Measuring Campaign Effectiveness
2. March 2009 01:54How do you measure the effectiveness of your marketing action? Do you take into account dynamic effects and interactions of multiple marketing-mix variables?
Dynamic effects:Response to marketing actions does not often take place instantly. The effect of an ad campaign does not end when campaign is over; the effect, or part of it, will continue in a diminished way for some time. Many customers purchase more then they can consume of a product during a short term price promotion.
Carryover effects - is the general term used to describe the influence of a current marketing expenditure on sales in future periods. We can distinguish several types of carryover effects.
Delayed-response effect , arises from delays between when marketing dollars are spent and their impact. Delayed response is especially evident in industrial markets, where the delay, especially for capital equipment, can be a year or more.
Customer-holdover effect, arises when new customers created by the marketing expenditures remain customers for many subsequent periods. Some percentage of such new customers will be retained in each subsequent period; this phenomenon gives rise to the notion of the customer retention rate and in its converse, the customer decay rate (also called the attrition or erosion rate)
New trier effects, in which sales reach a peak before settling down to steady state, are common foe frequently purchased products, for which many customers try a new brand but only a few become regular users
Stocking effects occur when a sales promotion not only attracts new customers but encourages existing customers to stock up or buy ahead. The stocking effort often leads to sales trough in the period following the promotion
The most common dynamic or carryover effects model used in marketing is:
Yt=a0+atXt+λYt-1
Says that at time t (Yt) are made up of a constant minimum base (a0 ) an effect of current activity atXt and a proportion of last period sales (λ) that carries over to this period.
Managers can easily guess λ directly as the proportion of sales that carries over from one period to the next or estimate it by using linear regression.
Multiple marketing-mix elements: interactions
When we consider multiple marketing-mix variables , we should account for their interactions.
Interactions are usually treated in one of three ways:
- by assuming they do not exist
- by assuming that they are multiplicative
-
by assuming that they are multiplicative and additive
assumption (1) gives us :
Y=af(X1)+bg(X2)
assumption (2) gives us :
Y=af(X1)*g(X2)
assumption (3) gives us:
Y=af(X1)+bg(X2)+c*f(X1)*g(X2)
In practice when multiple marketing-mix elements are involved, we can resort to one of two forms:
(1) The full linear interactive model
Y=a+b*(X1)+c*X2+ d*X1*X2
(2) The multiplicative form is :
Y=a*X1b*X2c
b,c are the constant elasticities of the first and second marketing- mix variables, respectively, at all effort levels X1 and X2.
Email 
