forecast Chapter case study

• DR/ MAGDY ABDELFATTAH

• CAIRO UNIVERSITY
• MBA / FIRST SEMESTER 2018
• PRAPARED BY
• 1- AHMED ZALAT
• 2- MOHAMED SAMIR
• 4- ABDELHAMED
• 5- ELHAYTHAM ABDALLAH
 Southwestern University: (B) *
Southwestern University (SWU), a large state college in Stephenville, Texas,
enrolls close to 20,000 students. The school is a dominant force in the small city, with more
students during fall and spring than permanent residents. Always a football powerhouse,
SWU is usually in the top 20in college football rankings. Since the legendary Phil Flamm was
hired as its head coach in 2009 (in hopes of reaching the elusive number 1 ranking),
attendance at the five Saturday home games
each year increased. Prior to Flamm’s arrival, attendance generally
averaged 25,000 to 29,000 per game. Season ticket sales bumped up
by 10,000 just with the announcement of the new coach’s arrival.
Stephenville and SWU were ready to move to the big time! The immediate issue facing SWU,
however, was not NCAA ranking. It was capacity. The existing SWU stadium, built in 1953, has
seating for 54,000 fans. The following table indicates
attendance at each game for the past 6 years.
Southwestern University: (B) *
One of Flamm’s demands upon joining SWU had been a stadium expansion, or possibly
even a new stadium. With attendance
increasing, SWU administrators began to face the issue head-on.
Flamm had wanted dormitories solely for his athletes in the stadium as an additional
feature of any expansion.
SWU’s president, Dr. Joel Wisner, decided it was time for
his vice president of development to forecast when the existing
stadium would “max out.” The expansion was, in his mind, a
given. But Wisner needed to know how long he could wait. He
also sought a revenue projection, assuming an average ticket
price of $50 in 2016 and a 5% increase each year in future prices.
Southwestern University: (B) *
Discussion Questions
1. Develop a forecasting model, justifying its selection over other techniques,
and project attendance through 2017.
2. What revenues are to be expected in 2016 and 2017?
3. Discuss the school’s options.
Increasing trend
250

200 game
5
game
150 4
game
3
100

50

0
2010 2011 2012 2013 2014 2015
 Q1) develop forecasting model
As its appear to as we have a trend in increasing in no. of fans every year .
In this case we will use the trend projection method ( least square method ).
Step 1) identify the use of forecasting
We will forecast the no of fans attendance for southwestern university football through 2016
and 2017
Step 2 ) identify the items that should be used in Forecasting
We should determine the no of car parking , no security men force that should be
available
Step3 ) identify the time horizon
Because the forecasting will be for two years in the future so , its will be in the
long run
Q1) develop forecasting model
Step 4 ) identify which technique will used As we said before we will used the
time horizon approach especially the least square method
Step 5 ) gathering the data the forecasting:-The attendance for each game
will be as follow :
 Q1) develop forecasting model
Step5 ) gathering the data about the forecast
Tickets price for 2016= 50$
While there will a 5% increase in the price Tickets price in 2017 = 50 *1.05 = 52.5$ And the capacity of the
stadium is 54,000
Step6 ) make the forecast :-
As the least square method equation as follow :

Where :-
Y hat = computed value that will be predicted
A= y axis intercept
B= slope of the regression line
X =the independent variable
Q1) develop forecasting model
We will calculate the B variable by the following equation:

We will calculate the A variable by the following equation :-

Q1) develop forecasting model
We will forecasting the first game attendance :-

x*y x*x attendance (y) years (x)

34200 1 34200 1

72700 4 36100 2

107700 9 35900 3

167600 16 41900 4

212500 25 42500 5

281400 36 46900 6
∑XY=876100 ∑X*x = 91 ∑Y= 237500 ∑X =21
Q1) develop forecasting model
Now we will calculate the B variable for first game as follow :-

876,100 - 6*3.5*39,583
91-6*3.5*3.5
B=
AS (B) variable equal 2563.35we will calculate A variable as follow :-
A = 39,583.3 – 2563.35*3.5
A= 30,611.3

now will forecast the no of attendance for first game for year 2016 which represent year no. (7)
= 30,611.3 + 2563.35 *7
Attendance for year 2016 = 48,553
Attendance for year 2017 = 51,115
Q1) develop forecasting model
• After we calculating the attendance for game 1 we will calculate the
attendance for game (2) by the same way.

x*y x*x attendence (y) years (x)

