Toffee Inc Case Report

Toffee Bar Inc

Group 1

Introduction

Toffee Inc is an organization that has experienced steady growth for the past 5 years.  Demand for its Seven Star chocolate bar has increased at an average rate of 6.25% each year over this time period, so it came as a surprise to the organization when the production manager declared, “If the ingredient inventory that was purchased and managed by the firm is not re-examined and re-worked to the firm’s advantage then [soon] the final products based on these ingredients will cease to yield the kind of profits the firm expects”.  This declaration caused the inventory manager to take action by 1) forecasting expected sales for 2011 and 2) based on the forecast, decide on most cost-effective strategy to order vendor supplied ingredients.

Forecasting        

To capture a wide range of perspectives and strategies in forecasting, 3 different models were separately created to determine the best way to move forward. These methods were aimed to decrease total costs for the Seven Star bars, while also providing information regarding optimal order points and holding inventory for ingredients. By calculating these models separately, our team was able to bring diverging points to light which led to discovering the model that would best meet our needs.

        One divergence in the models that were created was in the forecast method itself.  Two models forecasted the 2011 months by trending from equivalent months in years prior.  For example, the demands for the month of January by year between 2006 – 2010 were 742, 741, 896, 951 and 1030 respectively.  Therefore, the forecasted figure for 2011 was 1043, which was found by taking a monthly index of the demand and comparing each month with the average yearly demand.  This seemed to be a sound method that appropriately accounted for Toffee Inc.’s historically seasonal demand, but our third model forecasted decreases in demand for 2011, despite most months showing upward historical trends.  This was briefly debated until it was recognized that the last 4 months of 2010 was causing this discrepancy.  December for example, saw demand between 2006 – 2010 was 783, 901, 1023, 1209, and 1013 respectively, showing there was a significant dip of 16.2% in demand in 2010.  A method that could capture the most recent trends, without giving too much credence to wide variations was needed to best forecast the 2011 year.  Exponential smoothing was the found to be the best option for this situation as it provides a relatively accurate forecast when demand is fairly consistent. We found when we implemented exponential smoothing, we were able to show that it could not only assist with forecasting future demand but, we were also able to show that it was able to properly forecast prior year demand. This  showed that it could be a reliable method for us to use to predict the demand for 2011’s Seven Star Bars.

        To utilized exponential smoothing, an optimal α and β must be found.  The α will be applied to each data point using the forecast formula[pic 1]

The β will be used to forecast trend, using the formula[pic 2]

The α and β (forecast and trend) are then added together to arrive at a forecasted demand figure.  When applying these formulas to historical data, the goal seek function to arrive at the α and β that best fit the data points on historical demand.  Once the optimal α and β are found, they are applied to the formula to forecast future trends.  Based on the historical data for this group, the α that produced the best fit was .304 and the β was .207.  Figure 1 was created to show how well these figures (in red) fit historical data (in blue).  Note how well the forecasted years fit the actual demand. When applied to the 2011 figures, the predicted demand follows the seasonality appropriately but also show a downward trend in demand.  These results represent a model that the group felt best reflected historical data and should result in the most accurate forecast for demand in 2011.

Order Strategy

Now that a demand forecast was established, further action had to be taken to ensure costs of ordering and holding inventory were minimized.  To do this, an order strategy was developed.  Both Q-Systems (order when inventory drops to a certain level) and P-Systems (order up to a certain amount at fixed periods) were discussed, and due to the seasonality of this product, it was decided to use a Q-System for ordering to best meet these fluctuations.  This meant establishing order quantities and reorder points to ensure Toffee Inc. consistently held onto enough inventory to meet demands.  Factors that were considered in the ordering strategy were order fees, units costs, demand forecasts, usage rate for each ingredient and holding costs.

To arrive at a reorder point warranted a fair amount of debate amongst the group.  It was vital that stock was adequate to meet expected service levels of 95%, but how this was to be done took the form of two different approaches.  The first approach was to simply forecast an amount that represented the true forecast plus 1.64 times the standard deviation of all historical demand figures.  While this approach would have accomplished the goal of ensuring enough safety stock, it also meant holding a great deal of stock resulting in high holding costs. Instead, it was decided to find an adequate safety stock using the 95% level of confidence.  The safety stock calculation applied average weekly demand plus 1.64 standard deviations of monthly demand.  It is projected that stock levels will be adequate to meet expected service levels of 95%.  After applying this method, the safety stock (reorder point) came out to the numbers in Figure 2.

Inventory

It now had to be established how much inventory should be purchased to optimize costs.  To do this required data on holding costs, unit price and ordering fees. Figure 3 in the appendix provides figures and data on these costs.

Based on the data calculated in Figure 3, the optimal order amount was computed using the EOQ (Economic Order Quantity) formula.  This formula optimizes costs by considering the cost to order (Cp), demand (A) and holding costs (Ch). Figure 4 lists the EOQ for each main ingredient of the Seven Star Bars.