MODELING THE PRODUCT DEVELOPMENT FUNCTION FOR AN ENTREPRENEURIAL FIRM Developments in Business Simulation and Experiential Learning, Volume 30, 2003 MODELING THE PRODUCT DEVELOPMENT FUNCTION FOR AN ENTREPRENEURIAL FIRM Richard D. Teach Georgia Tech richard.teach@mgt.gatech.edu Robert G. Schwartz Southern Washington University rgssbd@earthlink.com ABSTRACT The purpose of this paper is to explore a variation of the sine function that represents the product development process in a small, one-product, entrepreneurial firm. The function utilizes two primary decision variables, cumulative time and cumulative capital expended on product development. Before the product may be marketed, it must successfully complete both an Alpha and Beta test, both of which are modeled as stochastic functions. The game that the functions have been developed for is DUALITY; a review of Duality was presented at the 2002 ISAGA conference. INTRODUCTION New products are very important to businesses. At Gillette, 50% of sales are from products less than 5 years old (Kahn 2001, page 17). A benchmarking project by the Product Development and Management Association (Griffin et al, 1997) found that one-third of company sales come from products introduced within 5 years of the measurement date. More than 90% of new product developments, however, did not successfully make it out of the product development stage (Kahn, 2001). Further, a Booz, Allen and Hamilton (1982) study found that only one in three new products were successful after market introduction. Thus, much goes wrong during the entire new product development process. Perhaps if simulations could effectively model this process it could help to reduce failure rates. BACKGROUND In spite of the fact that fifty percent of business profits are reported to be from products that are less than five years old (Lin and Dwyer, 1995), few business simulation authors have developed this important process within the decision framework of their games. Gold and Pray (1998, page 156) lament: “Although these dynamic changes are of growing importance and influences in the market, little attention has been given in the simulation and gaming literature to these considerations.” While not the major decision orientation of major business games, product development has been utilized in previous business games and simulations. The first game primarily devoted to product development was described in a paper by Wynne, Klosky and Snyder (1979). This game modeled the management of a product development project, where a cost plus incentive fee contact was the objective of the participant teams. The leadership simulation, LOOKING GLASS, McCall, and Lombrado (1982), developed an in-basket exercise and simulated product development along with capital investment, firm vision, mission and other items of importance to senior corporate officers to develop leadership skills. Recently, Nicholson and Oliphant (2002) developed a semester long experiential exercise in new product development. Business games that have used product improvement or development include: MICROMATIC (Scott and Strictland, 1985), which uses lump sum commitments for product improvements to allow for product differentiation; THE BUSINESS STRATEGY GAME (Thompson and Strappenbeck, 1995), which uses Quality control to minimize reject rates; THE BUSINESS POLICY GAME (Cotter and Fritzsche, 1991) which has players divide R&D expenditures into two parts, one part is for Process R&D to reduce costs and the other part is for new product development, which results in developing advanced designs of the current product; MARSTRAT (Larreche and Gatignon, 1997) which allows the marketing department the development of a new product. In MARSTRAT, there is a definite Product Development – Marketing interface. But in none of these games is the Product Development function one of the primary concerns of the game. DUALITY The product development algorithm described in this paper was designed for a business game DUALITY (Teach and Schwartz, 2002), which simulates both technology- based start-ups and a set of venture capital firms (VCs). 