39,800 1 39,800 1
80,400 4 40,200 2
139,500 9 46,500 3
184,400 16 46,100 4
241,000 25 48,200 5
300,600 36 50,100 6
∑XY=985,700 ∑X*x = 91 ∑Y= 270,900 ∑X =21
Q1) develop forecasting model
By repeating the last calculation we will find that
B variable = 2145.7
A variable = 37,640
Where the attendance for game 2 for year 2016 which is represent year no 7
=52,655
And for year 2017 which represent year no (8) = 54,800
Q1) develop forecasting model
After repeating the last calculations for the last years we identify the attendance
for the games as follow:-
2017 2016 GAME
51,115 48,553 1
54,800 52,655 2
49,420 47,860 3
38,966 37,023 4
55,628 52,482 5
Q1) develop forecasting model
STEP 7 ) Validate and implement the result
As its appear from the last figure we found that there is still a trend increasing
in the no. of attendance , however in year 2017 for game no.2 and no.3 the
attendance will excess the capacity of the stadium .
Q2) what revenue are to be expected in
year 2016 , 2017
As we mention before that the price of the
tickets will be in 2016 (50$)
Total attendance in year 2016 =238,573 fan
Total revenue for 2016 = 238,573 * 50
= 11,928,650 $
Ticket price for 2017 = 52.5 $
Total attendance for year 2017 =249,929 fan
Total revenue = 249,929 * 52.5
= 13,058,791 $
Q3) what is the school options
Because of the excess of the attendance in games no. 2 and no.5 for
year 2017 the school should start to solve this problem While it has
three options .
First option :
In these two games we can transfer the game into other stadium that
still in Stephenville , Texas that could has bigger capacity and can hold
the no. of attendance , but after search we found that there is one
stadium which called MEMORIAL STADIUM , which has capacity only
for 7,000 fans
Q3) what is the school options
Second option : prepare big display screens out of stadium to
be close to the festival , but as the layout of the stadium show
there is no enough space out of the stadium and it will take
apart from the parking .
Q3) what is the school options
Third option :
build anew extension As we are in 2015 and the excess
of attendance will be in 2017 , we can build anew
extension for the stadium and will cover the excess of
the attendance .
day of the week, or time of the season. If an upper-deck seat sold for $10 in the
first game of the year, it likewise sold for $10 for every game. But when Anthony
Perez, director of business strategy, finished his MBA at the University of Florida,
he developed a valuable database of ticket sales. Analysis of the data led him to
build a forecasting model he hoped would increase ticket revenue. Perez
hypothesized that selling a ticket for similar seats should differ based on
demand.
Studying individual sales of Magic tickets on the open Stub Hub marketplace
during the prior season, Perez determined the additional potential sales revenue
the Magic could have made had they charged prices the fans had proven they
were willing to pay on Stub Hub. This became his dependent variable, y , in a
multiple-regression model. He also found that three variables would help him
build the “true market” seat price for every game. With his model, it was
possible that the same seat in the arena would have as many as seven different
prices created at season onset—sometimes higher than expected on average and
sometimes lower. The major factors he found to be statistically significant in
determining how high the demand for a game ticket, and hence, its price, would
be were:
◆ The day of the week ( x 1 )
◆ A rating of how popular the opponent was ( x 2 )
◆ The time of the year ( x 3 )
Forecasting Ticket Revenue for Orlando
Magic Basketball Games
• For the day of the week, Perez found that Mondays were the least-favored game days (and he assigned
them a value of 1). The rest of the weekdays increased in popularity, up to a Saturday game, which he
rated a 6. Sundays and Fridays received 5 ratings, and holidays a 3 (refer to the footnote in Table 4.3 ).
His ratings of opponents, done just before the start of the season, were subjective and range from a low
of 0 to a high of 8. A very high-rated team in that particular season may have had one or more
superstars on its roster, or have won the NBA finals the prior season, making it a popular fan draw.
Finally, Perez believed that the NBA season could be divided into four periods in popularity:
◆ Early games (which he assigned 0 scores)
◆ Games during the Christmas season (assigned a 3)
◆ Games until the All-Star break (given a 2)
◆ Games leading into the play-ofs (scored with a 3)
The first year Perez built his multiple-regression model, the dependent variable y , which was a
“potential premium revenue score,” yielded an r 2 = .86 with this equation: y = 14,996 + 10,801x1 +
23,397x2 + 10,784x3 Table 4.3 illustrates, for brevity in this case study, a sample of 12 games that year
(out of the total 41 home game regular season), including the potential extra revenue per game ( y ) to
be expected using the variable pricing model. A leader in NBA variable pricing, the Orlando Magic have
learned that regression analysis is indeed a profitable forecasting tool.
1. Use the data in Table to build a regression model
with day of the week as the only independent variable.

