Skip to content

4 min read

Microtargeting

Summary

Microtargeting is the campaign practice of using voter-file data, demographic model outputs, and persuasion/turnout propensities to direct distinct messages to distinct individuals. It is the operating application of data-science models to campaign decisions: door-knock targeting, direct-mail segmentation, message-matching, and donor-prospect scoring.

Body

Microtargeting in the modern sense traces to the 2004 Bush-Cheney campaign’s data-driven voter-contact program and matured through campaigns of the 2010s. The discipline rests on three components: a voter file (the consumer/registration/voting-history dataset that anchors every score); a model (a propensity or persuasion score trained on polling or turnout data, often with causal-inference correction); and a delivery mechanism (door-knocks, mail, phone, digital) that can route the chosen message to the chosen voter. [source: civis-analytics-methodology]

The empirical literature on microtargeting has expanded rapidly since the Cambridge Analytica controversy. Recent academic work distinguishes audience segmentation (group-level message variants) from individual persuasion modeling (per-voter scores fed by granular features), and finds that the latter’s effect sizes are uneven and dependent on modelling transparency. The Wesleyan Media Project’s CREATIVE dataset enables academic content analysis of cross-cycle ad creative, which sits downstream of microtargeting. [source: wesleyan-media-project]

Microtargeting’s center of gravity has shifted toward social-media platforms, where delivery is cheaper and targetability is finer. This raises the question of where the trade-off between reach and precision lies: a campaign that microtargets 10 percent of voters with a high-match message may underperform a campaign that broadcasts to 60 percent with a lower-match message. The mobilization-vs-organization literature (Hahrie Han; Marshall Ganz) treats this as a strategic choice, not a technical optimisation. [source: get-out-the-vote]

The ethical and regulatory perimeter of microtargeting is contested. EU data-protection regulation (GDPR) and US state-level “voter choice” disclosures shape what is permissible; political-consulting ethics codes (EAPC Lisbon Declaration; IAPC global code) provide voluntary professional norms. The discipline lives at the intersection of behavioural-science effect sizes, data-protection rules, and the consulting-ethics frame. [source: eapc]

Use it for

Designing or auditing a voter-file-driven campaign targeting plan; choosing between audience-segmentation and individual-persuasion modeling for a given campaign context; calibrating data-protection compliance for a microtargeting program; making the strategic trade-off between broadcast reach and targeted depth explicit in a campaign plan.

FAQ

What is microtargeting?

Microtargeting is the campaign practice of using voter-file data, demographic model outputs, and persuasion/turnout propensities to direct distinct messages to distinct individuals [source: civis-analytics-methodology]. It is the operating application of data-science models to campaign decisions: door-knock targeting, direct-mail segmentation, message-matching, and donor-prospect scoring.

What three components does microtargeting rest on?

The discipline rests on three components: a voter file (the consumer/registration/voting-history dataset that anchors every score); a model (a propensity or persuasion score trained on polling or turnout data, often with causal-inference correction); and a delivery mechanism (door-knocks, mail, phone, digital) that can route the chosen message to the chosen voter [source: civis-analytics-methodology]. The model is only useful when both the voter file and the delivery channel can carry the model’s chosen message to the chosen individual.

How does microtargeting differ from audience segmentation?

Recent academic work distinguishes audience segmentation (group-level message variants) from individual persuasion modeling (per-voter scores fed by granular features), and finds that the latter’s effect sizes are uneven and dependent on modelling transparency [source: wesleyan-media-project]. The Wesleyan Media Project’s CREATIVE dataset enables academic content analysis of cross-cycle ad creative, which sits downstream of microtargeting.

What is the strategic trade-off between reach and precision?

The center of gravity has shifted toward social-media platforms where delivery is cheaper and targetability is finer, and this raises the question of where the trade-off between reach and precision lies [source: get-out-the-vote]. A campaign that microtargets 10 percent of voters with a high-match message may underperform a campaign that broadcasts to 60 percent with a lower-match message, which is why the mobilisation-vs-organisation literature treats this as a strategic choice rather than a technical optimisation.

What ethics and compliance rules govern microtargeting?

EU data-protection regulation (GDPR) and US state-level “voter choice” disclosures shape what is permissible; political-consulting ethics codes (EAPC Lisbon Declaration; IAPC global code) provide voluntary professional norms [source: eapc]. The discipline lives at the intersection of behavioural-science effect sizes, data-protection rules, and the consulting-ethics frame, which is why microtargeting is also a question for political-consulting-ethics.

Open Questions

  • The persistence-of-effect question: do microtargeting-driven voter contacts develop durable attitudinal and turnout effects, or do the gains wash out after the campaign?
  • The transparency question: should microtargeting models, training data, and effect-size estimates be public, and what is the minimum-disclosure regime that protects against manipulation without banning the practice?

Learn more