New Product Example

Suppose you are evaluating a project to develop a new product. You estimate development will cost $2.3 million. There is an 80% chance that the development will be successful. If it is, you are not sure how much of the product you can sell and you are also unsure about how much it will cost to produce. You expect lifetime demand, at $23 per unit, to be about 1 million units with a standard deviation of about 0.3 million. You also expect the production costs to be about $4 per unit +/- $1. Marketing and tooling costs will be $1.2 million. You expect the entire development time and commercial viability to be only a couple of years, so you can ignore the time value of money. You want to know what range of return can be expected.

The first step in solving this problem is to construct a model that calculates the return for a single scenario where the scenario values are chosen randomly. Here is the model tree.

The three uncertain quantities (Succ, Units, and Cost) are modeled using DecisionPro's random number generators brand and nrand. Brand returns a 0 or 1 indicating development failure or success. Nrand returns a number from a normally distributed population.

Running a Monte Carlo simulation with this model produces the following results.

The simulation summary tells you that this project has a positive expected value ($12.227 million) and shows the range of possible outcomes.

The frequency distribution graphically illustrates the range of outcomes. In this example there is a strong possibility that the project will lose money (about $2 million). But if successful, the project will generate a return centered at about $15 million.

The cumulative distribution shows you the probability of meeting specific targets. For example, there is about a 20% chance of losing money and a 20% chance of earning more than $20 million (80% chance of less than $20 million).