244 mailto:richard.teach@mgt.gatech.edu Developments in Business Simulation and Experiential Learning, Volume 30, 2003 Y(p) = A * sine ((π*(d’))/(2B)) C (1)1 The VCs’ provide capital in exchange for shares of the start- up firms’ common stock. Whenever a start-up firm has need for additional capital resources, it negotiates the best deal it can obtain from one of the VC firms in the game. The start-ups and the VCs co-exist in the same game with participants managing both sets of firms. In this equation d’ is the value (money or time) used in the simulation for the actual decision for a particular period and A, B and C are controllable parameters. B is the amount of the budget (time and money) the product development process will actually take. The parameter p is the period being simulated. Estimates of Bp are provided to the game players, initially a randomly selected number centered on 70 percent of the actual values. This estimate is continually updated and approaches the actual values as game play progresses. The input variable d (the actual decision) is conditionally exponentially smoothed as shown in equation 2. In DUALITY, players assign hours of their simulated time among the various tasks needed to operate their specific firm, i.e., either a VC or start-up. They must devote time to product development as well as, searching for capital, fighting fires, (a euphemism for solving the immediate and serious problems that arise suddenly in organizations), planning and strategy development, marketing, managing people, networking and selling. Both time and the more traditional monetary expenditures play a role in the game. Capital must be allocated to product development, salaries for the officers, new hires, etc. There are general overhead costs and costs for marketing and sales, materials and supplies. Budgets need to be developed and cash flows must be constantly monitored in the start-ups. DUALITY is a game that deals with technology based firm developments and the paradigms that involve the interchange between venture capitalists and the firms’ developers, i.e., a duality. Thus, a product development algorithm is central to this game. Value of d’ at periodp = Value of d’ at period p-1 + ES (d)p (2) where ES(d)p = [ES(d’)p-1*α] + (d)p*(1-α) and values of (d)p that are greater than or equal to [(d)p-1 *(1- α )]. When (d)p is less than [(d)p-1 *(1- α )] then: ES(d)p = [ES(d’)p-1*α] + (d)p (3) The value of (d)p is the actual decision for period p. and (d’)p is the value of the accumulated and exponentially smoothed decision used in the simulation for period p. At the starting point when p = 1, ES(d)p-1 is set to a specific value selected by either the game designer or the game facilitator. SIMULATION OF THE PRODUCT DEVELOPMENT PROCESS The product development process for an entrepreneurial firm is far simpler than that of a large firm. An entrepreneurial firm usually starts with a single product concept, whereas a large, existing firm has a portfolio of new and developing products. This paper describes the algorithm developed for the simulation, DUALITY, for simulating the product development process for a single product. It is a stochastic function and uses two variables for decision-making: the investment of capital and the amount of time devoted to the overall product development process. In the previously mentioned games, only money invested in R&D or product development determined the simulation outcomes. (MARKSRAT included product attributes). Perotti and Pray (2002) used only a monetary budget for their product development function. The process of exponentially smoothing prevents sudden and dramatic changes resulting from radical decisions that can result in an unwanted impact on the outcome. Simulations results are a function of the mathematics included in a game. If sudden and radically different levels of inputs occur, most algorithms fail to provide realistic results. (See Perotti and Pray, 2000, while this article does not directly address this specific issue, it does show that algorithms may not behave as expected under unforeseen decisions. Exponential smoothing simply dampens the affect of radical decision changes. The function for accumulated capital spent on product development is the same one used for the accumulated time devoted to product development. The parameters A, B and C and the exponential smoothing constant Alpha may all be different. Figure 1 graphically depicts the shape of the function Y(p) across multiple values of the input variable d’. There are substitution effects in this function at all points except when the value of the function is at it maximum value. Thus, if more time is spent on product development, then less capital needs to be invested in product development and vice-versa. THE FUNCTION The product development algorithm has three parameters that can be used to control its behavior. The basic function is a modified sine curve and is in the form of: 1 The authors recommend that interested readers program this function into a spreadsheet and plot the results for about 100 points. Then they should vary the parameters and watch how the function performs. 