Harmon Foods (A)

Sales Forecasting Models

Managerial Problem

Harmons management needs an accurate forecast of sales for Treat Cereal in order to efficiently manage production and inventory. Seriously underestimating demand could cause stock outs, poor customer relations and a loss of profitable business. Overestimating demand could create costly over production of Treat cereal.

A strong model will forecast sales with sufficient accuracy to aid management in planning. In short, your goal is to develop a model that minimizes forecasting error.

Modeling Challenge

1. Calibrate at least two forecasting models. Model 1 should forecast quarterly sales using two predictor variables, seasonality and a linear trend. This model is simple, easy to estimate and to apply. Model 2, a more complex multiple regression model, should use any combination of predictor variables that help to accurately forecast quarterly sales.

2. Evaluate the predictive power of each model. Does either model forecast sales with sufficient precision to help in the planning process? Examine several statistical measures of forecasting accuracy.

3. Write a recommendation for management.

Harmon Foods (B)

Evaluating the Impact of Promotions on Sales Volume

Managerial Problem

Harmons management spends heavily to promote Treat cereal. However, they do not fully understand the impact of promotional spending on sales. Management wants to evaluate the effectiveness of promotional spending. They are particularly interested in three questions:

- How do promotions impact sales?

- Is promotional spending profitable?

- On the margin, which type of promotion (trade or consumer) is most effective? That is, if management had extra funding for promotions, how should they spend it?

Modeling Challenge

1. MODEL. Present and discuss at least one regression model demonstrating the impact of promotions on sales. Your model should highlight:

- lagged promotion variables

- the net impact of a promotion on sales

2. EXHIBIT. Design at least one clear exhibit summarizing your model. Ideally, the exhibit will present coefficients (with measures of significance) and a measure of the net effect of each promotion. An exhibit requires an explanatory title and clear labeling.

3. MODEL FIT. Briefly assess the overall fit/validity of the model. This could be accomplished with one short paragraph, which might include a mix of the following: R-square, standard error of the estimate, significance of estimated coefficients, face validity of model.

4. PROFIT IMPACT. Compare the profit impact of the promotions. Any method of assessing profit impact is OK, as long as it would be understandable to management.

5. MANAGERIAL FINDINGS. Clearly and succinctly explain your managerial findings.

Data:

		Case	Consumer	Dealer	Index
YEAR	MONTH	Shipments	Packs	Allowance	Season	Trend
1983	Oct	na	0	224,028
1983	Nov	na	0	304,004
1983	Dec	na	0	352,872
1984	Jan	425,075	75,253	457,732	113	1
1984	Feb	315,305	15,036	254,396	98	2
1984	Mar	367,286	134,440	259,952	102	3
1984	Apr	429,432	119,740	267,368	107	4
1984	May	347,874	135,590	158,504	119	5
1984	Jun	435,529	189,636	430,012	104	6
1984	Jul	299,403	9,308	388,516	107	7
1984	Aug	296,505	41,099	225,616	81	8
1984	Sep	426,701	9,391	1,042,304	113	9
1984	Oct	329,722	942	974,092	97	10
1984	Nov	281,783	1,818	301,892	95	11
1984	Dec	166,391	672	76,148	65	12
1985	Jan	629,404	548,704	0	113	13
1985	Feb	263,467	52,819	315,196	98	14
1985	Mar	398,320	2,793	703,624	102	15
1985	Apr	376,569	27,749	198,464	107	16
1985	May	444,404	21,887	478,880	119	17
1985	Jun	386,986	1,110	457,172	104	18
1985	Jul	414,314	436	709,480	107	19
1985	Aug	253,493	1,407	45,380	81	20
1985	Sep	484,365	376,650	28,080	113	21
1985	Oct	305,989	122,906	111,520	97	22
1985	Nov	315,407	15,138	267,200	95	23
1985	Dec	182,784	5,532	354,304	65	24
1986	Jan	655,748	544,807	664,712	113	25
1986	Feb	270,483	43,704	536,824	98	26
1986	Mar	365,058	5,740	551,560	102	27
1986	Apr	313,135	9,614	150,080	107	28
1986	May	528,210	1,507	580,800	119	29
1986	Jun	379,856	13,620	435,080	104	30
1986	Jul	472,058	101,179	361,144	107	31
1986	Aug	254,516	80,309	97,844	81	32
1986	Sep	551,354	335,768	30,372	113	33
1986	Oct	335,826	91,710	150,324	97	34
1986	Nov	320,408	9,856	293,044	95	35
1986	Dec	276,901	107,172	162,788	65	36
1987	Jan	455,136	299,781	32,532	113	37
1987	Feb	247,570	21,218	23,468	98	38
1987	Mar	622,204	157	4,503,456	102	39
1987	Apr	429,331	12,961	500,904	107	40
1987	May	453,156	333,529	0	119	41
1987	Jun	320,103	178,105	0	104	42
1987	Jul	451,779	315,564	46,104	107	43
1987	Aug	249,482	80,206	92,252	81	44
1987	Sep	744,583	5,940	4,869,592	113	45
1987	Oct	421,186	36,819	376,556	97	46
1987	Nov	397,367	234,562	376,556	95	47
1987	Dec	269,096	71,881	552,536	65	48