About 100 years ago, John Wanamaker, a pioneer of American retail, is supposed to have made the comment, “Half the money I spend on advertising is wasted; the trouble is, I don’t know which half.” More recently, Google’s Eric Schmidt made the infamous comment, “Corporate marketing is the last bastion of unaccountable spending in corporate America.” The core issue of accountability for success (or the lack thereof) of advertising campaigns and marketing programs is one that continues to linger well into the digital age despite the inherent traceability built into digital media.
Healthcare advertising and marketing offers a unique set of challenges and opportunities on the path to modeling accountability. The challenges largely arise from the fact that every healthcare decision has multiple stakeholders, a.k.a. customers who influence it. The customer who uses the product (the patient) isn’t the customer who chooses it (the doctor), or who typically pays for it (insurance companies or the government). The opportunities too are unique—the core selling process depends on scientific claims (derived from clinical trials) and the ability to shape opinion and therefore choice is rooted in these validated claims. These multiple layers make it hard to drive accountability of individual healthcare marketing initiatives, and modeling the probability of success harder still.
User-initiated Actions Are Better Indicators of Desired Outcomes Than View and Clicks
There is, however, a halfway point between rigorous control settings to establish program ROI and spewing tactic by tactic usage metrics, and it relies on measuring actions that require user initiative and then calibrating them against the probability of a successful outcome (as defined by the brand)—the hypothesis being that actions requiring initiative are better indicators of “brand engagement” than passive metrics like views and clicks. These could be actions like registering for a newsletter, making a call to a nurse, downloading a financial support form or a copay card, etc. Or a measurable shift in mindset as a result of exposure to an educational asset captured through a poll or a quiz.
S.C.O.R.E.: Predictive Modeling to Project Brand Impact
S.C.O.R.E. is an acronym that clusters user-initiated actions and attitude shifts that are measurable across channels: Subscriptions to brand-supported assets such as a newsletter; Conversion-oriented actions such as copy card downloads; Outcomes presentations such as longitudinal case studies or real-world patient reported outcomes; Response measures that include use of social features such as send to friend; and Educate, a measurable state defined by a shift in understanding. A SCOREboard delivers a view of active customer engagement, not just interaction.
Taking the next step towards ROI means layering this methodology with a probability rating for each user-initiated action that projects what percent these actions will result in the desired outcome (be it product conversion or a shift in mindset). Let’s say the probability rating for a co-pay card download is 20% (meaning 2 of 10 downloads result in redemption). This tells us that for every five thousand downloads with a 20% probability of conversion, we’ll convert a thousand customers. This is easily translated into dollars based on lifetime value of customer benchmarks.
The key is to get the probability ranking right. In some cases, there’s the ability to calibrate on an ongoing basis by an analysis of customer-specific engagement metrics and sales. In others, it has to be modeled based on control groups. While probability modeling adds a layer of complexity, the value in connecting the dots between actions and projected brand impact doesn’t just stop with ROI for digital initiatives, but also applies to traditional marketing that typically lack traceability.
By Prodeep Bose
Prodeep Bose is SVP, director of multichannel strategy at The CementBloc. Over the past decade, he has led digital marketing groups at Digitas and Ogilvy.
Courtesy of MediaPost