Cutting Edge Case Assignment (Kenny D. Neto) PDF

Preview

Summary

Case study...

Description

Nova Southeastern University Wayne Huizenga Graduate School of Business & Entrepreneurship

Assignment for Course:

QNT 5160 – Data Driven Decision Making

Submitted to:

Yuliya Yurova

Submitted by: Kenny D. Neto

Date of Submission:

February 28, 2018

Title of Assignment: Individual Case Project- Cutting Edge

CERTIFICATION OF AUTHORSHIP: I certify that I am the author of this paper and that any assistance I received in its preparation is fully acknowledge and disclosed in the paper. I have also cited any sources from which I used data, ideas of words, whether quoted directly or paraphrased. I also certify that this paper was prepared by me specifically for this course.

Student Signature: ___________________________

******************************************* Instructor’s Grade on Assignment: Instructor’s Comments:

1

QNT 5160, Fall 2016 Semester Individual Case Assignment: Cutting Edge This case was adapted from Hiller, Frederick S. & Mark S. Hillier (2014). Introduction to Management Science: A Modeling and Case Studies Approach with Spreadsheets, 5th ed., McGraw-Hill/Irwin, pp 429-432.

Mark Lawrence has been pursuing a vision for more than two years. This pursuit began when he became frustrated in his role as director of Human Resources at Cutting Edge, a large company manufacturing computers and computer peripherals. At that time the Human Resources Department under his direction provided records and benefits administration to the 60,000 Cutting Edge employees throughout the United States, and 35 separate records and benefits administration centers existed across the country. Employees contact these records and benefits centers to obtain information about dental plans and stock options, change tax forms and personal information, and process leaves of absence and retirements. The decentralization of these administration centers caused numerous headaches for Mark. He had to deal with employee complaints often since each center interpreted company policies differently – communicating inconsistent and sometimes inaccurate answers to employees. His department also suffered high operating costs since operating 35 separate centers created inefficiency. His vision? To centralize records and benefits administration by establishing one administration center. This centralized records and benefits administration center would perform two distinct functions: data management and customer service. The data management function would include updating employee records after performance reviews and maintaining the human resource management system. The customer service function would include establishing a call center to answer employee questions concerning records and benefits and to process records and benefits changes over the phone. One year after proposing his vision to management, Mark received the go-ahead from Cutting Edge corporate headquarters. He prepared his “to do” list – specifying computer and phone systems requirements, installing hardware and software, integrating data from the 35 separate administration centers, standardizing record-keeping and response procedures, and staffing the administration center. Mark delegated the systems requirements, installation, and integration jobs to a competent group of technology specialists. He took on the responsibility of standardizing procedures and staffing the administration center. Mark had spent many years in human resources and therefore had little problem with standardizing record-keeping and response procedures. He encountered trouble in determining the number of representatives needed to staff the center, however. He was particularly worried about staffing the call center since the representatives answering phones interact directly with customers – the 60,000 Cutting Edge employees. The customer service representatives would receive extensive training so that they would know the records and benefits policies backwards and forwards – enabling them to answer questions accurately and process changes efficiently. Overstaffing would cause Mark to suffer the high 2

costs of training unneeded representatives and paying the surplus representatives the high salaries that go along with such an intense job. Understaffing would cause Mark to continue to suffer the headaches from customer complaints – something he definitely wanted to avoid. The number of customer service representatives Mark needed to hire depended on the number of calls that the records and benefits call center would receive. Mark therefore needed to forecast the number of calls that the new centralized center would receive. He approached the forecasting problem by using judgmental forecasting. He studied data from one of the 35 decentralized administration centers and learned that the decentralized center had serviced 15,000 customers and had received 2,000 calls per month. He concluded that since the new centralized center would service four times the number of customers – 60,000 customers – it would receive four times the number of calls – 8,000 calls per month. Mark slowly checked off the items on his “to do” list, and the centralized records and benefits center opened one year after Mark had received the go-ahead from corporate headquarters. Now, after operating the new center for 13 weeks, Mark’s call center forecasts are proving to be terribly inaccurate. The number of calls the center receives is roughly three times as large as the 8,000 calls per month that Mark had forecasted. Because of demand overload, the call center is slowly going to hell in a handbasket. Customers calling the center must wait an average of five minutes before speaking to a representative, and Mark is receiving numerous complaints. At the same time, the customer service representatives are unhappy and on the verge of quitting because of the stress created by the demand overload. Even corporate headquarters has become aware of the staff and service inadequacies, and executives have been breathing down Mark’s neck demanding improvements. Answer Questions 1a through 1c below: Question 1a: Define a problem statement which reflects the challenge facing Mark as he planned for the opening of the new center. Mark’s challenge is to try to centralize records and benefits administration by establishing one administration center. Furthermore, he needs to find an equilibrium between cost reduction and effiency, which leads to the following question: what is the best way/method to forcast the expected amount of calls so that Mark knows exactly how many people to hire so that he doesn’t have to overpay for staff (causing costs to go up) and he doesn’t get to be understaffed (causing effiency to go down)?