y= a + bx 1

Where :
Y is the dependent variable (addition sales potential)
A is the y-axis intercept
B is the slop of the regression line
X1 is the dependent variable (Day rating)
 Day Day rating (X1) Add. Sales potential x2 xy
(y)
Wednesday 3 $12,331 9 36,993
Friday
Friday 5
5 $29,004
$29,004 25
25 145,020
145,020
Wednesday
Wednesday 3
3 $109,412
$109,412 9
9 328,236
328,236
Wednesday
Wednesday 3
3 $75,783
$75,783 9
9 227,349
227,349
Wednesday 3 $42,557 9 127,671
Wednesday 3 $42,557 9 127,671
Thursday 4 $120,212 16 480,848
Thursday 4 $120,212 16 480,848
Monday 1 $20,459 1 20,459
Monday 1 $20,459 1 20,459
Sunday 5 $231,020 25 1,155,100
Sunday 5 $231,020 25 1,155,100
Wednesday 3 $28,455 9 85,365
Wednesday 3 $28,455 9 85,365
Sunday 5 $110,561 25 552,805
Sunday 5 $110,561 25 552,805
Friday 5 $44,971 25 224,855
Friday
Wednesday 5
3 $44,971
$30,257 25
9 224,855
90,771
Wednesday
Total 3
∑x=43 $30,257
∑y=855,022 9
∑=171 90,771
∑xy=3,475,472
Total ∑x=43
=3.58 ∑y=855,022
=71.252 ∑xy=3,475,472

 
b= =24,098 a = -b = -15,018
y= -15,018 + 24,098 X
 y= a + bx
250

200

150
Sales *1000

100

50

0
0 1 2 3 4 5 6 7
Day Rating
2. Use the data to build a model with rating of the
opponent as the sole independent variable.

y= a + bx 2

Where :
Y is the dependent variable (addition sales potential)
A is the y-axis intercept
B is the slop of the regression line
X2 is the dependent variable (opponent rating)
 Team Team rating Add. Sales x2 xy
(X2) potential (y)
Phoenix Suns 0 $12,331 0 0
Detroit
Detroit Pistons
Pistons 1
1 $29,004
$29,004 1
1 29,004
29,004
Cleveland
Cleveland 6
6 $109,412
$109,412 36
36 656,472
656,472
Miami
Miami Heat
Heat 3
3 $75,783
$75,783 9
9 227,349
227,349
Houston
Houston Rockets 2 $42,557 4 85,114
Rockets 2 $42,557 4 85,114
Boston Celtics 4 $120,212 16 480,848
Boston Celtics 4 $120,212 16 480,848
New Orleans 1 $20,459 1 20,459
New Orleans 1 $20,459 1 20,459
L. A. Lakers 8 $231,020 64 1,848,160
L. A. Lakers 8 $231,020 64 1,848,160
San Antonio 1 $28,455 1 28,455
San Antonio 1 $28,455 1 28,455
Denver Nuggets 1 $110,561 1 110,561
Denver Nuggets 1 $110,561 1 110,561
NY Knicks 0 $44,971 0 0
NY Knicks
Philadelphia 0
1 $44,971
$30,257 0
1 0
30,257
Philadelphia
Total 1
∑x=28 $30,257
∑y=855,022 1
∑=134 30,257
∑xy=3,516,679
Total ∑x=28
=2.33 ∑y=855,022
=71,252 ∑xy=3,516,679

 
b= =22,141 a = -b = 19,661
y= 19,661 + 22,141 X
 120

100

80

Sales *1000 60

40

20

0
0 1 2 3 4 5 6 7

opponent Rating
 3. Using Perez’s multiple-regression model, what
would be the additional sales potential of a Thursday
Miami Heat game played during the Christmas
holiday.

There are 3 independent variables X1,X2and X3

y = a + b1x1 + b2x2 +b3x3
Where
y = dependent variable, sales
a = a constant, the y intercept
x 1 and x 2 X3 = values of the two independent variables, Day rating ,
opponent rates, and season rate respectively
b1 , b 2 and b3= coefficients for the 3 independent variables “calculated”
Y= a + (24,098)(4)+ (22,141)(3)+(1,380.48)(3)
 date Time rating(X3) Add. Sales potential x2 xy
(y)
4
4 nov
nov 0
0 $12,331
$12,331 0
0 0
0
6
6 nov
nov 0
0 $29,004
$29,004 0
0 0
0
11
11 nov
nov 0
0 $109,412
$109,412 0
0 0
0
25 nov 0 $75,783 0 0
25 nov 0 $75,783 0 0
23 dec 3 $42,557 9 127,671
23 dec 3 $42,557 9 127,671
28 jan 1 $120,212 1 120,212
28 jan 1 $120,212 1 120,212
3 feb 1 $20,459 1 20,459
3 feb 1 $20,459 1 20,459
7 mar 2 $231,020 4 462,040
7 mar
17 mar 2
2 $231,020
$28,455 4
4 462,040
56,910
17
23 mar
mar 2
2 $28,455
$110,561 4
4 56,910
221,122
23 mar
9 april 2
3 $110,561
$44,971 4
9 221,122
134,913
9 april
14 april 3 $44,971
$30,257 9 134,913
90,771
14Total
april ∑x=17
3 ∑y=855,022
$30,257 ∑=41
9 ∑xy=1,234,098
90,771
Total =1.416
∑x=17 =71,252
∑y=855,022 ∑xy=1,234,098

