Climbing Toward Peak Sales


By Scott Setrakian, Co-Founder and Managing Director, APT


According to the Tufts Center for the Study of Drug Development, the average new drug that gains marketing approval costs over $1 billion to develop. Yet, according to Bain & Company and EvaluatePharma, half of all launches fail to achieve expected peak sales. With these astronomical investments and less than stellar success rates, it is no surprise that every life sciences company and its shareholders are looking for new ways to improve commercial outcomes. So how can life sciences companies set themselves up for more successful launches? They can begin by supplementing traditional pre-launch market research with empirical data based on how physicians and patients actually respond to commercial strategies.

Today, pre-launch market research focuses on driving an accurate expectation of physician and patient behavior upon release of the new drug. Brand teams are challenged to predict future behavior based on direct customer interviews, statistical measurement tools, such as conjoint analysis, and other market research strategies designed to identify and segment the market, understand the barriers and drivers of adoptions, and more.

Such upfront research has been used for decades across a wide variety of industries, from consumer products to financial services to hospitality to retail. It provides a rich set of indicators that can help guide first steps, but practitioners have learned the hard way that up front research tools are not highly accurate when it comes to predicting true customer behavior. There are many reasons for this gap, all of which come back to a single truth: no matter how elegant and complex the research tool, it cannot truly simulate the real world in which the actual decision to buy or not buy is made. This research is useful but hypothetical.

Practitioners have learned the hard way that up-front research tools are not highly accurate when it comes to predicting true customer behavior.

Many reasons have been posed for differences between customer-self-predicted and actual behavior. One is that respondents to market research do not have accountability to their answers; there are no repercussions for answering however they feel at the time. A second, market research takes place in an environment that ignores the influence of other factors, such as media messaging, key opinion leaders, formulary, and more. A third, the theoretical presentation of a choice in the research path is often not a true reflection of how the alternative of making a decision at all develops. Physicians and patients simply do not always behave as they say they will once they are given the opportunity to make a real decision. Market research plays a valuable role, but an over-reliance on market research can lead to committing to a sub-optimal path post-launch.

Organizations can accurately isolate the cause-and-effect relationship between an action and its target KPIs (e.g. scripts, adherence, market share, etc.).

There is an incredibly useful step in the launch process that has been embraced by leaders in other industries – such as the above-mentioned consumer goods, financial services, etc. – over the last decade. Operators rapidly test launch decisions in real time, and adjust on the fly to pour resources into winning tactics and discard unsuccessful ones. This empirical approach is very familiar to life sciences companies, which are the pioneers of randomized control trials for the drug development process.

Organizations can also leverage this same method of scientific experimentation to understand the incremental impact of their commercial strategies: try the idea (e.g., changing sales detail strategy) in some situations and compare the test group (receiving a given “treatment”) to a highly similar control group, operating as normal. By comparing physicians, sales reps, or patients receiving a given action to a highly similar group that does not receive that action, organizations can accurately isolate the cause-and-effect relationship between that action and its target KPIs (e.g., scripts, adherence, market share, etc.).

Brand teams can use the insights generated through market research and other pre-launch analysis to build rapid experimentation into the launch plan and early stages of the product life cycle. Given the massive investments in bringing a drug to market, many leaders may think that they cannot afford to experiment, opting instead for an “all of the above” approach with significant investments across channels. However there is ample evidence that the opposite is true – organizations can receive a handsome payoff by using experimentation to generate empirically credible data and insights that inform optimal go-forward strategies. With the right analytic methods and technologies, organizations can design low-risk, rapid experiments in which executives will be confident.

The use of testing to optimize launch activities is a natural extension of the current adoption of this approach by leading life sciences organizations in other stages of the product lifecycle. Over the last few years, several leading life sciences organizations have accelerated their use of in-market experimentation to optimize patient engagement programs, new sales roles, speaker programs, DTC marketing, sampling effectiveness, and more.


Speaker Programs: A brand team was spending millions of dollars annually on its speaker programs. However, since physicians self-select to attend an event, they tend to be very different than the average physician based on prescribing behavior and other key characteristics. As such, it was very difficult for the organization to understand the true incremental impact of the investment and determine how to improve ROI going forward. Using innovative control group techniques, the organization was able to find highly similar control physicians (who did not attend the events). With the right analytics in place, the organization found that, while test physicians (those attending the events) began prescribing the drug at measurably higher rate than control physicians, the speaker program overall was not paying back. However, when the organization dove into the details, they found that certain types of events – such as during specific times of the day – were much more effective. Further, they found specifically which physicians had the greatest likelihood of responding incrementally. Through a combination of optimizing the type of event, who they focused on inviting, and more, the organization identified an opportunity to more than triple ROI of the program versus what they would have otherwise continued to do. This outcome would not have been possible based on market research alone.

