Cover Story: Invincible Marketing
"This was brand new for everybody," Dant says. "It was sort of uncharted territory, because it was a brand new program."
Lawmakers passing the Affordable Care Act in 2010 wanted the nation's uninsured residents—48 million in 2012, according to the U.S. Census Bureau (opens as a PDF)—to become insured.
So, by definition, EmblemHealth couldn't create lookalike predictive models from an existing database of customers. But the goal was clear: to acquire customers.
To do so, EmblemHealth had to create a database, gather consumer data, analyze it, segment it and employ it across channels, including digital.
EmblemHealth and Merkle gathered data from a government survey in which respondents stated whether they would sign up for insurance on the state exchange. (New York State opted in and created the New York Health Benefit Exchange.)
"We knew, for example, … in any given ZIP code, what proportion of people living in that ZIP code qualify for an ACA product because of their age and because of their income bracket," Dant says. "Then we used the data source, Merkle's data source, to identify people that met those age and income criteria. There were some other variables also that we looked at, but those were the primary ones. That is how we developed our model for deciding who we should be targeting our direct mail to."
So segmenting more than 8 million New Yorkers primarily based on age, income and where they lived, EmblemHealth and Merkle whittled the list down to 1 million. Then, Dant says, more than half of the 1 million remaining were Medicaid members. Ultimately, 370,000 names matched the qualified health plan prospect pool guidelines.
EmblemHealth had its goal: Ask 370,000 New Yorkers to pick its insurance plans.
Despite the types of targets being completely new to EmblemHealth, Dant says the company had an idea of whom it wanted to insure.