 
b3= =1,380.48 a = -b = 69,297
y= 69,297 + 1,380.48 X3
 120

100

80

Sales *1000 60

40

20

0
0 1 2 3 4 5 6

Date Rating
4. What additional independent variables might
you suggest to include in Perez’s model.

• Average ages of the public.

• Gender
• Race
• Time of day
CASE STUDY
FORECASTING AT
HARD ROCK CAFÉ

PREPEARED BY/ ELHAYTHAM ABDALLAH

FORECASTING AT HARD ROCK CAFE
• With the growth of Hard Rock Cafe—from one pub in London in 1971 to more than
145 restaurants in 60 countries today—came a corporatewide demand for better
forecasting. Hard Rock uses long-range forecasting in setting a capacity plan and
intermediate-term forecasting for locking in contracts for leather goods (used in
jackets) and for such food items as beef, chicken, and pork. Its short-term sales
forecasts are conducted each month, by cafe, and then aggregated for a
headquarters view. The heart of the sales forecasting system is the point-of-
sale(POS) system, which, in efect, captures transaction data on nearly every person
who walks through a cafe’s door. The sale of each entrée represents one customer;
the entrée sales data are transmit-ted daily to the Orlando corporate headquarters’
database. There, the financial team, headed by Todd Lindsey, begins the forecast
process. Lindsey forecasts monthly guest counts, retail sales, ban- quet sales, and
concert sales (if applicable) at each cafe. The general managers of individual cafes
tap into the same database to prepare
FORECASTING AT HARD ROCK CAFE
• a daily forecast for their sites. A cafe manager pulls up prior years’ sales for that day,
adding information from the local Chamber of Commerce or Tourist Board on
upcoming events such as a major convention, sporting event, or concert in the city
where the cafe is located. The daily forecast is further broken into hourly sales, which
drives employee scheduling. An hourly forecast of $5,500 in sales translates into 19
workstations, which are further broken down into a specific number of waitstaf,
hosts, bartenders, and kitchen staf. Computerized scheduling software plugs in
people based on their availability. Variances between forecast and actual sales are
then examined to see why errors occurred.

• Hard Rock doesn’t limit its use of forecasting tools to sales . To evaluate managers and
set bonuses, a 3-year weighted moving average is applied to cafe sales. If cafe general
managers exceed their targets, a bonus is computed. Todd Lindsey, at corporate
headquarters, applies weights of 40% to the most recent year’s sales, 40% to the year
before, and 20% to sales 2 years ago in reaching his moving average.
FORECASTING AT HARD ROCK CAFE
• An even more sophisticated application of statistics is found in Hard Rock’s menu
planning. Using multiple regression, managers can compute the impact on demand
of other menu items if the price of one item is changed. For example, if the price of
a cheeseburger increases from $7.99 to $8.99, Hard Rock can predict the efect this
will have on sales of chicken sandwiches, pork sandwiches, and salads. Managers do
the same analysis on menu placement, with the center section driving higher sales
volumes. When an item such as a hamburger is moved of the center to one of the
side flaps, the corresponding efect on related items, say French fries, is determined
figures are used for purposes of this case study
1 2 3 4 5 6 7 8 9 10
month
Guest in 21 24 27 32 29 37 43 43 54 66
1000
Advertising 14 17 25 25 35 35 45 50 60 60
in 1000 $
 1- Describe three different forecasting applications at Hard Rock. Name three
other areas in which you think Hard Rock could
use forecasting models.

