# Microtargeting The decomposition of a campaign's audience into thousands of narrow segments, with a different advertisement, message, or appeal delivered to each — so that what any particular voter sees is incompatible with what other voters see, and the overall content of the campaign is not visible from any single vantage. The technique is descriptively *targeted advertising* applied to politics, but the analytical interest is in how it dissolves the public-square assumption that political communication is a *shared* artifact subject to common scrutiny. ## The provenance Microtargeting as a verb migrated from commercial direct-mail and database marketing — the 1990s vocabulary of consumer-segment analysis — into political work through three sequential moments: 1. **Hal Malchow / DSCC (late 1990s)** — Democratic data-targeting consultant who brought commercial-database modeling techniques into Senate-campaign voter contact. The first wave was small-bore: which households on a street to door-knock based on inferred persuadability 2. **Karl Rove / Bush 2004** — first national-scale deployment. The Bush reelection identified roughly 80 narrow voter segments and assigned each a custom appeal package; the campaign credited microtargeting as a decisive contributor to the Ohio outcome. The Republican-side infrastructure (Voter Vault, later i360) accumulated household-level data for tens of millions of Americans 3. **Obama 2008 and 2012** — Democratic-side scaling using the Catalist data cooperative and Project Narwhal / Dashboard custom-software stack. The 2012 campaign's microtargeting and modeling work was widely-credited and widely-imitated; it set the modern template By 2014 the *infrastructure* of microtargeting was a normal feature of both parties' campaigns. What made the post-2016 cycle distinct was the addition of one more layer: psychographic modeling. ## The Cambridge Analytica disclosures [[Cambridge Analytica Scandal]] — the [[SCL Group]] political-affiliate funded by [[Robert Mercer]], run by Alexander Nix, and contracted by the 2016 Cruz and Trump campaigns and the Vote Leave Brexit campaign — claimed to use psychographic profiles built from harvested Facebook data (~87 million users via the [[Aleksandr Kogan]] "thisisyourdigitallife" personality-quiz app) to identify voter segments by inferred OCEAN personality traits (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism) and tailor messaging to each segment's psychological profile. The 2018 [[Christopher Wylie]] whistleblower disclosures and Channel 4's hidden-camera footage of Nix discussing entrapment-and-honeypot political tactics turned Cambridge Analytica into the public face of psychographic microtargeting. The firm collapsed in May 2018; the FTC and ICO investigations produced multimillion-dollar fines for Facebook (the FTC's $5 billion penalty in 2019 was the largest in FTC history). **Methodological caveat**: many independent researchers concluded after the fact that Cambridge Analytica's *psychographic* claims were overstated — the company's technical capability did not necessarily match its marketing pitch. The deeper concern, regardless, is structural: a market for political-targeting services using personal data exists at scale, and the regulatory regime governing it is still uneven. ## Why microtargeting matters analytically Three features make microtargeting distinctively dangerous for democratic discourse, independent of any particular campaign's technical sophistication: 1. **No shared text** — when there is no common version of a campaign's appeal, opposition research, fact-checking, and citizen deliberation lose their object. A claim made to one segment is invisible to the segment that would object to it 2. **Dark money pipeline compatibility** — microtargeted ads can be funded through opaque vehicles, bypass disclosure requirements that assume broadcast-era media buying, and avoid the contextual signaling (which network, which programming) that historically helped voters interpret ad sponsors 3. **A/B optimization at scale** — when a campaign can test thousands of message variants per day against narrow segments and select for engagement, the optimization process selects *for whatever provokes reaction in that segment*, which empirically is more often outrage, fear, or in-group flattery than information The Trump 2016 digital team under [[Brad Parscale]] ran roughly 5.9 million distinct ad variants on Facebook during the campaign cycle; the Clinton campaign ran roughly 66,000. The asymmetry is not necessarily about quality — it is about industrial-scale A/B testing as a campaign instrument. ## Regulatory responses The legal regime governing political microtargeting remains patchwork: - **EU** — GDPR (2018) and the Digital Services Act (2022) impose meaningful constraints on political-data processing and require platforms to maintain ad transparency archives - **United States** — no comparable federal regulation; the Federal Election Commission's online-ad rules have remained largely unchanged since the broadcast era. State-level laws (California's CCPA, Washington's I-940) provide partial coverage - **Platform self-regulation** — Google banned political microtargeting on Search and [[YouTube]] (2019); Facebook / Meta narrowed its political-targeting categories (2021) but retained broad demographic targeting; X removed most political-ad restrictions under Musk ownership (2023) The 2022 *DSA* and 2024 EU Regulation on the Transparency and Targeting of Political Advertising represent the most-developed regulatory model; whether US political microtargeting will face comparable disclosure or restriction in the 2026–2028 cycles is an open question. ## Adjacent vocabulary Microtargeting sits inside a vocabulary cluster: - **Psychographic profiling** — the underlying personality-inference technique (often Big Five / OCEAN modeling) - **Surveillance capitalism** — [[Shoshana Zuboff]]'s 2019 framing of the underlying data-harvesting business model - **Dark patterns** — UX manipulation techniques that often appear together with microtargeted appeals - **Dark ads** — the British term for unlogged paid political content; partially addressed by Facebook's Political Ad Library - **Targeted advertising** — the commercial parent category; microtargeting is its political-domain instance ## See Also - [[Cambridge Analytica Scandal]] — the canonical case study and brand-name scandal - [[Christopher Wylie]] — the whistleblower - [[Karl Rove]] — the political-microtargeting practitioner who scaled it to national reelection - [[Brad Parscale]] — the digital-director architect of the Trump 2016 Facebook ad operation - [[Surveillance Capitalism]] — the structural-economic frame - [[Disinformation]] — the broader categorical context - [[Disinformation Hub]] — Subject hub - [[Disinformation Lexicon]] — alphabetical MOC