SAVE | EMAIL | PRINT | MOST POPULAR |  RSS | | REPRINT
|
The Science of 1:1 Marketing
April 02, 2008
By Dr. Mark Klein
The mantra of 1:1 Marketing has always been to make the right offer to the right customer at the right time … But that's much easier said than done. If your company has a million customers, are you supposed to have a million different marketing plans, one for each customer? Or, is it good enough to segment the customers into a more manageable number of groups with a different plan for each group?
The granularity of your plan is just one of the problems. Accuracy is another big issue. How do you figure out what to do and what to offer, regardless of whether your target audience is a group of one or one thousand? A third problem area is the availability of data on which to base your plan. On one hand, there is the Federal Trade Commission raising the possibility of "Do Not Track" lists to curb privacy invasion. On the other lies the cost and difficulty of trying to append demographic, psychographic or firmographics data to your customer file. Yet, without this customer data, there is no hope of making intelligent offers.
As formidable as these obstacles may seem, several forward-looking companies are overcoming them by employing a collection of techniques, coming to be known as Mathematical Marketing (MM). Mathematical Marketing is the process of marketing to existing customers based on a scientific understanding of how past customer behavior predicts future purchases.
Comparison with Search Engine Marketing
One way to understand MM is to compare its uses and elements with the more familiar concepts of Search Engine Marketing (SEM), such as customer acquisition.
The prototypical SEM application is AdWords from Google. SEM elements include keywords, ad impressions, click-through, search engine optimization, and web analytics.
Elements of Mathematical Marketing
There is a similar set of elements for mathematical marketing:
Behavioral tracking. Transaction records are the input data. While external data might be used to supplement the purchase transaction data, it is not necessary. A big advantage of basing MM on transaction records is that every company doing direct marketing already has this data and there is no conflict involved in their using it. Further, past purchase behavior is by far the best predictor of future buying behavior.
• Predictive analytics. Mathematical algorithms calculate purchase propensities. They predict which customers are going to buy next, what they are likely to buy, and which customers are likely to defect. Of course some algorithms and methods are better than others. RFM delivers minimal accuracy. Done properly, the analysis of a company with tens of thousands of customers and thousands of products should only take about one hour.
• Behavioral targeting. Algorithms show marketers how to make the right offer to the right customer at the right time. The challenge is to both calculate the probability for each customer to buy each product and to make this information available to marketers in an easy-to-use format.
• What-if analysis. There is no reason why a what-if analysis should be limited to financial types and spreadsheet jockeys. Predictive tools enable manipulation of target lists to achieve the best outcome. Marketers can now accurately estimate the results of a campaign before creating the collateral or paying for postage.
• Testing is one of the two best ways to improve the results of marketing campaigns yet most companies bypass this step because they think it is too expensive or time consuming. However new techniques such as factorial design enable ten, twenty, even forty variables to be tested simultaneously, and an e-mail campaign can be executed in 36 hours. Now, expert systems for automated testing are available to determine the best target groups and offers.
• Results measurement. Despite all the science, there is still a considerable component of art in marketing. It is important to measure campaign results in terms of revenue and yield so you can adjust methods to achieve maximum sales. Determining ROI is another benefit, in line with Peter Drucker's maxim: "If you can't measure it, you can't manage it."
What's Next
The individual elements of MM have been around for a few years, but until recently their use has been limited to sophisticated practitioners at larger enterprises. That situation is changing due to three developments:
1. There is a growing realization that transaction data by itself can be an effective predictor of future buying behavior.
2. New, lower cost services are appearing that incorporate all of the elements of MM.
3. The newer services are more user-friendly and don't require knowledge of higher mathematics, making them more accessible to more people, especially at SMB-sized companies.
Coupling these developments with the fact that most companies have many more existing customers than newly acquired ones, and it becomes easy to predict the rapid growth of technology services in the existing customer space to parallel Google and SEM in the customer acquisition arena.
Dr. Mark Klein is the CEO and founder of Loyalty Builders LLC, a leader in the new science of mathematical marketing, plus three earlier companies. Prior to his career in business, he earned a Ph.D. in Theoretical Physics from Indiana University, and taught and did research at several other colleges. His first novel (http://www.whencomessuchanother.net) will appear soon. Contact him at markk@loyaltybuilders.com.
Sales & Marketing Management Magazine
This article is brought to you by Sales & Marketing Management, the leading authority for executives in the sales and marketing field.
|
|
SAVE | EMAIL | PRINT | MOST POPULAR | RSS |
|
|
| Back to Marketing Index |
|
|
