If you ever played marbles, you know that there is a difference between Swirls and Agates. Let's say Swirls cost $.25 while Agates cost $5.00. If you pay $5.00 for a marble, you are going to want to make sure you bought an Agate and not a Swirl.

Pretend for a moment that each marble represents a unique prospect. Some are valuable ($5.00 Agates) and some are not as valuable ($.25 Swirls). Unfortunately, many marketing programs are run as though there is no difference between Agate prospects and Swirl prospects. It is as though prospect names come in at random, get shaken up in a jar and then programs are developed without regard to the value of each prospect. It would be the same thing as selling a jar of marbles without regard to the mix between Swirls and Agates.

Over twenty years ago I ran a relatively large and sophisticated business-to-consumer catalog company that mailed out over 100,000,000 pieces of direct mail per year. You can bet we knew the difference between our Agates and Swirls. We had three PhD statisticians pouring over results and running multivariate regression analysis against huge files based on what was then called RFM (recency, frequency and monetary) data resulting in precise measurements based on previous results, investment hurdles and availability of goods.

Business-to-business marketers do not have the same rich data sources as did (and do) b-to-c marketers, but they have more data than they think. What I tell my clients is that any data file can be segmented based on history, overlay of information and intuitive judgment. With a little effort it is possible to substantially increase results over what is commonly accepted today.

The sidebar table is a simple illustration of what can be an immensely complex topic—however, it does provide a framework for very simple segmentation that can produce very powerful results:

Note that a database of 1,000 prospects has be segmented (based on any known or attainable information) into five groups of 200. Note, too, that the response rate on the highest ranking group is nine times that of the response rate on the lowest ranking group. The rankings are generally based on SIC or NAICS, revenue and/ or number of employees, business type (Headquarters, Branch...) and can include geography and other "firmographics".

In the example shown, it would be possible to produce 96% of the results with just 80% of the investment or 84% of the results with just 60% of the investment...

Honestly, it does not take a rocket scientist (or PhD statistician) to enjoy very positive results from segmentation. You just have to segment, measure, re-segment... and the work is never "done".

Closed loop marketing fits in this equation because so few companies actually close the loop on their marketing investments.
I recommend two places to measure: First is lead rate (taking into account the relative amount of time and or dollar investment each lead type requires) and second is close rate. I have worked with one client for years and it is easy for us to generate leads in one particular vertical—yet, their sales force cannot close any of those leads. It's not the fault of the sales force. It's the nature of the vertical. So, despite the fact that our lead rate is relatively high in that vertical and as such the lead cost is relatively low, we have eliminated contact of that vertical from our prospect list.

If you are struggling with how to identify, segment and then close the loop on your database marketing, we can help! Let me know what challenges you are having in these areas and we can work together to come up with solutions.

Download the complete article How Relational Segmentation Techniques Help Achieve Higher Sales at Lower Cost.

By Dan McDade