Everyone is talking about customer value. And no wonder. If you have a customer database, the only way to value it is with customer value.
Consumers are growing adept at tuning out the more than 3,000 marketing messages that bombard them daily. We’re all becoming equally numb with the constant barrage of e-mails. My daily junk e-mails easily total 100. I’m sure that some well-intended messages are getting trashed in the onslaught.
Customers also are growing more sophisticated, demanding and selective regarding marketing pitches. They are determining when and where they wish to interact with companies. While some consumers want a relationship with a company they buy from, the vast majority do not.
As a result, companies are searching for a competitive edge, knowing that the same old marketing and selling techniques no longer work. Because the consumer has a choice of channels – e-mail, Web and telephone – marketing, sales and service no longer can be selling opportunities.
The businesses that will be most successful in the future are those that invest in understanding their customers. But understanding the customer is not as easy as it seems.
First, you need to know the difference between a customer and a prospect or attriter in your database. For many companies, just determining who is a customer creates all types of meetings. Is a customer an individual or a household? Have they purchased once or several times in the past few months? If they returned merchandise for credit are they as important as the top quintile in recency, frequency and monetary value?
An even more fundamental issue is the type and amount of data kept on a customer. Many records don’t even have adequate mailing addresses. What can an e-mail address tell you about a household? And what about all those null values? When no data are associated with a customer or prospect, it becomes difficult to create value.
Up to 20 percent to 50 percent of many customer databases are estimated to lack enough data to create customer value.
The world moves at such lightning speed that the time to create a positive impression on a prospect or a bad impression with a customer has diminished from months just a few years ago to a few minutes today. Companies need to continue refining their abilities to deliver relevant, timely and value-oriented communications and to know with confidence what audience they are addressing.
At the very least, a company needs to understand the strengths and weaknesses of its customer data. Once enough records have complete data fields, data mining and segmentation can begin without the risks of guessing or creating inferred data substitutes. Known information that is accurately collected always outperforms inferred or derived data. Inferred or derived data are good for marketing to large groups but lousy for one-to-one marketing.
Modeling and scoring help companies apply value to customers and prospects, segment values and perform response modeling and cross-selling. Before a program of scoring a customer record can take place, an understanding is needed of how much the customer has purchased and over what time.
Though it’s not applicable for all customer segmentation, RFM analysis should be considered. This can be the test to see what purchase history exists on the customers.
For firms where purchase data cannot be tracked – such as hard-goods companies, business-to-business companies selling heavy equipment or consumer-packaged goods – RFM does not work well. But for companies where purchase data can be tracked, RFM analysis provides a clear, historical perspective of the value of a customer.
But watch out. RFM does not help in understanding share of wallet among groups. Nor does it let a company model prospective good customers who haven’t yet emerged.
For this reason, profiling is often supplemented to better understand customers and prospects. Profiling counts on a variety of data elements. Since profiling doesn’t rely on purchase information and discount demographics, you can use it to help define your best customers.
Generally, buying third-party data will be essential to fill in the blanks in the data. In your initial test you might choose to have two or three firms supply information on the household, financial strength of the household or lifestyle behavior.
The most important characteristics will emerge, and they will help you with future modeling, scoring and ranking.
Marketing automation systems now can evaluate the characteristics of the “gold,” “silver,” “bronze” and “lead” customers and prospects. Assuming the profiling work has been done correctly, weighted values can be applied to the most predictive elements. LTV (Life Time Value) scores now can take a strong role in the valuing of the customer database.
Once this is achieved, automated marketing programs can be set with trigger-event business rules. Offers and communications can be delivered in a timely, relevant manner, with value for that segment of the database. As more experiences are achieved and more models are refined, ever more targeted marketing campaigns can be set.
