company searches: site:companieshouse.gov.uk [director name]
  • Charity searches: site:charitycommission.gov.uk [org name]
  • Court records: site:caselaw.nationalarchives.gov.uk [name]

Data Collection

  • Download Electoral Commission bulk data quarterly
  • Companies House API for structured company/person data
  • Charity Commission API for registered charity searches
  • FOI requests to Electoral Commission for refused donations

  • Electoral Commission data is public record — safe to use.
  • Companies House data is public record — safe to use.
  • Charity Commission data is public record — safe to use.
  • GDPR: living individuals have rights. Public interest defence applies to political research, but be careful with personal financial data not in public record.
  • Defamation: always distinguish between fact and allegation. Don't allege criminal conduct without evidence.
  • Contempt of court: be careful around active proceedings.