BUS 250
Case Study
Presented by:-
Shubham Singh
Presented to:-
Prof. Soumitra Mookherjee
Harmon Foods
• John MacIntye, manager of Harmon Foods, wants to
forecast the sales of Treat, a ready-to-eat breakfast.
• The difficulty arises from the wide variability in sales.
Actual sales varied from 50% to 200% of his forecast.
Accurate forecasts are essential for the business.
• Huge variations in the forecasting of sales for the Treat product of
Harmon Foods, a ready-to-eat breakfast cereal allowed the
company to develop a robust forecasting technique that could
help them in their forecasting.
• The company was also facing inventory problems, increase in their
cost of production, and the profitability of the company became
low.
• The advertisement and the marketing cost is also increased as a
result of poor forecasting techniques adopted by the company.
The company must develop a robust sales
forecasting technique which can allow the company
in decoding the actual future sales
To avoid the Problem
Key Factors affecting the Demand and Shipments
• Market Share: It is expected that the market share for Treat will
remain steady in the market, and as well as in the commodity group
as it was discovered in a research conducted by A.C Nielsen.
• Seasonal Factors: The sales for Treat varied in different seasons as
due to the vacations of their sales personnel in the summer season
and plant shutdowns, the sales were comparatively lower than
expected.
• Non-Media Promotions: This factor has reasonably performed well
for the company as it has strong influence over the company’s
product. This strategy includes discounts that are passed to
consumers and the dealers so that they could promote the Treat
product through monetary benefits.
Consumer Packs
• Consumer Packs were usually a 20-cent-per-package reduction.
• Sales are favorably affected during the month in which the
packs were shipped.
• This resulted in inventory build-ups by stores and consumers.
• Hence, the consumer packs might have a negative influence in
subsequent months.
• It also has negative effect on the company as the company had
excess inventories that resulted in the failure of selling those.
Dealer Allowances
• Dealers promoted Treat by using spectacular end-of-aisle
displays, ads, coupons, fliers, etc.
• Much of the sales increase also lead to inventory build-ups
which may last up to two months.
• Dealer allowances are promotions to push dealer to sell their
products.
Factors for sales forecasting
• Competitors’ Strategies: The Company also needs to have a close watch on its competitors’
promotional strategies and their pricing because this will help in devising a proper strategy for
the company and will affect its sales. The brand managers of Treat has taken the necessary
information of their competitors which will help the company in forecasting of its sales.
• Substitute and Compliment Products: The price variation in the substitute and the compliment
products for Treat which probably includes milk and oatmeal, it can either positively or
negatively affect the sales for the company.
• Changes in Income Levels: The variations in income level also affects the sales of the company
because a target segment of Treat also target those customer which are price-sensitive which
would be greatly affected by this move.
Budget and Control
• Contribution to division overhead and profits,
Long term dividend policy and corporate
expansion plan was forecast
• Spurious accounting profit in the quarters
where sales exceed forecast , to decrease
overspending on advertising .
• After taking these steps in fourth quarter
they have profit of $4 million dollars .
The Brand Manager
• Donald Carswell Who was Brand manager
for Budget control , his idea was to
forecast the governed monthly advertising
and promotional expenditure
• Donald carswell furnished knowledge of
market trends for their brand and their
brand’s competitors , however he also
keeps report of market research for be
aware of new products , package design
and product formulation .
• Donald Carswell Find the solution for
companywide application.
Advertising expenditure
• Inaccurate sales forecast also lead to
ineffectiveness of treats advertising
expenditure.
• They advertised on morning kids shows
because they believed it was most effective.
• Ad slots have to be booked a year In
advance so when the sales forecast figures
are inaccurate additional advertising time is
required which means other less effective
times and Tv shows has to be chosen .
• Which not only incurs extra cost but the
effectiveness of ads also decrease
Factors Affecting Sales
• “Caseship” is the monthly average Sales of Treat measured in
cases (24 packs.)
• “Sindex” is the Seasonal Factors (one for each month)
• Conpacks” is the history of Consumer Packs measured in cases
(24 packs).
• “DealAl” is the actual expenditure in Dealer Allowances
measured in dollars.
• Competitive advertising and prices also affect sales. They are
considered unpredictable and, hence, left out of the regression.
• Sales of Treat (in cases) 17
Seasonality Index
60
70
80
90
100
110
120
130
SeasonalityIndex
Month
Figure 2
Seasonality Index
Question 1
Using only data on monthly shipments of Treat, provide a forecast
for shipments of Treat in January, 1988. Give a 95% prediction
interval for this forecast.
Question 1
• One solution is to use the average of past monthly shipments
(382,522) as a forecast and then build a 95% PI around it.
• Another way is to run a regression on a time index (Tindex, a
variable running from 1 to 48).
Question 1
• Switch to the prediction.
• Enter 49 as the value for prediction for Tindex.
