home * about us * contact us * past features * columns * resource links * site map
 Home
 About Us
 Contact Us
 Site Map
 Past Features 
 Columns 
 Resource Links 


9/11 Remembered
Marketing With Numbers
Marketing literature is written mainly to create interest and arouse curiosity and often contains more hype than substance. Along these, the following is an excerpt from a press release from a data mining solution provider that was posted on the KDNuggets August newsletter. Note: The names of the modeling solution and the credit union in question have been left out at the discretion of the editors.
    Prior to using [the model, the credit union] typically received a one percent response rate from leasing direct mail campaigns with over 150,000 prospects. With the [model], a similar campaign targets only 15,000 prospects and achieves a 4.5 percent response yielding $510,000 in annual revenue from just one campaign. As a result, the [credit union] has reduced its per customer acquisition cost by 78 percent.

On the surface the figures sound very impressive. In reality, however, the numbers do not provide much insight into how much the customer actually benefited from using the product. In fact, based on the figures provided, one can come to the conclusion that the credit union would have been better off without the model.

According to the press release, prior to using the model, the credit union achieved a 1% response rate from a direct mail campaign of 150,000 prospects, i.e. 1,500 responses. In contrast, the model targeted only 15,000 prospects and achieved a 4.5% response rate. While 4.5% represents a 450% increase over 1%, 4.5% of 15,000 is in fact only 675 responses, 825 less than 1,500.

As for revenue, the press release states that using the model, the credit union brought in $510,000, which works out to roughly $755 per response. Applying this figure to original campaign involving all 150,000 prospects, the credit union would have brought in $1,132,500, more than $600,000 in additional revenue.

Of course, it is possible that by contacting only 15,000 prospects, the credit union so reduced its hard costs as to make the reduction in revenue worthwhile. Still this seems unlikely given the magnitude of the difference, not to mention the cost of acquiring the model, which could easily range in the $50,000 to $100,000 range. One is left to wonder if the analysts were so excited by the increase in response rate that they forgot what it was they were actually hired to do: increase profits. Or possibly it was simply the marketing department seizing upon figures that at first glance seem quite impressive.

All Contents Copyright © 2003 by Stone Analytics, Inc., except where otherwise noted.