Case Study 29.1: Nadia's Message Matrix — Designing the Garza Persuasion Program

Background

By the second week of September, six weeks before Election Day, Nadia Osei and the Garza campaign's communications director, Rafael Ortega, had a problem: the campaign had a sophisticated targeting model but had not yet resolved how to connect model outputs to message strategy. They could identify which voters were in the persuasion universe. They could score those voters on six issue affinity dimensions. But they did not have an agreed-upon approach to translating those scores into distinct message tracks delivered through different channels.

This case study follows the three weeks of work that produced the Garza campaign's message matrix — the framework that connected voter characteristics to message content and delivery channel across the campaign's final six weeks.

The Persuasion Universe Profile

Nadia's targeting model identified approximately 180,000 voters in the Garza persuasion universe — the voters with support scores between 38 and 62, turnout propensity above 55, and persuadability scores above the 55th percentile. She presented a demographic and geographic profile of this universe to the communications team.

Geographic concentration: 67% of the persuasion universe was concentrated in four suburban counties surrounding the state's largest city. The remaining 33% was spread across smaller cities and inner suburbs.

Age distribution: Heavier toward 35–54 than the broader registered voter population. The 35–44 cohort was particularly large — voters who had come of age politically in the Obama era but whose allegiances had loosened over subsequent cycles.

Educational composition: Substantially more college-educated than the statewide voter file. In the four primary suburban counties, 58% of persuasion targets were college graduates, compared to 38% statewide.

Partisan registration: Mixed. About 35% were registered Democrats (whose current-cycle support scores fell in the persuasion range because of issue cross-pressure or general dissatisfaction). About 28% were registered Republicans (whose scores reflected weakening partisan attachment). About 37% were unaffiliated or registered with minor parties.

Issue affinity profile: The persuasion universe skewed toward high scores on economic fairness, healthcare access, and education — and lower relative scores on immigration and criminal justice. This pattern was consistent with what the campaign's qualitative research suggested about this segment of suburban voters: engaged on economic and institutional issues, less engaged on social issues that the broader Democratic base prioritized.

Building the Message Matrix

Rafael Ortega's communications team had developed what they called the "Garza brand pillars" — five thematic frames that the campaign had tested with focus groups and found to resonate with this segment:

1. Competence and Record: Garza's twelve years in state government, her record as Attorney General holding corporations accountable.
2. Healthcare Stability: Her support for the ACA's Medicaid expansion and opposition to Whitfield's stated preference for state-level flexibility (which focus groups read as potential cuts).
3. Education Investment: Her record supporting public school funding during her time as AG and her platform commitment to teacher pay.
4. Economic Accountability: Her record suing predatory lenders and opposing financial deregulation.
5. Institutional Trustworthiness: A contrast with Whitfield's endorsement of election integrity claims that focus groups in this segment rated negatively.

The challenge was that the communications team wanted to run one main persuasion message across the entire universe, while Nadia's targeting model suggested the universe was not homogeneous — different sub-segments showed meaningfully different issue affinity patterns.

Nadia brought data to make the case. She pulled a cross-tabulation of the persuasion universe showing how issue affinity scores varied by sub-segment:

• Registered Democrats in the persuasion universe showed strongest affinity for healthcare (mean score: 7.2/10) and education (7.0/10).
• Unaffiliated voters showed strongest affinity for economic accountability (6.8/10) and competence (6.5/10).
• Soft Republicans showed strongest affinity for competence and institutional trustworthiness (6.6/10 and 6.3/10) and lowest affinity for healthcare (4.8/10).

The heterogeneity was substantial enough to warrant distinct message tracks — but the practical question was how many tracks were feasible given the campaign's production budget and Rafael's communications team capacity.

The Compromise: Three Tracks

After two weeks of negotiation between the analytics and communications teams, they settled on three message tracks for direct mail and two slightly different versions for digital.