Question 1b: Why was Mark’s initial forecast of call volume so far off? What could have been the reasons for this? 3

Mark’s initial forecast of call volume was so far off because the sample size (one call center out of 35 centers) that he analyzed was too small which resulted in a bias forcast. Also using the judgemental forecasting method (which is mostly used when there is no available data in which statistical methods can be used) was not the best idea because with such complex data (big amount of data to be a analyzed) a different and more analytical forecasting method should have been used.

Question 1c: What could Mark have done differently to improve his initial forecast? To improve his initial forecast, Mark could have increased the sample size and analyze all 35 centers. Then Mark could opt for a diffentent analyzes method do to the complexity of the data. Intstead of using the judgemental forcasting method, Mark could have collected data from different time periods and taken into consideration the seasonal effects (since there can be a relationship between call levels and seasonality) and adjust the data accordingly by calculating the seasonal factors.

4

Mark needed help, and he approached Harry, a corporate analyst, to forecast demand for the call center more accurately. Luckily, when Mark first established the call center, he realized the importance of keeping operational data, and he provided Harry with the number of calls received on each day of the week over the last 13 weeks. The data (refer to Cutting Edge Student File No. 1) begins in week 44 of the last year (2012) and continues to week 5 of the current year (2013). Mark indicates that the days where no calls were received were holidays. As a start, Harry used the data from the past 13 weeks and applied five different time-series forecasting methods in preparing a trial forecast of the call volume for each day of the upcoming week (Week 6). He provided a different forecast for each day of the week by treating the forecast for a single day as being the actual call volume on that day. From plotting the data, Harry could see that demand follows “seasonal” patterns within the week. For example, more employees call at the beginning of the week when they are fresh and productive than at the end of the week when they are planning for the weekend. Therefore, Mark prepared and used seasonally adjusted call volumes for the past 13 weeks. After Week 6 ended, Harry compared the five forecasts with the actual volumes and calculated the Mean Absolute Deviation (MAD) values for each method. The result of Harry’s work is summarized below:

Cutting Edge Week 6 Forecast vs. Actual Daily Call Volume

Answer Questions 2a through 2e below: Question 2: Describe the details of each forecasting method used by Harry and explain its accuracy (MAD value) in comparison with the accuracy of the other methods. (Hint: In answering this question, it is helpful to review a time-series plot of the 13 weeks of data.)

5

2a) Last Value The last value method focus solely on the last value of the time series. Is the easiest method to be uses and statistician naïve because this method is based only on one sample size even when there is more data available. This method works well with economic and financial time series. When comparing to other methods is not the best method but is also not the best. As shown in the table , its MAD accurate with a deviation on only 171 calls of diference when comparing forecasting vs. actual call volume.

2b) Averaging The averaging method uses all data points from the time series to find the mean. It is usually used under stable conditions where the data from the time series is pretty similar at each point. This method is also good when doing a cross-sectional data analyzes- which help us predict a value that is not included in the data. In terms of accuracy MAD value, for this case is the worst method to be uses because of the variation (big variation/ not a stable condition) in the amount of calls between each day which causes the forecasting to bee way off. As shown in the table when using the averagin method there is a 399 MAD calls of difference when comparing forecasting vs. actual call volume.

2c) Moving Average (5 days) The moving average method is a bit similar to the averaging method expect, this method only averages the data for the most recently period (for the last 5 days in this particular case). This method is usually used for short range forecasts or to forecast the demand of a product with no trend during its maturity stage. The moving average method is more accurate than all other methods used in this case because its MAD value accuracy shows the smallest deviation- 104 calls of difference when comparing forecasting vs. actual call volume.

2d) Exponential Smoothing (alpha = 0.1) In the exponential smoothing method the weight assigned to the time series exponentially decays. In another words this method places most of the weight on the last value and the rest of the weight on the remaning values that came before the last. Its forecast is a weighted average of the actual volume calls from previous period and the forecast from the previous period where alpha (0.1 for this case) is the weight that is applied to the actual call volume for the pervious period. In terms of MAD

6

value accuracy for this case, this method is in 4 th place out of 5 with 184 calls of diference when comparing forecasting vs. actual call volume.

