Research Services

Demand Forecasting

One of the most important missions for market research is predicting reactions to changes in the marketing landscape — new product offerings, pricing changes, new technologies, and competitive activities. Our Farsight forecasting service, as well as our Prescience service customized for the pharmaceutical industry, provide robust predictive modeling to help our clients:

Project the effects of pricing shifts on sales volumes and profitability
Forecast the effect of promotional incentives on short-term and long-term sales, purchase cycles, and net profits
Forecast the adoption of new products over multi-year time horizons
Forecast uptake in different market segments and manage product portfolios to maximize profitability and loyalty across segments while minimizing cannibalization
Forecast the impact of product enhancements on market penetration

Different market environments and technologies each present unique challenges to demand forecasters. National Analysts Worldwide has a long history of forecasting new product performance for clients in a wide range of markets, including pharmaceuticals and biotechnology, consumer products, financial services, package transport, information technology, and energy/utilities. Our Farsight service offers extensive, continually expanding, capabilities building on sophisticated survey design, statistical modeling proficiency, a suite of comprehensive analytical tools and depth of experience in all the markets we forecast.

For mature products, forecast modeling may be driven primarily by sales histories and knowledge of emerging market trends. For new or repositioned products, forecasts are more typically informed by survey research among prospective buyers. Key requirements for successful forecast modeling include:

Understanding the competitive landscape and emerging market context.
Framing the purchase/usage decision context in meaningful ways.
Applying a well-chosen approach from our versatile and validated suite of Farsight forecasting techniques.
Developing integrated models that reflect the influence of multiple decision makers on acquisition and usage.
Calibrating demand estimates to reflect market knowledge/norms and to address overstatement of purchase/usage intent.
Placing the output in context and drawing market-savvy interpretations.

Prescience

Prescience provides comprehensive and powerful decision support for the Pharmaceutical Industry by enabling clients to take advantage of custom models and modeling applications that reflect the depth of our industry experience and our advanced analytics. Specialized modeling services include:

Demand Forecasts (propelled by a suite of choice models) for new molecules and platform technologies
Clinical Program Optimization Models designed to guide trials strategy
Integrated MD-Patient Demand Models that reflect new "democratic" models of therapy decision-making
DTC ROI Models to validate investment decisions and optimize media strategies

Prescience modeling architecture is both sturdy and elegant; it can be designed to sit atop complex Bayesian choice model platforms and accommodate a very large array of multi-year assumptions about the behaviors and interactions of physicians, payors, patients, and competitors. Ultimately, what sets it apart is not merely design finesse and sophistication, but the way in which decades of real world experience in pharmaceutical marketing inform and guide interpretation of the output. Prescience is not merely a set of advanced algorithms and high-tech modeling mechanics. It's a thoughtful and inventive problem-solving process powered by creative analytics and fueled by keen sensitivity to trends, changing market structures, and evolving medical paradigms.

Competitive Landscape and Emerging Market Context

Valid demand forecast requires sensitizing research respondents to environmental and competitive conditions that are likely to exist when the product will be introduced. Depicting that future environment is often difficult, however, even for those with well-informed industry perspectives. The path of technology and other market events is often uncertain, and it is difficult to place respondents in the proper frame of mind simply by "fast-forwarding" the scenario with a set of future "facts."

For that reason, the fine art of survey design – including a sensitivity to how much new information respondents can absorb and how best to evoke a "future world" – is every bit as important as modeling mechanics. Ensuring that respondents are properly "educated" about new technologies and that the potential market scenarios are clearly articulated to them is a critical focus of our forecasting approach.

Framing the Purchase/Usage Decision

Asking customers to help us predict the future requires that we frame the offering in accurate, meaningful ways as we evoke all other relevant aspects of the market environment we aim to forecast. It also requires us to couch the decision (purchase or selection) process sensibly and realistically. We need to ask respondents for the kinds of judgments they will actually make and give them the tools to do it.

Although we can account for multiple product contingences and other sources of uncertainties using conjoint models, evoking the purchase context involves more than correctly presenting the future contingencies. There are decisions to be made about how much information to provide and how "promotional" a story to tell in order to mimic what customers will know and think when.

The challenge is especially notable when we need to model new products or technologies whose performance may not be easily envisioned – like innovation capable of changing our world and our decision priorities in profound ways. Research craftsmanship can make all the difference between responses that merely answer the question and responses that meaningfully anticipate the future.

National Analysts Worldwide is often called upon to help companies forecast demand for "paradigm-changing" innovation, and we have found that simple, "canned" forecast algorithms often fall short. Extensive qualitative research, careful selection of prototype customers, and extensive concept immersion are among the steps we take to increase the likelihood that research can do justice to novel ideas and that new product forecasts do not suffer from a failure of anyone's imagination.

Forecasting Support

Using historical information as well as primary data and well-engineered modeling engines, National Analysts Worldwide helps clients anticipate the response to future market conditions and potential business scenarios. The models rely on a variety of different techniques, including conjoint analyses, multivariate VAR, co-integration, and panel co-integration methods. Forecasts are calibrated to properly capture current market conditions and temporal dynamics, and validated to ensure that the underlying models are robust. They are also adjusted for overstatement using algorithms tested over decades to verify predictive accuracy. The resulting forecasts can be delivered with confidence to senior management and relied upon to make important business decisions.

Where future market scenarios and product options or outcomes are uncertain, various forms of conjoint analysis may be used as a platform for modeling customer choices. The power of conjoint lies in the ability to accommodate a relatively large number of product variables or market scenarios, while also measuring the importance (and desirability) of various product/service attributes in accounting for market outcomes. By having customers evaluate a systematic subset of all possible product/service permutations, it is possible to simulate interest in product/service profiles not explicitly tested; and expose the decision dynamics that account for product/service selection. The result is a data set of customer choice or preference input from which predictive algorithms can be constructed to model potential market penetration or share predictions.

Conjont models (and related multivariate techniques) are the primary engines that drive our market response simulators, used by clients to:

Make key investment decisions and guide product design or commercialization strategy
Manage product portfolios in order to maximize penetration and loyalty across market segments while minimizing cannibalization
Anticipate implications of pricing shifts on sales volumes and profitability
Forecast the effect of promotional incentives on short-term and long-term sales, purchase cycles, and net profits
Forecast the adoption of new products over multi-year time horizons
Forecast uptake in different segments

Calibrating Survey-Derived Demand Estimates

Where forecasts are developed from survey-derived demand estimates, adjustments for over-projection of purchase intent are needed to avoid mis-estimation of potential sales. The degree of overstatement can be highly dependent on several factors, including the specific market, the type of customer, the novelty of the product in question, the length of the purchase cycle, and the nature of the purchase intent question.

Rather than relying on a single standardized algorithm to adjust for modeling error, National Analysts Worldwide has found that variations in product, market, and competitive conditions make single canned approaches unreliable. Our extensive experience across a broad array of markets guides a flexible, customized approach to calibration based on internal validation processes and deep knowledge of specific industries – particularly their historic responses to market innovation.

For more information on National Analysts Worldwide, please e-mail us or call (215) 496-6800.