Katie Carr, SVP of Sales at Swoop, explains how pharma brands can leverage condition-specific patient segments to connect with more relevant consumer audiences online. This episode delves into the building blocks of custom audiences — real-world health data (RWD), in conjunction with artificial intelligence (AI) and machine learning (ML) — and how custom segments provide a better way than traditional targeting methodologies for brands to drive script lift. 


Real-world health data in conjunction with artificial intelligence and machine learning offers an unprecedented opportunity for pharmaceutical marketing to drive script lift. It allows marketers to connect with more relevant consumer audiences through condition-specific patient segments.

During a sponsored MM+M podcast, editor-at-large Marc Iskowitz sat down with Katie Carr, SVP of sales at Swoop, to discuss the limits of traditional targeting methodologies and the potential of custom audience alternatives for modern healthcare marketing.

Carr kicked off the discussion by explaining that Swoop offers a number of different solutions for brand marketers today. 

“What Swoop really understands is that every marketer, every pharmaceutical advertiser has its own unique challenges and therapeutic advantages,” she said. “And Swoop creates customized segments to meet those specific needs of the brands.” She went on to note that is essentially why the company is trusted by the top 42 out of 50 pharmaceutical manufacturers and top 18 out of 20 pharmaceutical agencies.

Moving toward real-data solutions

Traditionally, marketing solutions for pharma advertisers have included “demographic targeting, zip nine targeting, off-the-shelf segments and behavioral targeting such as clickstream,” Carr explained. However, these methods “are just not targeted enough,” she said, resulting in “a lot of waste for your media dollar when you’re not able to identify the specific target market, the specific patient that the advertiser would like to reach.”

The reason is that “demographic is essentially the entire population against criteria such as age, gender, ethnicity, household income” but “there is no health data within demographic targeting,” Carr noted. While “off-the-shelf” segments use real-world data, it’s generic and if you have “an oncology client that’s interested in finding patients for second- or third-line treatment, you can’t do that with a generic segment,” she explained. As far as behavioral targeting, “the data that’s collected online is all attached to an ID” and “a lot of these solutions are not going to be available for advertisers” when third-party cookies are disabled on sites such as Google or as “other platforms make privacy updates” like Apple has to iOS 14.5.

This is why brands need to “maximize their dollars across real-world data solutions,” said Carr, through precision healthcare marketing. Swoop, for example, creates “custom segments” to “hone in on the specific target market.” The real-world data derives from “a combination of claims data and social determinants of health,” she explained. Information such as “a diagnosis code, hospital visit, treatment and insurance notes or the NPI associated with the claim,” is accessed through claims, and social determinants of health provide insight into questions such as “can the patient access clean water or transportation in order to get healthcare?”

Building a custom audience

By way of example, Swoop has “over 300 million de-identified unique patient journeys in the U.S.,” dating back more than a decade, explained Carr. The data is updated on a weekly basis offering pharma companies the opportunity to reach “fresh, recently diagnosed patients.” Swoop then curates that ”data in conjunction with artificial intelligence machine learning to create privacy safe segments.”

To build segments, Swoop uses HIPAA-compliant privacy-by-design data architecture and works “within guidelines that have been created by the Network Advertising Initiative (NAI),” Carr said. In addition, Swoop has “a proprietary methodology that allows us to create a very high Audience Quality,” which drives engagement and script lift. In the case of rare diseases, she added, “the Audience Quality is going to be much lower but will perform better than other solutions in the market,” because the audience will have a higher multiple of patients in it than the prevalence in the general population.

Beware of any vendor “supplying you with segments that are over 50%,” Carr cautioned. Ask about its HIPAA certification, if it is a member of NAI and “the likelihood of a patient potentially being re-identified” because that would “put any manufacturer or agency at risk,” she said.

At Swoop, “we always put privacy first,” Carr noted. That includes the “privacy of the consumer but also being able to market to the consumer so that you can make their quality of life better” by “educating them on the treatment that’s out there and on the condition they could potentially have.”

Using data in advertising

Once the custom audience is built, it can be “activated across your entire media mix,” said Carr, including connected TV (CTV), addressable TV and linear TV. “As dollars begin to shift from traditional linear over to CTV, we begin to see those budgets increase,” she added. “It’s more important to have really strong targeting as we start to see those shifts happen.”

As long as advertisers are working with good data that perform well, being able to use “the same data asset for planning, activation and measurement and having that all aligned is really the dream state,” Carr said. Swoop will sit down with a pharma company “to understand what your brand needs are” and “how you want to measure the success of the campaign.”

Since the pandemic, media consumption habits have changed and the pharma industry accelerated its digital transformation. “Pharma has been known to be late to adapt,” she said. “Some brands are still stuck on measuring campaigns on click-through rates and cost per click.”

Rather than rely on “all sorts of vanity metrics,” Carr said, “we recommend that every single brand use a third party to validate its data” and “measure script lift for these campaigns.” At Swoop, she concluded, “we don’t grade our own homework.”