Chapter 29 Key Takeaways: Voter Targeting and Microtargeting

Core Concepts

1. Targeting is resource allocation, not magic. Campaigns have finite money, staff time, and volunteer hours. Targeting is the systematic attempt to direct those resources toward voters where they will have the greatest electoral impact. It is about efficiency, not omniscience. Targeting improves your probability of winning within your structural constraints — it does not transcend them.

2. The three core universes serve distinct purposes. The GOTV universe (likely supporters, uncertain turnout) requires mobilization messaging. The persuasion universe (genuinely undecided, likely to vote) requires persuasion messaging. The fundraising universe (likely supporters, financial capacity) requires donation asks. Using the wrong contact approach with the wrong universe wastes resources and can backfire — a mobilization message delivered to a persuasion target, for example, may implicitly assume support that hasn't been established.

3. Targeting requires answering three questions simultaneously: Who, What, and Through What Channel. Identifying the right voters (who) but delivering the wrong message (what) or through the wrong medium (channel) still fails. Effective targeting integrates all three dimensions — voter selection, message design, and channel allocation — into a coherent, coordinated program.

4. Predictive scores are probabilities, not facts. A support score of 75 means a 75% estimated probability of supporting the candidate — not a guarantee. In a large targeting universe, many voters with high support scores will vote for the opponent. Treating model predictions as certainties is one of the most consequential errors in campaign analytics. Score-based targeting is about improving the expected value of outreach, not eliminating uncertainty.

5. Persuadability is the hardest score to build and the most important to be skeptical of. Persuadability cannot be directly observed — it must be inferred from behavioral correlates and experimental data. The signal-to-noise ratio in persuadability modeling is substantially lower than in turnout or support modeling. Use persuadability scores for rough categorization (top third vs. bottom third of the distribution), not for precise individual-level predictions.

Microtargeting Mechanics

6. Microtargeting customizes both message and channel to individual voter characteristics. Standard targeting identifies who to contact. Microtargeting goes further, calibrating what to say and how to say it based on each voter's inferred issue priorities, channel preferences, and psychological characteristics. The efficiency gains are real; so are the ethical risks.

7. Facebook Custom Audiences enabled identity-based digital targeting. The ability to match voter file records to social media accounts allowed campaigns to move from demographic/interest targeting (reaching a type of person) to identity-based targeting (reaching a specific, named individual). This shift fundamentally changed the scale and precision of digital political advertising — and raised corresponding accountability concerns.

8. Third-party consumer data has significant reliability problems. Match rates between consumer data files and voter files can be misleading because average match rate conceals wide variation in record quality. Consumer data ages quickly. Aggregate consumer-political correlations are real but noisy at the individual level. Canvass data from direct voter contact provides a much higher-quality signal than consumer data inference and should be weighted accordingly.

Ethical Dimensions (Central Theme)

9. The line between targeting and manipulation runs through intent and content. Targeting that delivers relevant, substantive political information to voters who care about an issue is a form of democratic communication. Targeting that identifies psychological vulnerabilities and exploits them with fear-based or deceptive messaging is manipulation. The same infrastructure enables both. The ethical line is not built into the tools — it is enforced by professional norms, legal constraints, and imperfect external scrutiny.

10. Microtargeting creates accountability gaps that broadcast political communication does not. A TV ad is public — it can be seen, fact-checked, and critiqued by journalists, opponents, and any citizen. A microtargeted digital ad visible only to specifically identified voters contains claims that may never face external scrutiny. The fragmentation of political communication into privately targeted messages is a democratic accountability problem that disclosure reforms have only partially addressed.

11. Efficient targeting has distributional consequences for democratic participation. Campaigns that target efficiently reach voters where they'll have the most electoral impact — and systematically ignore voters who aren't worth the investment. The voters who fall outside the target universe (committed opponents, very low-turnout voters, hard-to-reach communities) receive less political information and less mobilization pressure. The demographic and geographic correlates of non-contact are not random; they often amplify existing inequalities in political resources and voice.

12. Voter suppression through messaging is a real and documented concern. Microtargeting infrastructure can be used to deliver demobilizing messages — false voting information, persuasive arguments to stay home — to specifically identified opposing voters. This use of targeting is distinct from legitimate political communication, but it uses the same data infrastructure and is difficult to detect from outside the campaign.

Evidence on Effectiveness

13. GOTV targeting improves efficiency; the evidence is strong. Field experiments consistently show that personal canvassing and phone banking increase turnout. Campaigns that direct these resources toward likely supporters with uncertain turnout — rather than random registered voters — see better net gains. Targeting doesn't make the contacts more effective; it makes sure the contacts are delivered where they help the right candidate.

14. Persuasion targeting evidence is more mixed. Studies of issue-specific mail targeted to voters with matching issue affinities find some improvements over generic mail, but effect sizes are small and many experiments find no significant difference. The reliability problem in persuadability scoring limits the efficiency gains from persuasion targeting.

15. Digital microtargeting effectiveness is contested and likely overstated by vendors. Platform-published studies and vendor claims find large digital ad effects. Academic studies using rigorous experimental designs generally find much smaller effects, sometimes indistinguishable from zero. Campaigns should be skeptical of vendor effectiveness claims and demand to see methodology and validation data before investing in novel digital targeting techniques.

Campaign-Specific Lessons

16. Nadia's three-track message matrix demonstrates the tradeoff between analytical sophistication and operational feasibility. The optimal number of message tracks is not infinite — campaigns face production costs, staff capacity constraints, and coordination challenges that limit how many distinct targeted messages they can manage simultaneously. Analytical sophistication should be calibrated to operational capacity.

17. Jake's base mobilization focus reflects a coherent theory of electoral math. Mobilizing committed supporters in high-support territory is a legitimate alternative to persuasion targeting, particularly when the supporting coalition is geographically concentrated and the persuasion universe is small and difficult to move. Neither mobilization nor persuasion targeting is always superior — the right balance depends on the specific electoral math of the race.

18. The Cambridge Analytica case illustrates the vendor credibility problem. Campaigns pay premium fees for targeting techniques that have not been independently validated. The episode demonstrates both that sophisticated-sounding targeting claims are commercially persuasive regardless of empirical support, and that the data acquisition practices underlying some claimed innovations may themselves violate ethical and legal norms. Rigorous evaluation of vendor claims is an essential campaign analytics competency.