Goal: Increase The Success Rate Of Telemarketing

The marketing department of a medium sized software company was charged with developing a strategy for the introduction of a new product. A strategic decision was made to design and implement a telemarketing campaign to sell the new item to previous customers.  The customer database consisted of 30,000 names, as well as several fields for each customer; such as, address, date of previous purchases, modules purchased, etc.

First the marketing group randomly selected 400 customers.  Of that sample, 50 cases had incorrect phone numbers, leaving a viable sample of 350.  All were called and pitched the new product.  Sales were recorded in 9 calls.  The success  rate with this method was 2.5%.

WizWhy was then run on the same sample.  The key prediction parameter

(dependent variable) in the WizWhy analysis was whether or not the customer had purchased the new product. WizWhy automatically discovered all patterns and characteristics of the customers who were purchasers, as well as those who did not purchase.

The full customer database (less the sample) was then subjected to the WizWhy prediction mode.  WizWhymade a prediction of the conclusive probability of  purchase for each customer in the population, based upon the profiles it had revealed in the sample.  The database was sorted according to this conclusive probability.  A telemarketing campaign was then targeted at the 1000 customers with the highest conclusive probability, i.e., those most likely to purchase.  Of the 1000 customers called, 60 in fact purchased the new product.

Using WizWhy to determine the profile of likely buyers increased the success rate from 2.5% to 6%!