Evolved Targeting: Artificial Intelligence and The Birth of Smart MCM

Brad Davidson, Phd., Medical Anthropologist and Senior Strategy Consultant

You know you’re getting old when what was once science fiction is now coming up in 2017 brand planning discussions.

There’s been a lot of chatter around the concept of “Artificial Intelligence,” or AI, in the news recently—and also among pharma marketers, in terms of the role AI may play in healthcare. Like very few other technical, hard science concepts, AI has captured the imagination of the general public, sparking robust discussions around the ethics, impending virtues, and potential disasters of “machines that can think independently.”

But what is AI, exactly? And how can we, as pharmaceutical marketers, harness it meaningfully and appropriately…today? Computers that think The term Artificial Intelligence originally meant something like “a complex machine that could think.” In short, a created human brain, one that could learn, problem solve, grow, and (heaven forbid) gain awareness of its own existence.

That’s why it makes for such good t.v.

But today, that definition and expectation of AI has evolved to something more practical. What we have come to focus on are “discrete problem solving machines that can learn from their own experience”—a machine that can play chess, and get better at it, or a set of algorithms that can predict with increasing accuracy purchasing behaviors, based on previous predictions, which were in turn based on lots of input data (that a human first assembled).

Now it’s starting to sound like something less like an existential threat to humanity and more like something that could reasonably support multichannel marketing efforts.

Adding the science of algorithms to the art of marketing

Of course, computers work in a very specific way: any computational solution requires data as input, a series of algorithms (fancy word for computer code) to analyze that data, and some kind of output, whether that be a chess move, a diagnosis, or a set of ads sent to an individual who seems on the verge of purchasing more camping gear (don’t ask). The algorithms that work on one set of data rarely, if ever, work without editing on another set of data, to solve a different problem.

If AI has come to mean a “problem solving machine that can learn as it goes within a very specific context,” then we have a ready-made use for AI in creating integrated multichannel engagement plans.

Our ability to meaningfully connect the various touchpoints we have at our disposal with which to communicate to our audiences (face to face, virtual, non-personal, passive digital like websites, etc.) has somewhat eroded, because there are just so many to choose from, and so our multichannel plans have become an exercise in educated guessing (at least when compared to the purchasing algorithms used by, say, Amazon). They needn’t be.

Applying intelligence – artificial and real—to multichannel engagement strategies

Our task, when applying AI to creating integrated channel plans, is actually relatively simple, conceptually speaking: we only need to create a set of algorithms that will take the input we provide (in this case, channel usage, observed impact of usage, typical patterns of interaction between the channels, meaning, what interactions with what channels lead to further interactions with different channels…), and use the output (a series of findings about what channels pair together best for most measurable impact) to guide our channel plans.

The subtle variations within the question are nearly infinite: do primary care physicians use the same channels, in the same way and in the same order, as specialists? Does this pattern change depending on the category, say, CNS versus diabetes? Does the number of available therapies change our basic algorithm? Does the modality, or cost, or side effect profile of the drug or intervention change how the healthcare provider learns about or seeks for answers about the therapy itself? The list, potentially, is long.

But as a first step, we know what we want to do. We have the data and we have the desired output (a data driven, targeted multichannel engagement plan) to help our newly minted algorithms produce usable solutions that get better at creating impact the more iterations we create and study.

Key takeaway: Aptus Health has evolved to the point where this first step is a reality, and generating results. So while we will all be watching to see how elements of AI such as augmented reality and advanced analytics transform how health and life sciences companies connect with HCPs and consumers, the future, in the form of an AI-driven channel planning function, is already here.

  1. ← Previous
  2. Next →