245 Developments in Business Simulation and Experiential Learning, Volume 30, 2003 Figure 1 A graph of the function values for a range of the input variable 0.00 0.20 0.40 0.60 0.80 1.00 1.20 0 20 40 60 80 100 120 Cumulative Budget (the input variable) Y( d' ) The arrows point to the inflection point The value of the parameter “A” has been set to the value 1. The parameter A is a scalar and alters the maximum value of each sub-function. At a value above 1, the product of the two sub-functions changes shape. The parameter “B” is the point at which the first derivative of the function is equal to zero. The value of B is the maximum amount of either time or the maximum amount of capital to be invested in the project. The inflection point is influenced by the selection of the parameter C, and for the purposes of this paper, C is defined as equal to 2. The selection of this value results in the inflection point always being at the mid- point of the function. As noted, there are two variables used in the simulation of product development, the accumulated capital invested and the accumulated time expended. The function can represent both time and money variables and the values of the two functions are multiplied together to produce the joint function for both variables. Since the values of both functions are less than one, the combined function has its inflection point shifted to the right. The shape of the total product development function is shown in Figure 2. 246 Developments in Business Simulation and Experiential Learning, Volume 30, 2003 Figure 2 The Product Development Function 0% 20% 40% 60% 80% 100% 120% 0 20 40 60 80 100 120 Value of the function of the accumulated moetary budget and time when the two values are equal Pe rc en t c om pl et ed The arrows point to the inflection point Certain marketing processes accompany the overall product development process. Alpha and Beta testing are two of those. Alpha tests are done earlier in product development when the product is only partially ready for market. Beta tests are generally conducted when the final product form is in place and ready for limited test market. The simulation can model these tests. ALPHA AND BETA TESTING A priori, the new product will be defined to be ready for Alpha testing whenever the product development function is equal to 0.80 or greater. In reality, product development has no upper limit. When successful in-house (alpha) testing occurs and when the market is willing to accept the product (a successful beta test), it can be marketed. However, one needs to realize that further development may increase the product’s utility to its buyers. For the purpose of the game, the probability of a successful alpha test P(Alpha Success) should be about 0.50 when Y is equal to 0.80 and P(Alpha Success) should be about 0.60 when Y is equal to 0.85 and P(Alpha Success) should be about 0.90 when Y is equal to 1.00 (See Tables 1 and 2). Beta testing can only take place after a successful Alpha test. Beta test probabilities would be expected to be lower that the probabilities of a successful Alpha test when the value of the function Y are comparable. These numbers are arbitrary and can be varied, thus influencing the outcome of the simulation. The functions used to produce these probabilities are discussed in appendix A Costs for an Alpha Test can be defined by the game administrator and would vary depending on what product developments are being simulated. Beta testing is done with a limited number of select customers and either “free” or at reduced prices in order to be attractive to potential users and thus costs are much more to run a Beta Test. The game has been set so that three Beta test failures lead to market failure (although this number can easily be changed). Probability functions To determine the function for the probability statements, Microsoft EXCEL was utilized. The coefficients and the resulting alpha test functional values are shown in Table 1. 