PE: How can market research reasonably expect to build accurate metrics that forecast the uptake of a certain innovation?

SHARMA: Statistical controls should always be incorporated in the analytical package to guard against the tendency to overstate potential demand for a product. Survey respondents often try to please the research sponsor with the "right" answer, and there is little awareness of how inertia can slow real-life uptake of a new drug. In addition, we have to account for the fact that highly innovative products are subject to a failure of imagination about how the technology might transform medical practice. This means that raw projections might actually understate ultimate demand. Lilly's initial forecast numbers for Prozac prescriptions never encompassed the impact that enhanced tolerability and safety would have on the market for antidepressants, while Pfizer's analysis failed to take into account how interest in Viagra would be fueled by the health information revolution fostered by the Internet.

PE: So do numbers really count as a benchmark for evaluating the prospective response to a new therapy?

MCDONALD: Both of those cases [Prozac and Viagra] demonstrate the importance of using market research techniques that tap a vision, rather than simply tally votes. Even when statistics are required to advance the proposition, emphasis should be placed less on large numbers—the reliability test—than on developing a product proposition or scenario that the survey population can meaningfully evaluate. A validity test. In other words, selecting the right targets in survey research is essential to ensuring that the proposition is applicable to a commercial strategy. For example, key opinion leaders are often the best "prophets" of change, and can tell us more about the prospects for an innovation than would a survey of community physicians. But when seeking to identify barriers to the uptake of a new medicine, the community physician's perspective might be more relevant due to their tendency to resist the "re-engineering" of clinical paradigms.

In addition, careful targeting of a survey audience should not be limited to products promising a dramatic technical advance; it applies also to situations where deeply ingrained customer thinking must be adjusted only a few degrees to render distinctions where none have been made. So market research has to first establish how best to tell a story, and determine to whom that story should be told, deferring costly quantitative research until there is confidence that customers are in the right frame of mind to validate a path forward.

PE: A standard tool of market research is the product positioning study, designed to help establish a clear therapeutic footprint and drive acceptance by customers. Is this approach still relevant in ensuring success in an increasingly crowded marketplace?

SHARMA: Positioning research is applied to a variety of uses, and can be described as a tool to inform the marketing process, to identify a product attribute or benefit, or assess a marketing outcome. Learning what we need to know about product positioning typically begins when the product is in early gestation, where we start to establish what outcomes or characteristics are necessary to ensure success with the FDA and with customers. It is based on a program of research that allows us to understand how competitors are viewed, what space is available in the marketplace, how well our product meets unmet needs, and how differentiated from competitors our customers believe it to be. The culmination is the positioning study, which tests one or more positioning statements with different customer constituencies. Our experience is that such work is a waste of money and resources. It is also potentially dangerous, in that it limits potential market opportunities by framing positioning as a vote on various slogans. It risks inviting customers to vote on the wrong thing—turns of phrase—at the wrong moment—before those phrases have been vetted by regulatory staff or set in a creative context by your brand agency.