WorkflowDigital CampaignsIntermediate

Investigating Political Ads Workflow

A workflow for researching political advertising with public ad-transparency libraries — scope, search, extract, structure, analyse spend and targeting, and cite.

ElectionsTechUpdated March 20, 20252 min read
Share:

This workflow turns public ad-transparency libraries into a structured, analysable dataset — the practical companion to the investigating political advertising guide. The goal is to move from scrolling an ad library to producing citable evidence about funders, spend, and targeting.

Overview

  • Purpose: research political advertising transparently.
  • Objective: a structured dataset and spend timeline.
  • Difficulty: Intermediate.
  • Time: 1–3 hours.
  • Output: a structured ad dataset, a spend timeline, and an evidence log.

Step 1 — Define scope

Choose an advertiser, keyword, region, and date range. A tight scope keeps the dataset manageable and the analysis focused on an answerable question.

Step 2 — Search the libraries

Query the Meta Ad Library and the Google Ads Transparency Center. Record advertiser identity and verification status, which are often as revealing as the ad content.

Step 3 — Extract the fields

For each ad, capture: advertiser, funding entity, spend range, impressions, dates, targeting, and the creative. For scale, use the bulk Meta and Google datasets rather than manual collection.

Step 4 — Structure the data

Load everything into a spreadsheet. Normalise advertiser names and currencies so records are comparable, and flag any that need manual review.

Step 5 — Analyse

  • Build a spend timeline to see how spending tracks events and dates.
  • Look for duplicated creatives across many pages — a signal of coordination or astroturfing.
  • Identify funder mismatches between the disclosed payer and the apparent beneficiary.

Step 6 — Preserve and cite

Archive every ad page and cite the transparency source for each figure. Ads get removed; a preserved copy protects your finding.

Outputs

  • A structured ad dataset with normalised fields.
  • A spend-over-time timeline.
  • An evidence log with archived pages and citations.

Common mistakes

  • Comparing spend across platforms with different definitions.
  • Ignoring each platform's "political" classification caveats.
  • Failing to archive ads that are later removed.

Next steps

Read the accompanying investigating political advertising guide for context, and if the data suggests coordination, continue with investigating disinformation campaigns.

#political-advertising#transparency#osint#data-journalism

Frequently asked questions

How do I research who is behind a political ad campaign?

Search the platform ad libraries by advertiser, page, or keyword, then extract the funding entity, spend, impressions, dates, and targeting into a spreadsheet. Look for funder mismatches and duplicated creatives across pages, and corroborate across multiple libraries.

Can I compare ad spend across platforms?

Cautiously. Each platform defines "political," measures spend, and reports impressions differently, so cross-platform totals are not directly comparable. Analyse each platform on its own terms and disclose the differences.

What outputs should a political ad investigation produce?

A structured dataset of ads and their metadata, a spend-over-time timeline, and an evidence log with archived ad pages and citations for every figure.