February 19, 2018
February 19, 2018
Healthcare brands have two primary customers: those who deliver care and those who seek it out. Reaching them effectively through digital channels calls for effective use of data, AI, and machine learning, and that's PulsePoint's area of expertise.
I spoke with SVP and general manager of Pulsepoint's health division, Chris Neuner, about how first-party data is combined with AI to uncover insights about the target market for the health industry. That includes both physicians and patients.
Neuner says, they see “significant engagement based on first party data,” both from physicians and patients. He offered the example of clicking through to diabetes information on a brand's website. The brand can then capture that data to identify its target audience. Segmenting the audience on the basis of the indicated interest then allows the brand to reach the physician or patient “wherever they might be online,” Neuner said. Now pharmaceutical companies can tailor their messages to individual interests indicated by their online behavior.
When doctors enter their National Provider Identifier (NPI) on sites, it leaves a deterministic digital trail of their web behavior. Brands can then “marry that” with the brand's first party data and craft their content and messaging accordingly.
The challenge with working off data, Neuner pointed out is that it's not always clear what we're going to find. That's where AI come in to identify “those signals and patterns” that indicate who would be receptive to what type of information. It also plays a role in targeting without violating HIPAA privacy rules, since you are able to target based on identified patterns of behavior instead of on an individual basis.
While physicians may be targeted with information on any condition of interest, under HIPAA, patients cannot be targeted without consent with information that comes through a healthcare facility, or a condition that is included among those classified as “sensitive.”
One way that brands can address the challenge of restricted information is through “lookalike modelling” he explained. They can “look at aggregated browsing behaviors of a group” rather than the individual and then “let algorithms detect patterns and combine first-party data with segmentation and machine learning” to identify audiences that can be targeted “at scale, and where to find them.”
This kind of approach is particularly helpful for marketing for clinical trials, which, Neuner pointed out, amounts to $8 billion a year. AI can identify a diabetes patient who was looking at information that may indicate being open to information on new treatments.
The second thing that AI can provide is precision, location-based targeting based on geolocation data. It can then pair signs of interest and proximity to a relevant clinical trial to identify potential participants and “drive better outcomes for them.”
As the healthcare industry does require strong awareness of what information can or cannot be used , “PulsePoint recently launched the Programmatic Health Council to help clarify programmatic advertising for healthcare marketers. It functions as a kind of “self-regulatory group” made up of about 30 healthcare advertising experts in the US and UK, Neuner said.
Such a group is needed, he explained, because currently “technology is ahead of the regulations out there.” Consequently, there is a need to get the industry to agree to “a safe, compliant environment that produces optimal results for consumers and brands.”