• Our forecast for January, 1988 is 414,137
• The 95% PI is (160,089 , 668,186). This is a wide PI !
Question 2
Use your estimated model to comment on the
impact/effectiveness of consumer promotions and dealer
promotions.
Question 2
• Each additional consumer pack sent out will increase
deseasonalized sales of Treat by an average of 0.402 cases in
that month.
• But will decrease average deseasonalized sales by 0.192 next
month and 0.0023 two months ahead.
• The overall effect is to increase average deseasonalized sales by
0.402 - 0.192 - 0.0023 = 0.2077 cases.
Question 2
• Each additional dollar of dealer allowances is estimated to
increase average deseasonalized sales by 0.066 + 0.0066 -
0.0155 = 0.0578 cases.
• Consider spending an additional dollar on dealer allowances in
January.
• The predicted effect is 0.066 * 1.13 + 0.0066 * 0.98 - 0.0155 *
1.02 = 0.06609 cases
Question 2
• 1 case = 24 packs = $4.80 dollars discount. (each pack is 20
cents discount)
• The 0.2077 increase in cases corresponds to 0.2077 / 4.80 =
0.04327 increase in dollars.
• Dealer allowances are more effective.
Question 3
What improvements, if any, would you recommend to the product
manager in terms of the timing and amounts of dealer promotions
and consumer promotions in the future?
Question 3
• Promotions are more effective if done in months with high
seasonal indices followed by months with lower indices.
• Dealer allowances seems, dollar for dollar, to be more effective
than consumer packs.
• November is an exception.
Recommendations
Case Shipments = -80,315.68 + 946.56T + 3877.95S + .096C + .076D - .048CL - .02DL-2
Where
T - Time in months since Dec 1983
S - Seasonal Index as given by the case
C - Consumer Pack spending for the current month (packs*$.2 discount *24 packages/case)
D - Dealer Allowance for the current month
CL- Consumer Pack spending lagged from 1 month ago (packs*$.2 discount *24 packages/case)
DL-2- Dealer Allowance lagged from 2 months ago
We assume that one consumer pack includes 24 coupons (a case) at a value of $.20/ea. Based on
information in the case, we weighted the consumer packs and dealer allowance to be equal to 90% of the
current month and 5% from each of the two previous months. Also, we assumed that these allowances
would immediately increase sales and then decrease future sales because consumers and dealers would
stock up during canvassing periods. Therefore, we included values for lagged effects of previous consumer
packs and dealer allowances.
Harmon food

Harmon food

  • 1.
    BUS 250 Case Study Presentedby:- Shubham Singh Presented to:- Prof. Soumitra Mookherjee
  • 2.
    Harmon Foods • JohnMacIntye, manager of Harmon Foods, wants to forecast the sales of Treat, a ready-to-eat breakfast. • The difficulty arises from the wide variability in sales. Actual sales varied from 50% to 200% of his forecast. Accurate forecasts are essential for the business.
  • 3.
    • Huge variationsin the forecasting of sales for the Treat product of Harmon Foods, a ready-to-eat breakfast cereal allowed the company to develop a robust forecasting technique that could help them in their forecasting. • The company was also facing inventory problems, increase in their cost of production, and the profitability of the company became low. • The advertisement and the marketing cost is also increased as a result of poor forecasting techniques adopted by the company.
  • 4.
    The company mustdevelop a robust sales forecasting technique which can allow the company in decoding the actual future sales To avoid the Problem
  • 5.
    Key Factors affectingthe Demand and Shipments • Market Share: It is expected that the market share for Treat will remain steady in the market, and as well as in the commodity group as it was discovered in a research conducted by A.C Nielsen. • Seasonal Factors: The sales for Treat varied in different seasons as due to the vacations of their sales personnel in the summer season and plant shutdowns, the sales were comparatively lower than expected. • Non-Media Promotions: This factor has reasonably performed well for the company as it has strong influence over the company’s product. This strategy includes discounts that are passed to consumers and the dealers so that they could promote the Treat product through monetary benefits.
  • 6.
    Consumer Packs • ConsumerPacks were usually a 20-cent-per-package reduction. • Sales are favorably affected during the month in which the packs were shipped. • This resulted in inventory build-ups by stores and consumers. • Hence, the consumer packs might have a negative influence in subsequent months. • It also has negative effect on the company as the company had excess inventories that resulted in the failure of selling those.
  • 7.
    Dealer Allowances • Dealerspromoted Treat by using spectacular end-of-aisle displays, ads, coupons, fliers, etc. • Much of the sales increase also lead to inventory build-ups which may last up to two months. • Dealer allowances are promotions to push dealer to sell their products.
  • 8.