2e) Exponential Smoothing (alpha = 0.5) By increasing alpha the weight placed on the last value is way greater than the weight placed on the values before the last causing the last value to become more relavant. Also, due to the fact that changes on the actual call volumes occur frequently by increasing the alpha it results on a more accurate MAD value and that’s why the MAD value for the ES (alpha=0.5- 104 calls of difference) is way better than the MAD value for the ES (alpha=0.1- 184 calls of difference).

After many months of work and with Harry’s help, Mark has been able to stabilize the call center operation. Mark now has a better handle on how to forecast the daily call demand and he is able to prepare effective weekly staffing schedules for handling the daily variation in volume. However, Mark is still experiencing difficulty in forecasting the volume from month to month. Cutting Edge has been very active in acquiring new companies while, at the same time, selling off portions of their existing business. Mark believes that this activity is causing fluctuations in call volume because it is affecting the employee head count of Cutting Edge. Mark has assembled monthly data for call volume and head count for the past 18 months (refer to Cutting Edge Student File No. 2). Mark also suspects that there are other factors which may be affecting the call volume, and he has noted these factors on the attached spreadsheet. Based on the upcoming acquisition of Cutter Corp on 7/1/2015, the forecast of head count for July 2015 is 77,000.

Answer Questions 3a through 3d below: Question 3a: Prepare a forecast of call volume for July 2015 by applying Exponential Smoothing (with alpha = 0.5) to the prior 18 months of data. Use the appropriate Excel template from the Hillier text to prepare your forecast and assume that initial call volume is 24,000. Show your forecast below and attach the completed template. Call Volume Forecast for July 2015 (Exponential Smoothing, alpha=0.5): ______34,035.00___________

7

July 2015 forecast

Question 3b: Apply Linear Regression to predict call volume from head count using the appropriate Excel template. Show your forecast below and attach the completed Excel template. Call Volume Forecast for July 2015 (Causal Forecasting based on head count): ____34,000.87______

Question 3c: Calculate the Mean absolute deviation value of the Exponential Smoothing model (Question 3a) and the Average Estimation Error of the Linear Regression model (Question 3b). Explain the difference between these two values. Mean absolute deviation of Exponential Smoothing model, alpha=0.5: _______1,361.7_______________ Average Estimation Error for Causal Forecasting model based on headcount: _____483.85_________ Explanation of the difference in values: the diference in values is due to the fact that the MAD value only take into consideratio the volume of calls and it tells us how far off is the actual call volume from the forcasted call volumes; while the average estimation error, is based on the volume calls and the headcount (number of employees). By using the linea`r regression model we are able to take into 8

consideration the peek days where they get more calls and the changes in the staff which allow us to have a more acurate forecast.

Question 3d: Considering your answers to Questions 3a, 3b and 3c and all the factors that have been described above, prepare your best forecast for July 2015. Show your forecast value below and explain and justify how you came up with this forecast. Call Volume Forecast for July 2015 (My forecast): _____34,000.87_______ Explanation and Justification of Your Method: The method that I chose was the Average Estimation Error for Causal Forecasting model based on headcount. As I’ve mentioned before this method is more accurate than any other method discussed in class because this method is based on the volume calls and the headcount (number of employees). By using the linear regression model we are able to take into consideration the peek days where they get more calls and the changes in the staff which allow us to have a more acurate forecast. I came up with this result by copying the available actual calls volume data and the employee head count data into the Template for Linear Regration (independent variables and dependent variables, respectively). Then I calculated the actual call volume by using the Linear regration formula line where y= a + bx where a=-11,368.67, b= 0.59, and x=77,000 which gave me the result showed above.

9

Bibliography Hillier, F. S., & Hillier, M. S. (2019). Introduction to management science: a modeling and case studies approach with spreadsheets. New York, NY: McGraw-Hill Education. Hiller, Frederick S. & Mark S. Hillier (2014). Introduction to Management Science: A Modeling and Case Studies Approach with Spreadsheets, 5th ed., McGraw-Hill/Irwin, pp 429-432. Oracle. (2016, November 14). JD Edwards EnterpriseOne Applications Forecast Management Implementation Guide. Retrieved February 28, 2018, from https://docs.oracle.com/cd/E16582_01/doc.91/e15111/und_forecast_levels_methods.htm#EOAF M00177

10...