How hard is it to spot an emerging threat or opportunity in time to actually do something about it? Is it as hard as spotting a scud missile in the deserts of Iraq? As hard as identifying an underwater threat to a submarine using only sonar? As hard as spotting a consumer trend in a vast and complex business like financial services?

Turns out, analytic techniques developed to help meet exactly these three challenges (and others) are migrating to the pharmaceutical industry. They may not be common yet, but several pharma companies are already employing these techniques, which identify useful patterns in highly complex data, to pursue some important research goals:

• Early detection of behavioral shifts, alerting marketers to changes well before they happen, as opposed to alerts that detect change after the fact
• Identification of interactions that lead to changes in behavior, by coming up with metrics that identify social networks and individual levels of influence
• Knowledge about the behavior of both individual customers and whole markets that can be layered into a broad range of sales and marketing initiatives to enhance outcomes.

Although pharma has only recently begun to use advanced analytical tools, other industries have been turning to them for over a decade, to cut costs and enhance customer relationships. In the consumer banking industry, for instance, these tools helped marketers allocate resources with greater precision by focusing on "segmentations of one," which helped bring about a revolution in the consumer banking industry a decade ago.

Kelly D. Myers

The Learning Curve

In the pharmaceutical industry, advanced analytical tools are particularly valuable in identifying key patterns in complex data flows. They help companies figure out how to deliver the right message to the right customer, at a time when the customer is likely to be most receptive. Analytics can be used to detect shifts in market and customer behavior patterns by using "learning" algorithms to identify problems, and then adapt and correct the situation early in the cycle.

The analysis is a two-step process. First, networks of physicians who are likely to interact with one another on a regular basis are identified. Then, changes in practice patterns are quantitatively captured along with the direction of influence.

Using data instead of opinion to quantify influence enables marketers to navigate areas where ordinarily there are blind spots. The quantitative approach ascertains who is influencing whom and confirms it with data. Traditional key opinion leader influence mapping identifies thought leaders, but quantitative influence mapping identifies networks operating on regional and local levels that were not previously perceived as influential.

At Esprit Pharma, Brent Herspiegel, director of marketing for Estrasorb, a topical estrogen treatment for menopausal hot flashes, is responsible for growing the brand in a crowded hormone therapy market, with a relatively small sales force of 175 reps. Using advanced analytics, Herspiegel's team measured the influence of customers within the entire therapeutic class, and identified clusters of physicians who regularly affect the behavior of one another. With influence metrics quantified and clusters identified, this knowledge was then layered on top of the company's speaker program.