When the Patient Pool Is Tiny, Every Targeting Decision Counts: How Precision Audience Modeling Transformed Rare Disease Patient Acquisition

The Challenge

Rare-disease patients often remain invisible to standard audience models, making broad targeting wasteful. An oncology brand with stalled adoption needed behavior-based targeting to reach qualified patients and drive new starts.

The Strategic Solution

• Patient-Likelihood Modeling: Identified qualified prospects through behavioral, clinical, and proprietary audience signals—not disease-state targeting alone.
• Crossix Data Integration: Used Crossix Prime and Reach to find high-probability candidates based on diagnostic, clinical, and prescription patterns.
• Behavioral Audience Proxies: Combined patient personas and high-indexing consumer signals to reach audiences standard health models miss.
• Machine Learning Segmentation: Scored audiences by adoption propensity, concentrating spend on high-value prospects.
• Geographic Precision: Matched HCP targets to patient ZIP codes to prioritize markets with the strongest prescribing opportunity.
• Channel Optimization: Shifted investment into precision programmatic and continuously optimized against real-time audience-quality signals.

Results and Learnings

• 41% increase in new patient starts year-over-year — reversing stagnant adoption with no increase in overall budget
• 70% of the qualified patient audience reached within just 2 weeks of campaign launch — compressing timelines that traditionally take months
• 73% lower CPM and 81% lower cost-per-click versus property-based targeting, proving that precision reaches further for less
• 877% lift in brand website visits among targeted patients versus control group
• 34% of all conversions driven by programmatic — at a fraction of the cost of other channels
• 120% reduction in new patient acquisition cost — transforming the program’s ROI profile and unlocking reinvestment into expanded reach

Relevance to Client

For niche and rare-disease brands, audience strategy is a science challenge. Combining clinical intelligence, behavioral signals, and machine learning identifies high-propensity patients standard models miss—expanding reach, improving acquisition efficiency, and maximizing every media dollar.