Tele-Detailing: For years, a leading pharmaceutical company had used only its own sales force to support its specialty drugs. However, cost concerns and shifting priorities made it difficult to continue allocating sales force investments to smaller brands. Executives knew that pulling investments from these drugs could damage profitability, but they were unsure which alternate channel would be best to support these drugs. Among the new channels executives evaluated was investing in a contract sales organization to tele-detail a specific segment of physicians. They had a prevailing hypothesis about which physicians would respond best, but did not have a credible way to validate that hypothesis. The organization turned to specialized software to compare the group of physicians that received a call to highly similar physicians that did not receive the call. This causal analysis revealed that, in aggregate, the investment only broke even. However, by understanding which physicians were most likely to change their behavior due to the call (based on test vs. control impact), the organization identified an opportunity to capture almost all of the revenue benefit with less than half of the initially-planned investment. This learning enabled the organization to reallocate the remaining dollars to other high-return opportunities.


What steps can your organization take today to supplement market research and build experimentation into the commercial process?

Use “natural tests” to hit the ground running: Natural tests are events that happened in some situations but not others and can be analyzed as a test to isolate the cause-and-effect relationship of that action on KPIs. For example, if a set of physicians are detailed two months before others, test vs. control analysis can reveal the incremental impact of those detailing investments.

Collect and prioritize ideas to design experiments: All of the investment in pre-launch market research generates great ideas for testing (e.g., it appears that a given message resonates well with physicians). These hypotheses should then be tested through small scale experiments to validate if the survey learnings hold true in the real world. Organizations should prioritize ideas for testing based on 1) whether they will generate a new learning for the organization, 2) will inform a near-term action, and 3) are tied to significant economic value.

Put the right analytics in place: With small sample sizes, different growth trends, and inherent differences between test and control physicians, patients, and markets, it is incredibly challenging to distill the signal of a business action from the noise inherent in the environment. As such, specialized techniques are necessary to find control groups that are as similar as possible to the test group, except for the action of interest. Further, every result should be associated with a specific level of statistical significance so that executives can be aware of what risk they take in rolling out an action more broadly.

Generate incremental value with tailoring and targeting: Many organizations stop their analysis after isolating the overall impact. However, as most programs work much better in some situations than others, organizations can generate substantial value by targeting actions to the right groups with the right investment level or message at the right time. Targeting models should always be based on incremental response, as opposed to who is most likely to respond overall (i.e., need to determine who will change their behavior due to the action as compared to what would have happened anyway).

Develop an organizational commitment to commercial experimentation: Market research is likely to remain a critical component of the commercial process. The most sophisticated organizations will view the resulting insights as hypotheses that should then be tested. As with all tests, however, a significant number of ideas will not work. Organizations that embrace commercial testing thus must also embrace a culture that rewards innovative ideas, even though many may not work as originally anticipated.

Rapid experimentation ultimately leads to step change improvements by enabling more successful commercial innovation with less risk. Organizations can quickly try new ideas, roll out the winners, discard the losers, and tailor and target implementation for maximum impact. As the healthcare environment changes, new marketing channels emerge, new drugs enter the market, formularies change, new patient tools become available, etc., testing is a powerful way to keep pace with change and innovate rapidly and successfully. •


Scott Setrakian, Managing Director of APT. Scott has over 20 years of experience advising many of the world’s largest and most successful companies. His work focuses on the development and implementation of strategies and processes that lead to significant improvements in profitability and corporate value. He has led the work of hundreds of management consultants in engagements in over 20 countries on six continents. Prior to co-founding APT, Scott sat on the Board of Directors of Oliver Wyman (formerly Mercer Management Consulting), one of the world’s largest general management consulting firms, where he led the firm’s global Energy and Process Industries practice.


APT is a leading cloud-based analytics software company that enables organizations to rapidly and precisely measure cause-and-effect relationships between business initiatives and outcomes to generate economic value. APT’s software suite is revolutionizing how executives leverage their data to improve commercial performance. It helps organizations determine cause-and-effect relationships between changes in business strategy and KPIs. APT has offices in Washington, D.C., San Francisco, London, Bentonville, Taipei, Tokyo, Sydney, and Chicago.

The Turnaround Tool-Box