HealthRankings datasets enable programmatic and television campaign targeting of condition-based audiences that are fully HIPAA-compliant and de-identified, available off-the-shelf and as customized audiences, and can be activated through your chosen media channel. 

We help clients answer:

  • How do I find my desired audience?
  • How do I activate or target them?
  • How do I do this in a cost effective and timely manner?
  • Who is likely to respond and take the desired action?
  • Is the campaign reaching the intended target audience?

Integrated Dataverse

Our health datasets come from one of the largest integrated, curated, and de-identified health data repositories – IDV®

The process of building digital, CTV, and TV activation audiences:

  • Define the population of interest
  • Profile them with attributes available in IDV®
  • Build a model that scores all patients in IDV®
  • Identify caregivers as living in the same household as patients
View our data warehouse

Precision and Scale

Our approach to propensity scoring allows for both precision and scale and addresses the concerns of marketers who have not seen the best results from some of the generalised approaches in market today.

Propensity-Based Scoring

Through healthcare utilisation data, and via machine learning and statistical modelling techniques, we have grouped individuals based on their propensity to have certain health characteristics.

100% Anonymisation

All scoring adheres to the strict requirements of HIPAA. Scoring is conducted on anonymised audiences only. Population sizes are large enough such that there is no risk of re-identification.

Behaviour Focus

All our scoring is conducted in accordance with NAI Code of Conduct and HIPAA. We focus on all therapeutic categories and strive towards understanding the drivers of behaviour.

2x Click-Through Rates

Some Symphony Health segments received more than twice the benchmark Click-Through Rates and higher number of clicks despite serving less than 50% impressions, compared to a competitor model.