    Factors for salesforecasting • Competitors’ Strategies: The Company also needs to have a close watch on its competitors’ promotional strategies and their pricing because this will help in devising a proper strategy for the company and will affect its sales. The brand managers of Treat has taken the necessary information of their competitors which will help the company in forecasting of its sales. • Substitute and Compliment Products: The price variation in the substitute and the compliment products for Treat which probably includes milk and oatmeal, it can either positively or negatively affect the sales for the company. • Changes in Income Levels: The variations in income level also affects the sales of the company because a target segment of Treat also target those customer which are price-sensitive which would be greatly affected by this move.
  • 9.
    Budget and Control •Contribution to division overhead and profits, Long term dividend policy and corporate expansion plan was forecast • Spurious accounting profit in the quarters where sales exceed forecast , to decrease overspending on advertising . • After taking these steps in fourth quarter they have profit of $4 million dollars .
  • 10.
    The Brand Manager •Donald Carswell Who was Brand manager for Budget control , his idea was to forecast the governed monthly advertising and promotional expenditure • Donald carswell furnished knowledge of market trends for their brand and their brand’s competitors , however he also keeps report of market research for be aware of new products , package design and product formulation . • Donald Carswell Find the solution for companywide application.
  • 11.
    Advertising expenditure • Inaccuratesales forecast also lead to ineffectiveness of treats advertising expenditure. • They advertised on morning kids shows because they believed it was most effective. • Ad slots have to be booked a year In advance so when the sales forecast figures are inaccurate additional advertising time is required which means other less effective times and Tv shows has to be chosen . • Which not only incurs extra cost but the effectiveness of ads also decrease
  • 12.
    Factors Affecting Sales •“Caseship” is the monthly average Sales of Treat measured in cases (24 packs.) • “Sindex” is the Seasonal Factors (one for each month) • Conpacks” is the history of Consumer Packs measured in cases (24 packs). • “DealAl” is the actual expenditure in Dealer Allowances measured in dollars. • Competitive advertising and prices also affect sales. They are considered unpredictable and, hence, left out of the regression.
  • 13.
    • Sales ofTreat (in cases) 17
  • 14.
  • 15.
    Question 1 Using onlydata on monthly shipments of Treat, provide a forecast for shipments of Treat in January, 1988. Give a 95% prediction interval for this forecast.
  • 16.
    Question 1 • Onesolution is to use the average of past monthly shipments (382,522) as a forecast and then build a 95% PI around it. • Another way is to run a regression on a time index (Tindex, a variable running from 1 to 48).
  • 18.
    Question 1 • Switchto the prediction. • Enter 49 as the value for prediction for Tindex. • Our forecast for January, 1988 is 414,137 • The 95% PI is (160,089 , 668,186). This is a wide PI !
  • 19.
    Question 2 Use yourestimated model to comment on the impact/effectiveness of consumer promotions and dealer promotions.
  • 20.
    Question 2 • Eachadditional consumer pack sent out will increase deseasonalized sales of Treat by an average of 0.402 cases in that month. • But will decrease average deseasonalized sales by 0.192 next month and 0.0023 two months ahead. • The overall effect is to increase average deseasonalized sales by 0.402 - 0.192 - 0.0023 = 0.2077 cases.
  • 21.
    Question 2 • Eachadditional dollar of dealer allowances is estimated to increase average deseasonalized sales by 0.066 + 0.0066 - 0.0155 = 0.0578 cases. • Consider spending an additional dollar on dealer allowances in January. • The predicted effect is 0.066 * 1.13 + 0.0066 * 0.98 - 0.0155 * 1.02 = 0.06609 cases
  • 22.
    Question 2 • 1case = 24 packs = $4.80 dollars discount. (each pack is 20 cents discount) • The 0.2077 increase in cases corresponds to 0.2077 / 4.80 = 0.04327 increase in dollars. • Dealer allowances are more effective.
  • 23.
    Question 3 What improvements,if any, would you recommend to the product manager in terms of the timing and amounts of dealer promotions and consumer promotions in the future?
  • 24.
    Question 3 • Promotionsare more effective if done in months with high seasonal indices followed by months with lower indices. • Dealer allowances seems, dollar for dollar, to be more effective than consumer packs. • November is an exception.
  • 25.
    Recommendations Case Shipments =-80,315.68 + 946.56T + 3877.95S + .096C + .076D - .048CL - .02DL-2 Where T - Time in months since Dec 1983 S - Seasonal Index as given by the case C - Consumer Pack spending for the current month (packs*$.2 discount *24 packages/case) D - Dealer Allowance for the current month CL- Consumer Pack spending lagged from 1 month ago (packs*$.2 discount *24 packages/case) DL-2- Dealer Allowance lagged from 2 months ago We assume that one consumer pack includes 24 coupons (a case) at a value of $.20/ea. Based on information in the case, we weighted the consumer packs and dealer allowance to be equal to 90% of the current month and 5% from each of the two previous months. Also, we assumed that these allowances would immediately increase sales and then decrease future sales because consumers and dealers would stock up during canvassing periods. Therefore, we included values for lagged effects of previous consumer packs and dealer allowances.