247 Developments in Business Simulation and Experiential Learning, Volume 30, 2003 Table 1: The coefficients and the resulting functional values for the Alpha Test Term Final Coefficients Functional Values Estimated Probabilities Resulting Probabilities Intercept 23.75 Y(0.80) 0.50 0.48 Linear term -92.00 Y(0.85) 0.60 0.61 Squared term 117.00 Y(0.90) 0.73 Cubic term -47.00 Y(0.95) 0.82 Y(1.00) 0.90 0.88 Table 2 shows the final coefficients that were found with their expected and approximated probabilities for the beta testing function. Table 2 The coefficients and the resulting functional values for the Beta Test Term Final Coefficients Functional Values Estimated Probabilities Resulting Probabilities Intercept -75.4 Y(0.80) 0.44 Linear term 232.0 Y(0.85) 0.60 0.63 Squared term -234.0 Y(0.90) 0.70 0.73 Cubic term 79.0 Y(0.95) 0.77 Y(1.00) 0.85 0.80 STAGE GATE PROCESS If one wanted to include a more sophisticated stage- gate process (O’Connor, 1994), a series of probability estimates could be developed for each gate. The gates would be placed to occur at different points along the function Y. Care needs to be taken, because a realistic simulation may have too few successes to be played in a normal classroom environment. It must be remembered that a business simulation is a learning device and not necessarily descriptive of an actual situation. SUMMARY This paper has discussed a product development algorithm that may be used in a game where new product development is the primary focus of a game or simulation. It original purpose was to be used in a game that emphasized entrepreneur endeavors. That is, when the firm must succeed in product development or fail. In large, well- financed firms, product development is a continuing process and frequently involves the development of multiple products. Games developed for large firms with multiple products “play the probabilities.” But, entrepreneurial firms, while facing the same probabilities, have only one time to “roll the dice.” If the product fails, the firm fails. This is what the game DUALITY simulates. APPENDIX A DETERMINING THE FUNCTIONS FOR THE PROBABILITY STATEMENTS The probability functions are easy to develop using Microsoft EXCEL. First put in the three function values in one column, the squares of these function values in the second column, the cubes of the function in the third column, then in the fourth column, key in the desired probabilities. Then add 2 or 3 additional lines with values close to the three values originally selected (This prevents a degrees of freedom error.). After the derived solution, check to make sure the function is monotonic (that is one that is always increasing, never decreasing). Write the equation on the spreadsheet and provide 20 to 30 values and plot the result. This plot will indicate if you have a non-monotonic function. If you do have a non-monotonic function, adjust the coefficients until you get a monotonic function. When using the suggested probabilities for the alpha test given in the previous section, a non-monotonic function resulted. After a few adjustments, coefficients were found that approximated the desired values. 248 Developments in Business Simulation and Experiential Learning, Volume 30, 2003 REFERENCES Booz, Allen and Hamilton (1982), “New Products Management for the ‘80s, Chicago. Lin, Binshan and Donald Dwyer (1995), “New product development in the information age: the `pharmaceutical industry,” Journal of Business and Industrial Marketing, Vol. 10, No. 3: pp.6 - 17 Cotter, Richard V and David J Fritzsche (1991), THE BUSINESS POLICY GAME, Prentice Hall, Englewood Cliffs, NJ Gold, Steven C. and Thomas F. Pray (1998), “Technological change and intertemporal movements in consumer preferences in design of computerized business simulations with market segmentation,” Developments in Business Simulation and Experiential Learning, Vol. 25: 156 - 167. Griffin, A., P. Belliveau, S. Markham, E.F. McDonough, III, D. Olson and A.L. Page (1997, October), “Drivers of NPD success: The 1997 PDMA report” Chicago. Kahn, Kenneth B. (2001), Product Planning Essentials, Sage Publications, Thousand Oaks, CA. Larreche, Jean-Claude and Hubert Gatignom (1997), MARKSTRAT A Marketing Strategy Game, The Scientific Press, Palo Alto, CA. McCall, M.W., Jr. and M.M. Lombrado (1982), “Using simulation for leadership and management research: through the looking glass,” Management Science, Vol. 28, No. 5: 533 – 549 Nicholson, Carolyn Y. and Becky J. Oliphant (2002), “Using project-based experiential learning groups in the principles of marketing course, ”Developments in Business Simulation and Experiential Learning, Vol. 29: 131 -133. O’Connor, P (1994), "Implementing a stage-gate