Sales & Marketing Forecast

Section 1

I decided to use my current employer, Engility Corporation, for the forecast assignment. Engility Corporation is “ a premier provider of integrated services for the U.S. government, we support the Department of Defense, intelligence community, space communities, federal civilian agencies and international customers”(About Us, 2017). This forecast would be useful for several reasons: it would help determine resource needs, forecast the amount of financial resources needed to complete the mission, and projecting industry growth. Determining what resources are needed is vital to supporting the mission of any business. Businesses need a complete understanding of the requirements for every mission, if this is not understood then a company may not be prepared for any sudden changes in the mission. If the mission is not fully understood then it can result in overages and/or inadequacies with personnel (staffing), budgeting, and advertising. A company must be able to correctly forecast what resources the mission will require and how much it will cost to implement said resources. Having a forecast of cash flow is needed so that the company will have an estimate of how much of those funds are needed to compensate for personnel, operating expenses, and any unexpected events of the business. Accurately forecasting the industry growth is key to Engility’s survival. Working with various government agencies requires the company to be prepared to move forward with a major financial decision at any given moment. An industry forecast is also vital because it’ll allow a manager to tailor fit any business strategy that would affect the business overall. For example, if the industry unexpectedly grows the strategy would need to adapt to accommodate the abrupt changes.

This 2 year (eight quarter) forecast would be useful to both a project manager and human resources manager. While their roles are different and respected in their own rights, a project manager could use this forecast in order to create tailor fit a business strategy for the next two years and a human resources manager could use this forecast to make sure the company has the personnel and budget needed to complete the mission.

Section 2

I will be using three different methods to forecast a 2 year (8 quarter) period based off of Engility’s annual financial reports (see figure 1 in the Appendix) from the past 5 years. In order to explain this, I will use Holts exponential smoothing method, a linear regression trend with a shift dummy variable, and a combination model of the two forecasts.

Holts exponential smoothing method is “an extension of the smoothing model; it adds a trend factor to the smoothing equation as a way of adjusting for the trend” (Keating and Wilson 2009). Holts method predicts two smoothing factors: the current level and the current trend. Holts method works well with data that shows trend over time, but does not work well when there are patterns of seasonality or cyclicality. Since Engility’s data has 2 sets of linear trends and no seasonality it is ideal for Holts method. Managers can expect a forecast that has made the necessary adjustments to take into account the trends in the data which will provide a more accurate forecast. After the forecast is provided an evaluation is conducted in order to determine how accurate it is and if it works the best for our data. Measuring the errors holds a vital role in capturing the accuracy of the forecast and establishes criterion for the forecasting process. The Mean Absolute Percentage Error (MAPE) is used as a metric to measure how well the model worked historically.

The linear regression trend method attempts to clarify the correlation of two or more variables by using a straight line. The regression model is listed as a linear model, Y = a + bx. This equation is used to show how a change in “X” influences “Y”. A represents the constant in the regression model, B is the slope which represents the change in Y as X changes. The linear regression does this by “setting up time index (T) to use as the independent or X variable in the basic regression model, where T is usually set equal to 1 for the first observation and increased by 1 for each subsequent observation” (Keating and Wilson 2009). Engility experienced a tremendous increase in revenue during the second and third quarter of 2015 ‘due to the finalization of a major acquisition’. (Press Release, 2017) In order to account for this dramatic change there has to be another independent variable, the event model added to the forecast. Whenever a non-qualitative event influences sales such as promotions, acquisitions, strikes, natural disasters, etc, there must be a dummy variable added to represent the event. An event model is used to measure the impact of the event on revenue and uses that information to improve the accuracy of the forecast. The dummy variable will represent the dependent or Y variable in the regression model. Instead of only using the MAPE as a metric to evaluate the accuracy of this method, there are 5 steps to evaluating any regression model. Even though there are 5 steps we will only use 4 of them. We will not use the fifth step because it is only applicable when there are at least 3 independent variables. First, we have to determine if the model makes sense logically. Second, we must determine if each slope for each independent variable is significantly more or less than 0. Third, we must determine how much explanatory power we have by looking at the coefficient of determination. Lastly, we must determine if there is any serial correlation in the data.

When building a forecast model there should be an effort to use all available and important information. It is important to accept the fact that all forecasts will be wrong however, the desire should be to do the best job that we can in order for the company to benefit from it. One way of doing this is to combine the two forecasts. Instead of being forced to choose the best method between the two, we will combine the forecasts from each model to obtain forecast improvement. Before combining the forecasts we must determine if there is any bias in the forecast. If there is any bias then we cannot proceed with the forecast. If a bias is present the forecast will be consistently too high or too low. After we know that combining the two methods will not create a bias we can proceed with the combined forecast. This combination forecast is considered a weighted average. After we perform a regression analysis it will provide the weights that should be used to represent the individual forecast.

Section 3

The above data represents Engility’s revenue in thousands of U.S. dollars over