Horizon Information Portal

MSOE Library Home

MSOE Library Home

Home

Using forecasting to understand how to increase (not just predict) new product sales / by William C. Bruhn.

by Bruhn, William C.

New products -- Marketing -- Forecasting.

New products -- Management

v, 125 leaves : ill. ; 29 cm.

Advisor: Gene Wright.

Committee members: Cecil Head, Kimbel Nap.

Introduction -- Hypotheses conclusion -- Best practices -- Market survey -- Annual reports -- Seminars -- Computer software -- Consultants -- Secondary research -- Conclusion.

The primary role of management is to improve the firm's processes. New product sales forecasting is one such process. The sales forecasting process can be used to increase, not just predict, new product sales. Forecasting new product sales is difficult due to market change and complexity. The changes are random and the complexity of the market always means that information is lacking. Traditional wisdom has forecasters concentrating on, 'how to increase the new product sales forecast accuracy?' This is important and it will be addressed in the course of this research. However, the better question to focus on is, 'how to increase the new product sales forecast?' If a logical rational can be established to this question, then the resultant actions from this exercise should indeed make this a self-fulfilling prophecy of increased sales. Management needs to answer both questions. Six information sources were used to compile this thesis: trade publications, an original market survey, corporate annual reports from the Dow Jones 30 Industrials, seminar literature, computer software information and consultant brochures. The Journal of Business Forecasting and Journal of Product Innovation were the most prolific trade literature sources. An original survey was done to gain insights from 'real world' practitioners. The annual reports yielded what was on the minds of the nation's top CEO’s -- new products were definitely critical. Seminars are another way to increase the forecasters skill base. The Marketing Institute and Institute of Business Forecasting each run forecasting seminars.

The forecasting specific software identified was: Forecasting Pro, Smart Forecasts, Sibyl / Runner. A number of consultants were also identified that specialize in the area of forecasting: The BASES Group, ADA Applied Decision Analysis, Hauser Furstace and Elrick & Lavidge. Each of these research information sources offered a new perspective on the seven hypotheses (H1 to H7) being studied. With five of the seven hypotheses there was sufficient proof to validate their claim. Strategic corporate measurements influence new product sales forecasting (H1). Forecasts evolve from a need for a broad numerical range to a narrow range (H2). The role of forecasting changes over the new product cycle from identifying winners and losers, to being the key measurement of success (H4). Sales forecasting identifies and defines the variables that are critical to improve the likelihood of new product success (H6). Finally, a team needs to recognize and avoid both resource costs and opportunity costs that result from inaccurate forecasts (H7). The other two hypotheses had elements of truth but were elected to need further evidence. External business environmental changes (e.g. recession, niche marketing., ...) in the 1970's changed forecasting practices in place from the 1960's. However, the internal business restructuring (eg. Teams, globalization, ...) of the 1990’s has not similarly impacted the forecasting process -- just yet that is (H3). It would seem logical that one or two key parameters may dictate proper forecasting method selection. However the reality seems to be that with minimal guidelines in existence for the forecaster, one uses any and all known methods (H5). A number of 'Best Practices' were noted that new product development teams can use in their search of how to increase forecast accuracy and actual sales. When new product development teams attempt to model their product’s sales with a diffusion model they need to answer three very good marketing questions: how will communications spread about the new product – mass media and/or word of mouth? When will adoptions peak? When will cumulative adoptions peak (market saturation)? Another excellent exercise for new product teams is to describe their new products with Roger’s General Attributes: divisibility (trialability), complexity, communicability, relative advantage, perceived risk and compatibility. Research shows that compatibility has a strong direct impact on purchase intent, as do perceived risk and relative advantage but to a lesser degree. New products are critical to a healthy future for any company. Not only do they gain a competitive advantage in the marketplace for the firm, they also earn better margins. In particular, firms should strive to: be innovators not imitators, reduce the time to market with new products (eg. use parallel rather than serial processes), find ways to view their markets as infinite (eg. global perspective) and recognize and reward teams and individuals who champion their new products. The corporation should use measurements to encourage new product development and hence, new product sales. Research and Development (R&D) is an expense for future growth. R&D expense is at 6% to 8% in America’s most innovative firms. Three different measurements can help the firm to track their progress: new product sales as a percentage of total sales, number of new products introduced and new product sales dollars. To improve the new product sales forecasting process itself, there are a number of suggestions to follow. Research seems to reveal that better application of known methods is needed, rather than invention of new techniques altogether. Approximately 40 forecasting methods were identified that the team can draw upon. Numerous studies point to the 'Jury of Executive Opinion' or management team opinion as the most popular technique. The key rule of thumb though is that using multiple methods and combining the knowledge gained from each will produce the most accurate forecast. For new products there are three main methods available: analogous products (past sales and future predictions), internal company opinion (management and sales force) and external company opinion (potential customers). All market research projects on new products should include purchase intention questions. There are two advisable improvements to this area of inquiry.

First, in the question itself use the probability phrases of: certain, high chance, even chance, low chance and never. Secondly, during results interpretation use the purchase intent translation of 75% (of certain respondents), 25% (of high chance respondents, etc. ...), 10%, 5% and 2%, to modify the survey results to what could be expected actual sales results. An interesting analysis of early sales data by the new product team should look at the purchase interval between the 1st and 2nd purchases. One study found that the shortest intervals usually belong to what later become the largest volume purchasers. Another rule of thumb to keep in mind is that 8 to 10 experts will create a very accurate forecast, and that additional opinions generally do not increase the accuracy. Forecasters also need to be aware of the forecasters' 'Survivor’s Curse' – products that survive to be actually market tested tend to disappoint in terms of their forecast. Also the corresponding bias of 'Prophet’s Fear' – forecasters may underestimate since low forecasted products never make it to market to allow judgment on the forecast accuracy. The basic forecasting formula used by consultants is a good guide for both entrepreneurs and corporate new product development teams. Consultants combine the marketing plan with a survey of potential customers plus add in their past experience usually captured in some model or numerical format. Note the trick is to have all three pieces of this puzzle completed. In closing, new product sales forecasting is a valuable dynamic viewpoint for the team to use throughout the new product development cycle and beyond. It can be used to increase (not just predict) new product sales.

Horizon Information Portal 3.25_9885

Powered by SirsiDynix

© 2001-2013 SirsiDynix All rights reserved.

Horizon Information Portal