Custom Patient Targeting is a new cookieless way to reach your patients now and in the future. Powered by Dstillery’s patented ID-free™ technology, it doesn’t rely on any form of user-based targeting, ensuring 100% compliance with all laws, policies, and guidelines. Read on to learn more about how Custom Patient Targeting was created and why.

What Problem is Custom Patient Targeting Designed to Solve?

Healthcare brands want to achieve more precise direct-to-consumer targeting but face strict data requirements. The industry grapples with rules from HIPAA, NAI, and DSPs and limitations on first-party data, and strict demographic requirements. These requirements have meant fewer, more narrow options in the marketplace, and have ultimately presented a precision problem. Brands are wasting time, money, impressions, and conversions by targeting people who aren’t their prospective patients. Less waste and more precision are needed. Enter: Custom Patient Targeting.

How Did Custom Patient Targeting Evolve From ID-free?

Dstillery was already in-market with ID-free Custom AI™, also powered by our ID-free tech, when we saw this problem that healthcare brands face. We knew that our ID-free technology could be part of the solution. Custom Patient Targeting (CPT) works the same way as ID-free but is specifically designed to address the needs of healthcare brands.

Like ID-free, we strategically built CPT with data minimization at its core. It only relies on a couple of data points, notably, a seed set with people who search for specific health terms such as product names, symptoms, or comorbidities. As Crossix and IQVIA results come in, our data science team will partner with you to optimize and tune your custom model to drive optimal patient outcomes.

This is incredibly valuable, but not enough to drive scale, performance, and understanding of your patients’ intent across their digital journey. That’s where our AI comes in. Using a continuous influx of patient journeys, our AI learns how, when, and from where people browse the web, finding patterns and applying the signals to all other websites.

This results in a just-for-your-condition behavioral intent score for every website, answering the question, “When someone visits this site, how likely are they to be interested in your brand’s message?” The intent scores give your brand the strategic advantage to predict the ad impressions most likely to convert for your brand. When we see a bid request that’s predicted to be your patient — such as someone visiting lowcarb-diet.com at 9pm, or someone from Austin visiting startsat60.com — we will bid and purchase that impression.

Custom Patient Targeting Key Differentiators

There are 3 key unique and differentiated enhancements that we’ve developed to make Custom Patient Targeting more aligned with what healthcare brands are specifically looking for.

  1. Healthcare search keywords as seeds: While first-party data is preferred, a lot of healthcare brands do not have access to (or are unwilling to share) this data with data vendors. By leveraging our opted-in panel dataset, we are able to use healthcare search keywords — ranging from product names, competitive product names, symptoms, and co-morbidities — as seeds for our ID-free models. Testing has shown that well-selected keywords can drive programmatic performance on par with first-party data.
  2. Model tuning for healthcare performance: Unique to healthcare advertisers, there is a need to not just drive growth in programmatic metrics (ROAS, CPA, CTR, etc.), but also drive high audience quality scores measured via vendors like Crossix and IQVIA. If audience quality results are shared with us, our data science team will partner with healthcare advertisers to optimize and tune the custom model to drive optimal patient outcomes like improved audience quality scores.
  3. (In development) ICD-10 Codes as seeds: We are in active discussions with leading RWD (Real-world data) providers rooted in healthcare data such as claims data, patient surveys, and more to use their data as seeds for our models. This would enable customers to simply provide ICD-10 codes or conditions they are trying to target, and for us to use this dataset to identify patients with the condition.

To learn more about Custom Patient Targeting, click here.