Product Inter-Relationships – Cannibalisation
OVERVIEW
This paper describes the methodology used by Promax to determine the reaction of baseline sales of one product caused by the promotion of other products and how the technology can aid the planner in developing a clearer understanding of the overall promotional plan. We deal with the concepts of historical analysis of promotion activity and the segmentation of the consumer response when there is no promotion (normal sales or baseline) and the lift in sales from the promotion.
Symptoms
Predicting what the effect will be of a promotion is a complex and involved task as there are many causal effects that need to be taken into consideration. An account manager who analyses the historical retail scan data will commonly observe that the promotion of one product can cause an effect on the baseline sales of another. A typical example is changing from the standard pack to a promoted pack when both packs remain on sale. The introduction of the promoted pack, which may contain 10% extra product, will inevitably cannibalise the base rate of sales of the standard pack. The extent to which this cannibalisation occurs is difficult to measure without sophisticated statistical analysis integrated into the promotion planning environment. Yet it is vital information for the account manager and should be easily interpreted at the time of planning the promotion.
Let’s look at the problem. We’ll use the familiar sample of baked beans to illustrate this effect. In our example, the standard pack of baked beans is a 375g can and its normal rate of sale through our sample retailer is 400 cases per week. Periodically we promote a 450g can for the same price as our 375g SKU. When we do this the sales of the 450g SKU is increased by 200% above the normal level and the 375g SKU’s sales dropped to 50% of the normal sales. If the normal retail price of the 375g unit uses $0.89 and the normal retail price for the 450g is $0.99, were we better off by promoting our 450g unit at $0.89 or should we choose another strategy?




Leave a Comment (* required fields)