TABLE OF CONTENTS 1.0 EXECUTIVE SUMMARY ...................................2 2.0 INTRODUCTION ................................4 3.0 EXPLORATORY ANALYSIS ................................5 3.1 DESCRIPTIVE STATISTICS ..............................5 3.2 INTERPRETATION OF THE GRAPHS ........................6 4.0 ANALYSIS OF THE REGRESSION OUTPUT ..................10 5.0 ANALYZING THE EFFECT OF A $1 REDUCTION IN AVERAGE PRICE (AVEP) ON SALES VOLUME ..................................17 6.0 ANALYZING THE EFFECT OF A $100,000 MEDIA CAMPAIGN SPLIT EVENLY OVER 10 WEEKS ....................................21 7.0 COMPARING PRICE PROMOTIONS AND ADVERTISING .........24 8.0 CONCLUSION ..............................27 9.0 APPENDIX A LINE GRAPHS 25 B SCATTER PLOTS 28 C DESCRIPTIVE STATISTICS 31 D 6 VARIABLE DATA 36 E 6 VARIABLE REGRESSION 39 F 6 VARIABLE HISTOGRAM 43 G 6 VARIABLES JARQUE BERA 44 H HETEROSCEDASTICITY 49 I 6 VARIABLES WHITE TEST REGRESSION 50 J 6 VARIABLES DURBIN WATSON TEST 55
Drink Me decision to which marketing activity to adopt
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
3.1 DESCRIPTIVE STATISTICS ..................................................................................53.2 INTERPRETATION OF THE GRAPHS .................................................................6
4.0 ANALYSIS OF THE REGRESSION OUTPUT ......................................................105.0 ANALYZING THE EFFECT OF A $1 REDUCTION IN AVERAGE PRICE ...........(AVEP) ON SALES VOLUME .......................................................................................176.0 ANALYZING THE EFFECT OF A $100,000 MEDIA CAMPAIGN SPLIT .............EVENLY OVER 10 WEEKS ...........................................................................................217.0 COMPARING PRICE PROMOTIONS AND ADVERTISING ..............................248.0 CONCLUSION ..........................................................................279.0 APPENDIX
A LINE GRAPHS 25B SCATTER PLOTS 28C DESCRIPTIVE STATISTICS 31D 6 VARIABLE DATA 36E 6 VARIABLE REGRESSION 39F 6 VARIABLE HISTOGRAM 43G 6 VARIABLES JARQUE BERA 44H HETEROSCEDASTICITY 49I 6 VARIABLES WHITE TEST REGRESSION 50J 6 VARIABLES DURBIN WATSON TEST 55K 3 VARIABLE DATA 58L 3 VARIABLE REGRESSION 61M 3 VARIABLE HISTOGRAM 65N 3 VARIABLES JARQUE BERA 660 3 VARIABLES WHITE TEST 69P 3 VARIABLES DURBIN WATSON TEST 74Q ANALYZING PRICE PROMOTIONS 77R ANALYZING MEDIA SPENDING 79
1.0 EXECUTIVE SUMMARY
DrinkMe, manufactured by GoodsCo, has to make a decision with regards to which
marketing activity to adopt. The report looks into the affect of Temporary Price
Reduction and Media Campaign over sales. Regression analysis has been done just to
find out that DrinkMe sales is not affected by the competitor’s price or their
advertisements and the main factors contributing to the sales of DrinkMe are DrinkMe’s
price, advertisement and the average temperature of Queensland. From the regression
output it has been noted that if average price increase by $1 volume would fall by 3.77
tonnes. For advertisements, an increase in advertisements would result in an increase in
sales volume by 0.063 tonnes and lastly an increase in temperature in Queensland would
result in a drop in sales by 0.61 tonnes.
The Jarque-Bera, White Test and Durbin-Watson test has been conducted to test for
normality, heteroscedasticity and serial correlation respectively. With a 95% confidence
level, all the tests have passed which concludes that the residuals are normally
distributed; there is no heteroscedasticity and no serial correlation.
With only three independent variables and when the assumptions of normality,
heteroscedasticity and serial correlation has been checked, the affect of price reduction on
sales has been worked out by holding all other variable constant. It came out that the
price reduction of 2.67 is the optimal point where the marginal difference is the
maximum. The marketing director is not recommended to reduce the price by 2.67 as it
would result in higher sales but an overall loss in the marketing activity. The marketing
director, depending on the outcome that is needed, needs to make a tradeoff decision
between high sales and low profit or vice versa. The media campaign on the other hand
provides the maximum marginal difference in the 10th week. The campaign is designed
only for 10 weeks but to make allowance for the residual effects of Adstock the analysis
is done over 20 weeks. The profit from this 10 week media campaign results to 13.84%.
It is recommended that the marketing director should use a combination of both price
reduction and media campaign. Both these would complement each other and the end
result would be more profitable than using only price reduction or media campaign.
2
The scope of the research is limited as outside the model competitors strategy does affect
DrinkMe’s sales and omitting them from the analysis would result in a different
conclusion. Whether the data collected represents the whole population is another
question that needs to be taken into account. Besides that, the research would assist the
marketing director in making the ultimate decisions regarding the selection between price
reductions and advertising promotions.
