White Paper
Digital Tools to Map COVID-19 Spread and Deliver Interventions
Published March 23, 2020 [Updated: April 4, 2020]
Overview
Over the past decade, Public Democracy, Inc. (a certified Public Benefit Corporation) has developed a proprietary psychometric database of over 100 million Americans with insights into the core values that motivate people to engage in the real world and join together in common purpose. In recent years, we have applied this data (and our decade-plus experience as a top communications firm) to build best-in-class behavioral models and empathy-driven machine learning systems. These systems operate in a PII/PHI-free environment – instead of tracking identifiable individuals, our system uses human and machine learning to adaptively analyze patterns of behavior.
Those patterns then allow us to identify individuals and groups by their behavior and intent. We can stratify and filter those behavioral segments using the hundreds of thousands of additional identifiers available through the commercial data market and our proprietary and iterative human-companion-in-the-loop process.
Over the weekend of March 13, our early COVID-19 models successfully identified localized outbreaks prior to any media or public health reports about a first case in numerous locations (Austin, TX; Blaine County, ID; Lucas County, OH; Santa Barbara, CA; Sonoma County, CA; and San Luis Obispo, CA). In another location with only a few confirmed cases that our model flagged as a hot spot (Kentucky), 10 new cases were reported in the following 48 hours, with over 100 cases over the following week. A few weeks later, we identified loss of smell and loss of taste as significant COVID symptoms before any major public mentions or medical consensus.
A Unique Solution
Our process utilizes a great deal of public and commercially-available data, and we also use a suite of non-proprietary enterprise-level marketing and adware tools. But our iterative process—comprising human insight, our proprietary Values Data™, and our empathy-driven machine learning techniques—is cutting edge in the marketplace. The Office of National Drug Control Policy determined that our unique ability to map risk and engage individuals with SUD justified their seeking legal authority for sole source contracts in support of our technology. Google also sought us out as a partner, providing free advertising and unlocking all the blocked keywords associated with SUD to improve our behavioral models. And our ability to track COVID-19 spread and identify contributing factors and related symptoms was recognized by the Atlantic Council, the Ad Council, the creators of the Johns Hopkins COVID-19 map, and most notably by our inclusion in the White House COVID-19 Data and Research Task Force.
In short, there are numerous data generation and analysis tools in the market, but our proprietary process, tradecraft, and ability to incorporate our best-in-class Values Data™ make our offering and capabilities uniquely different from others.