Prior to DeepIntent, most of my career was dedicated to the buy side. Having worked across many pharma brands at various media agencies, I’m all too familiar with the frustrations that media strategists and buyers face when trying to reach campaign-specific patient audiences. Simply put, 3rd party data usage was severely limiting, and patient modeling options were severely lacking. Imagine my excitement when we concepted, and then launched, a patient targeting solution for these challenges (and more). Here are the three things you should know about our reinvention, Patient Modeled Audiences:
1. You get precise, campaign-specific segments.
Patient audience models are not one-size-fits-all. Brand and campaign strategy play an important role in identifying your ideal patient audience. For example, let’s say you want to raise awareness of a new drug amongst two types of patients: those who are (1) newly diagnosed, and (2) currently being treated with a competitive drug. Before Patient Modeled Audiences, you’d be limited to “off-the-shelf” models, that would be unsuccessful in differentiating between these patient types. With Patient Modeled Audiences, you can create and target two distinct audience segments, based on a variety of data points and factors that are suited to your specific strategy and objectives. In fact, our modeled audiences are completely unique; they’re never recreated or reactivated.
2. It’s fully transparent.
Having been in your shoes, I know what you want to know about your audiences. So we give it to you. Advertisers have full visibility into model inputs, scores, and the audience selection process.
- The number of unique patients we’re modeling against. Once your strategy is implemented, our team provides the count of unique patients based on the campaign-specific claim codes used to direct the modeling.
- Audience scores and quality. Once the model is exported, we score all users in our platform to measure clinical relevance and direct segmentation. We share and discuss this information with you, and collectively select an effective threshold based on your KPIs.
For example, if your primary KPI is Audience Quality, we’ll select a high threshold (e.g. users who are 75% or more likely to be diagnosed with your condition). If your primary KPI is unique reach, it may make sense to lower the threshold so you reach more potential patients.
3. Audiences can be updated in-flight.
We collaborate with our clients to review in-flight performance and 3rd party measurement results, and optimize models accordingly, in-flight. For example, let’s say your campaign is live, and you’re reporting DeepIntent as #1 in audience quality, but #3 in unique reach. You have the option to rebalance your delivery by lowering the threshold (e.g. from 75% to 70%) to reach more unique users while sustaining acceptable audience quality. For these reasons—and more—Patient Modeled Audiences consistently outperforms competitive solutions. So if you’re still facing age-old frustrations, I suggest you give Patient Modeled Audiences a look. I’d be glad (and excited) to discuss it with you myself.