3
2.0 INTRODUCTION
This research is conducted to gain an insight into the effectiveness of a Temporary Price
Reduction (TPRs) activity and an Advertising Campaign (media spend) for the DrinkMe
brand in Queensland. The research will be able to prove whether a TPR or an increase in
media spent would be more effective.
Objectives
To measure the effects of a TPR of $1 on sales volume and sales value
To measure the effects of a media campaign worth $100,000 spread evenly over
10 weeks on DrinkMe’s sales volume and sales value
Data have been extracted from Electronic Point of Sale (EPOS) database dated 1st
September 2002 to 26th June 2005. The data includes DrinkMe and its competitor’s
weekly sales volume, weekly sales value, weekly distribution, weekly media spent and
the average temperature in Queensland.
For the purpose of the research, the data that has been collected is time series data as it
has been collected over a period of time. The measurement of the variables falls into the
category of ratio as the variables are conceptually quantitative measurement. The
numbers of units sold which in this case, volume, the dollar value of sales which in this
case value, are all conceptually quantified. The methodology that has been used is linear
regression. Multiple regressions have been used to find out the relationship between the
independent variables with the dependent variable which is volume in this case. The
affect of changes in 1 unit of independent variable on the dependent variable can be
determined from the regression summary output and in this case it is very important to
know this relationship as the ultimate objective is to find the effects of sales volume and
value due to the promotion of a $1 reduction in the price and the media campaign.
Moreover, as non normality of residuals, heteroscedasticity and serial correlation can
affect the explanation of the dependent variable accurately, Jarque-Bera test, White Test
and Durbin-Watson Tests are also done.
4
3.0 EXPLORATORY ANALYSIS
3.1 DESCRIPTIVE STATISTICS
A descriptive statistics was conducted and the following data was obtained. DrinkMe’s
sales volume shows an average of 14.35 tonnes while its competitor’s sales volume
shows an average of 0.36 tonnes..
DrinkMe’s sales value shows an average or mean of 131.11 ($’000) while its
competitor’s sales value shows a mean of 3.38 ($’000) DrinkMe’s sales value shows a
standard error 2.2893 and its competitor’s sales value has a standard error of 0.0817.
DrinkMe’s media spending has a mean of 2.2043 ($’000) while its competitor’s media
spending has a mean of 0.1421 ($’000). DrinkMe’s media spending has a standard error
of 0.3162 and its competitor’s media spending has a standard error of 0.0695.
DrinkMe’s average price has a mean of 9.1644 ($’000) while its competitor’s average
price has mean of 9.3959 ($’000). DrinkMe’s average price has a standard error of
0.0274 and its competitor’s average price has a mean of 0.0394.
DrinkMe’s Adstock has a mean of 4.3106 ($’000) while its competitor’s Adstock has a
mean of 0.2841 ($’000). DrinkMe’s Adstock has a standard error of 0.5173 and its
competitor’s Adstock has a standard error of 0.0958.
The average temperature in Queensland has a mean of 24.4179 degrees Celsius and has a
standard error of 0.3272. The maximum average temperature in Queensland was 30.4973
degrees Celsius while the minimum was 14.8676 degrees Celsius.
These averages will later be used in testing the effects of price reductions and media
spending.
5
3.2 INTERPRETATION OF THE GRAPHS
COMP VOL VS DM VOL
0
5
10
15
20
25
37500 37654 37808 37962 38116 38270 38424
TIME
VO
L VOLUME
COMP VOLUME
Graph 1: COMP Sales Volume vs DRINKME Sales Volume
By looking at Graph 1 it is found that DrinkMe’s sales are much higher compared to
DrinkMe’s competitor’s sales, in terms of volume. It is assumed that DrinkMe sales are
affected by price, temperature, media, competitor’s price and competitor’s media spent.
6
DM VOL VS AVE TEMP
0
5
10
15
20
25
30
35
37500 37654 37808 37962 38116 38270 38424
TIME
VO
L DM VOL
AVERAGE TEMP
Graph 2:
DRINKME Volume VS Average Temperature
Based on Graph 2, the temperature has a negative effect on sales, thus when temperature
increases, sales drop and vice versa. This relationship is also explained by the scatter plot
which is in Graph 3. The scatter plot shows that Drink Me volume has a negative
relationship with temperature.
DM VOL VS AVETEMP
0
5
10
15
20
25
0 10 20 30 40
DM VOL VS AVETEMP
Graph 3: DRINKME Sales Volume vs Queensland Average Temperature.
7
DM VOL VS DM ADVERTISING
0
5
10
15
20
25
30
9/1/
2002
12/1
/200
2
3/1/
2003
6/1/
2003
9/1/
2003
12/1
/200
3
3/1/
2004
6/1/
2004
9/1/
2004
12/1
/200
4
3/1/
2005
6/1/
2005
TIME
VO
L DM VOL
ADVERTISING
Graph 4: DRINKME Volume vs DRINKME Advertising
On the other hand, the effect of media expenditure has a positive effect on sales. Every
time the advertising has been done, there has been a delayed effect on the volume
Moreover, the effect of the media does help to increase sales and it is assumed that these
effects will have a 50% decay rate. The positive relationship is also clear in the scatter