process: A multi-company perspective,” Journal of Product Innovation Management Vol. 11, No. 3, 183 - 200. Perotti, Victor and Thomas Pray (2000) “Visual Modeling of Business Simulations,” in Developments in Business Simulation and Experiential Learning, Vol. 27, Edited by Diana Page and LT Snyder, Published by the Association for Business Simulation and Experiential Learning. Perotti, Vic and Tom Pray (2002), “New product development (NPD) simulations: some challenging questions and tough modeling issues,” Developments in Business Simulation and Experiential Learning, Vol. 29: pages 265 to 269. Scott, Timothy W. and Alonzo J. Strickland (1985), MICROMATIC, Houghton Mifflin Company. Boston, MA Teach, Richard D. and Robert G. Schwartz (2002), “Duality: A dual and congruent business simulation,” presented at the ISAGA2002 annual meeting, Edinburgh, Scotland. 26 to 30 August 2002. Paper to be published. Thompson, Jr., Arthur A. and Gregory J. Stappenbeck (1990) THE BUSINESS STRAEGY GAME, Irwin, Chicago, IL Wynne, A. James, J. Michael Klosky and Charles A. Snyder (1979), “Computer-aided performance control simulation: an interactive experiential gaming technique for managerial decision-making,” Insights into Experiential Pedagogy, Vol 6, 299 - 300. 249 Table of Contents Volume 30, 2003 The Optimal Timing For Introducing Business Simulations Can Handicapped Students Access Your Class Web Site? The Competition Game: Decision Making In A Dynamic Environment Pan-Pacific Enterprises: Strategic Decision Making Simulation Study Of Stochastic Channel Redistribution The Impact Of Business War Games: Quantifying Training Effectiveness Experiential Learning: Introducing Faculty And Staff To A University Leadership Development Program The Feasibility Of The Balanced Scorecard For Business Games A Misuse Of Pims For The Validation Of Marketing Management Simulation Games Incorporating Technology Into The 21st Century Classroom: Are We Facilitating Academic Dishonesty? Improving The Effectiveness Of Peer Evaluations The Use Of A Simulation In An Integrated Mba Curriculum Student Portfolios In Business Education Student Portfolios In Business Education At Ashland University Using SAP ERP Technology To Integrate The Undergraduate Business Curriculum Board Games And Teaching Textile Marketing And Finance Blogging: A New Threat To Student Research? The Way We Talk! Take II Strategic Management: An Evaluation Of The Use Of Three Learning Methods In Hong Kong Adoption Of Discussion-Based Teaching And Assessment In Teaching Strategic Management In Hong Kong The Tobin Q As A Company Performance Indicator Using Representative Nominal Group Technique For Course Review And An Interactive Solicitation Of Ways To Enhance Absel's Image Making Teaching Matter: The Art And Science Of Teaching Business Communication Beyond Sex, Age, And Race: Exploring The Deeper Contents Of Diversity Teaching & Learning The Facilitation Process A Brief On Debriefing: What It Is And What It Isn't Changing Perceptions Of The Importance Of Leadership: The Contribution Of Individual Spirit Harmonics In Leadership Knowledge, Skills And Sustainable Values In Learning Organizations: Some Implications From The Multicultural Virtual Classroom Challenges Of Teaching Undergraduate Organizational Behavior In A Nontraditional Time Format Interactive Online Positioning With The Web-Based Product Positioning Map Graphics Package The Longitudinal Effects Of Entrepreneurship Training On Risk Tolerance: A Look At Similarities And Differences Between Male And Female Undergraduate Students Revisiting Strategy Learning In A Total Enterprise Simulation What Are Simulations For?: Learning Objectives As A Simulation Selection Device Gaming Agency Markets Cooperate For Profits Or Compete For Market? Study Of Oligopolistic Pricing With A Business Game The Design Of A Business Simulation Using A System-Dynamics-Based Approach Modeling The Product Development Function For An Entrepreneurial Firm Simulation Performance And Forecast Accuracy? Is That All? Business Manager Identification Of Competitors In Real World And Simulation Settings Monte Carlo Simulation Analysis On The Costs Reduction Argument Of Interest Rate Swaps Ebiz Game: A Scalable Online Business Simulation Game For Entrepreneurship Training A Model For Online Education Delivery And A Look At Online Delivery Effectiveness Incorporating "Company Reputation" into Total Enterprise Simulations The Genesis And Future Of The Absel "Classicos" Initiative