1- HARD ROCK CAFÉ USES LONG RANGE FORECAST IN SETTING A CAPACITYPLAN( NUMBERS
OF VISITORS IN DIFFERENT TIME SLOTS SUCH AS BREAKFAST, LUNCH ,AND DINNER )
2- USE MEDUIM RANGE FORECAST FOR LOCKING IN CONTRACT FOR LEATHER GOODS(UESD
IN JACKETS ) AND FOR SUCH FOOD ITEMS AS BEEF, CHICKEN, AND PORK ( THE PREFERANCE
FOR MEALS FOR A MORE SELECT PUPLIC VEGETARINES )
3 - USE SHORT RANGE FORECAST IN SALES FORECASTING (SALES FORECASTING ARE
CONDUCTED EACH MONTHELY BY CAFÉ AND THEN AGGREGATED FOR A HEADQUARTER VIEW
(STUDY OF PRODUCT FOR THE SERVICE AT THE TABLE SUCH NAPKINS & CUTLERY)
AND ALSO CAN USE FORECATING IN SALES BY GUEST , RETAIL SALES ,CONCERT SALES , MNUE
PLANNING, EVALUEATE MANANGERS, PRODUCT DEMAND AND SALES AT EACH WORK
STATION
2- What is the role of the POS system in forecasting at Hard
Rock?

• The heart of the sales forecasting system is the point-of-sale(POS)

system, which, in efect, captures transaction data on nearly every
person who walks through a cafe’s door. The sale of each entrée
represents one customer; the entrée sales data are transmit-ted daily
to the Orlando corporate headquarters’ database. There, the financial
team, headed by Todd Lindsey, begins the forecast process. Lindsey
forecasts monthly guest counts, retail sales, banquet sales, and
concert sales (if applicable) at each cafe. The general managers of
individual cafes tap into the same database to prepare
3. Justify the use of
the weighting system used for evaluating
managers for annual bonuses?
• Hard Rock doesn’t limit its use of forecasting tools to sales . To evaluate
managers and set bonuses, a 3-year weighted moving average is applied to cafe
sales. If cafe general managers exceed their targets, a bonus is computed. Todd
Lindsey, at corporate headquarters, applies weights of 40% to the most recent
year’s sales, 40% to the year before, and 20% to sales 2 years ago in reaching
his moving average.
• IF GENERAL MANAGER EXCEED THEIR TARGET A BOUNCE IS COMPUTED .WITH THIS SYSTEM
WE CAN FIND HOW EMPLOYEES FIND OUT HOW THEY ARE DOING AND WHAT IS EXPEXTED
FROM THEM . ALSO IN THREE YEARS SYSTEM WE CAN TRUCKING PERFORMANCE AND WE
CAN INTEGRATED IN TO BUDGETING PROCESS .
• THE WEIGHTING SYSTEM IS REASNABLE AND ACTUALLY PROTECT MANAGERS FROM LARGE
SALES VARIATION OUTSIDE THEY CONTROL
4- Name several variables besides those mentioned in the case that
could be used as good predictors of daily sales in each cafe

• OTHER VARIABLES CAN COULD BE UESD AS GOOD PREDICTORS OF DAILY SALES IN EACH
CAFÉ
• 1- COMPER SALES VOLUME FOR SIMILAR STORES IN SIMILAR LOCATION AN SIMILAR SIZE
• 2- FOCUS ON HOUSE HOLD INCOMES
• 3- HOLIDAYS AND TOURISM
• 4-HOW MUCH WILL THEY SPEND ON THESE ITEMS ANNUALY
• 5-SEASONSE OF YEARS
• 6- HOTEL OCCUPANCY
• 7- SPRING BREAK OF COLLEGES
• 8- BEEF PRICE
• 9- PROMOTIONAL BUDGET
5- At Hard Rock’s Moscow restaurant, the manager is trying to evaluate how a new advertising
campaign affects guest counts. Using data for the past 10 months (see the table), develop a
least-squares regression relationship and then forecast the expected guest count when
advertising is $65,000.

MONTH ADVERTISING (X) GUEST (Y) (X)2 (Y)2 XY

1 14 21 196 441 294

2 17 24 289 576 408
3 25 27 625 729 675
4 25 32 625 1024 800
5 35 29 1225 841 1015
6 35 37 1225 1369 1295
7 45 43 2025 1849 1935
8 50 43 2500 1849 2150
9 60 54 3600 2916 3240
10 60 66 3600 4356 3960
TOTAL ⅀ 366 376 15910 15950 15772
WE USE REGRESSION ANALYSIS BECAUSE THAT IS DESCRIBE THE FUNCTINONAL
RELATIONSHIPS BETWEEN INDEPENDENT AND DEPENDENT VARIABLES

•• = = 36,6
• = = 37,6

• þ = = 0,8
•ɑ = ȳ - þ Ẍ = 37,6 – 0,8 * 36,6 = 8,3

• Ŷ = ɑ +þ X = 8,3 + 0,8 *( 65) = 60,3 (60300guests )