Download - Bevarage Market Research
Beverages Market Research
Caroline Fearon, Sherry ChiuIMC 302
Research Question
R1: Which Holidays are most effective for Beverage Sales Nationally?
R2: Which Holidays are most effective for Coke Classic* Sales versus Pepsi sales in Dallas and Atlanta?
* Whenever Coke is mentioned in this presentation, we are referring to the sub-brand Coke Classic
MotivationWhy is the cause or focal independent variable important?
It is important to look at major U.S. holidays in order to determine how different seasonalities affect beverage sales. Advertisers need to know how effective their advertising campaigns are depending on the time of year. By comparing two different Markets (Atlanta & Dallas) and two different brands (Coke Classic & Pepsi) we hope to determine how Pepsi and Coke sales fluctuate during the holidays in these two different markets. This will help determine where to target advertising and at what time of year, in order to more strategically sell both Coke and Pepsi products.
Data● Holidays: We chose seven major U.S. Holidays and used data entries of beverage sales from
approximately 10 days before and 10 days after the following holidays-St. Patrick’s Day March 10th - March 17th - March 24thEaster March 26th - April 6 - April 15thMemorial Day May 15th - May 25 - June 4July 4th June 24 - July 4 - July 14thHalloween Oct 21st - Oct 31st - Nov 10thThanksgiving/Black Friday Nov 17th - Nov 27th - Dec 7thChristmas/New years Dec 15th - Dec 25th - Jan 4st
● Markets: Atlanta and Dallas; 104 Coke Classic Entries and 104 Pepsi Entries for both markets ● Time period: January 2002 to December 2003● Dependent Variable: Volume of Sales● Independent Variable: Seven Major U.S. Holidays
Model Fits Nationally
Nationwide
Linear 0.202
Semi-Log 0.165
Log-Log 0.216
Which functional form had the best fit? - The log-log regression model showed the line of best fit since the R squared
variable was closest to 1 compared to all other models. The log-log model accounted for the greatest percent of sales.
Model
Results: Nationally
Results: Coke Classic in Atlanta
Results: Coke Classic in Dallas
Results: Pepsi in Atlanta
Results: Pepsi in Dallas
Summary of Findings - Nationally, we found that there is a significant increase in beverage sales
during Christmas/New Years (by .095%) and the 4th of July (by .179%)
- In Atlanta, we found that there is a significant increase in Coke Classic sales during Memorial Day (by .254%), Thanksgiving/Black Friday (by .198%), and Christmas/New Years (by .292%).
- In Dallas, we found that there is a significant increase in Coke Classic sales during Christmas/New Years (by .217%).
- In Atlanta, we found that there is a significant increase in Pepsi sales during Memorial Day (by .272%).
- In Dallas, we found that that there is no significant increase in Pepsi sales for any of the seven holidays.
How Holidays affect Revenue
Coke Classic Sales in Atlanta
Average Revenue: $1,638,224.43
Memorial Day: Revenue increases by $416,109.01
Thanksgiving/Black Friday: Revenue increases by $324,368.44
Christmas/New Years: Revenue increases by $478,361.53
Coke Classic Sales in Dallas
Average Revenue: $831,927.642
Christmas/New Years: Revenue increases by $180,528.30
Pepsi Sales in Atlanta
Average Revenue: $423,398.11
Memorial Day: Revenue increases by $115,164.29
Pepsi Sales in Dallas
Average Revenue: $287,952.50
Managerial ImplicationsCoke Classic:
● Focus on promoting sales around Christmas/New Years in both Atlanta and Dallas
● Also focus efforts around Memorial Day and Thanksgiving/Black Friday in Atlanta
Pepsi: ● Focus on promoting sales around
Memorial Day in Atlanta● In Dallas, no holiday significantly
affects Pepsi sales. Thus Pepsi should focus on promoting their overall brand in Dallas
Both:● Generally, Atlanta accrues more revenue than Dallas on and off holiday seasons● Both should focus on promoting sales in Atlanta
Atlanta Demographics
Population: 443,775
Median Age: 33.2
Median Income: $46,466
Unemployed: 14.7%
Household: Average size is 2.1 people, 43.4% are family households and 7.7% are comprised of unmarried partners
Dallas Demographics
Population: 1,241,162
Median Age: 31.8
Median Income: $41,354
Unemployed: 9.3%
Household: Average size is 2.6 people, 58% are family households and 7.3% are comprised of unmarried partners
Possible Demographic Interpretation
● Demographics:○ Atlanta= older households, less family households, slightly higher income, and
higher unemployment rates, predominantly Black○ Dallas = younger households, more family households, slightly lower income,
and lower unemployment rates, predominantly Hispanic ● There is a possible connection between race and beverage consumption. On
Memorial Day, Thanksgiving and Christmas in Atlanta (which is predominantly black) coke consumption increases. However, in Dallas which is predominantly Hispanic (whom are historically Christian) Coke Classic consumption only increases during Christmas/New Years
● Coke is headquartered in Atlanta, which may be the cause for their high Coke Classic consumption
Limitations of Study We advise interpreting our results with caution because…
- Our R-squared variables were low. - Most marketing R^2 variables are between .3-.7 and ours were around .2
- Coke is headquartered in Atlanta which may have affected Coke Classic sales in that area- The time frame of the demographic information does not perfectly align with the time frame of the
sample data- The outliers in the Coke Classic entries for Dallas may be affecting the data- The cities have slightly different population size
- Atlanta Pop = 443,775- Dallas Pop = 1,241,162
- We did not account for minor holidays- The dates we chose for the time frame of each holiday was arbitrary
- We tried to define it 10 days before and after the actual holiday, but when creating dummy variables in Excel, date selection was not precise
- Interpretation of demographic information is based on inference- For Coke Classic in Atlanta, the significance of Thanksgiving/Black Friday on sales is 0.055, which is
only slightly greater than the set alpha of 0.05, so we decided to include it in our analysis
The End