Track A — "The Record": Emphasizing Garza's competence and accountability record, with economic and institutional focus. Targeted at unaffiliated voters and soft Republicans in the persuasion universe — voters with relatively low healthcare affinity but higher economic accountability and competence affinity. Issue affinity criteria: economic accountability score ≥ 6.5, healthcare score ≤ 5.5.

Track B — "Healthcare and Schools": Emphasizing healthcare stability and education investment. Targeted at persuasion-range Democrats and unaffiliated voters with strong domestic policy affinity. Issue affinity criteria: healthcare score ≥ 6.5 or education score ≥ 6.5.

Track C — "Institutional Trust": A more explicitly contrast-oriented message emphasizing Garza's professional record and character relative to Whitfield's populist rhetoric. Targeted at soft Republicans who showed highest values on institutional trustworthiness — voters whose skepticism about Whitfield's style was the primary entry point for the Garza message.

The digital versions of Tracks A and C were collapsed into a single "competence and accountability" digital creative that the campaign could run against a broad match on the persuasion universe's Facebook and Instagram. Track B had a distinct digital version targeting Facebook audiences matched to the higher-healthcare-affinity voters.

Channel Assignments

The channel targeting logic followed from the persuasion universe profile and the evidence on contact effectiveness.

Direct mail: Used for all three tracks, with specific pieces targeted by track assignment. Mail was prioritized for voters aged 40 and above with reliable registration addresses and at least three election participations in the last six cycles — the cohort most likely to actually read physical mail.

Facebook/Instagram: Used for all tracks but with age targeting (25–55) and frequency capping to prevent oversaturation. The digital team linked the advertising to voter file match through a custom audiences upload.

Personal canvassing: Reserved for the highest-priority segment within the persuasion universe — voters with support scores between 48 and 58 (the genuinely competitive range, as opposed to those who were in the persuasion universe only because of statistical noise near the cutoffs) who lived in the four primary suburban counties. Approximately 28,000 voters met this criteria and were assigned to the personal canvass list.

Phone banking: Used for a broader segment of the persuasion universe in areas where canvassing density made personal visits difficult — primarily the more dispersed portion of the suburban counties and the smaller-city persuasion targets.

Results and Adjustment

Three weeks into the persuasion program, Nadia had enough incoming canvass survey responses to check whether the message matrix was working as intended. Canvassers were asking a standardized set of issue-priority questions and recording responses. She cross-tabulated those responses against the track assignments to check whether voters receiving Track A were indeed expressing the issue priorities that justified the track assignment.

The results were partially confirming and partially not. Track A (Record and Accountability) was performing well among the unaffiliated voters it was designed for. Among soft Republicans, however, the canvass responses suggested that the economic accountability frame in Track A was not resonating as expected — this group was responding more positively to the competence-and-character frame from Track C than to the economic regulation frame the campaign had assumed they'd care about.

Nadia recommended, and the communications team agreed to, a two-week adjustment: soft Republicans who had been in Track A were migrated to Track C for their final mail piece. The digital program was adjusted to reduce Track A creative frequency in the soft Republican segment and increase Track C.

Discussion Questions

1. Nadia uses cross-tabulated issue affinity scores to make the case for distinct message tracks. What are the advantages of this data-driven approach to message strategy over the communications team's preference for a single unified message?

2. The case describes a gap between the assumed issue priorities (from the targeting model) and the revealed priorities (from canvass survey responses) for soft Republican voters. What does this gap suggest about the limitations of issue affinity models built from consumer data?

3. Why was personal canvassing reserved for the narrowest band of the persuasion universe (support scores 48–58) rather than the entire 38–62 range? Is this the right cutoff? What are the tradeoffs?

4. The campaign settled on three message tracks as a compromise between analytical sophistication (which might suggest five or six) and production feasibility. How would you think about the optimal number of message tracks given a campaign's resources and time constraints?

5. Imagine you are the Whitfield campaign's targeting director. You know from your own model that Garza is running a persuasion program in the suburban counties. How would you respond? What targeting adjustments would you make in the final three weeks?