Analyzing the Data
Traditionally, customer databases have been analyzed with the intent to define customer segments. A variety of muUivariate statistical methods such as cluster and discriminant analysis have been used to group together customers with similar behavioral paiterns and descriptive data which are then used to develop different product offerings or direct marketing campaigns.** Direa marketers have used such techniques for many years. Their goals are to target the most profitable prospects for catalogue mailings and to tailor the catalogues to different groups.
More recently, such segmentation approaches have been heavily criticized.” Taking a large number of customers and forming groups or segments presumes a marketing effort towards an “average” customer in the group. Given the range of marketing tools available that can reach customers one at a time using tailored messages designed for small groups of customers (what has been referred to as ” 1 -to-1″ marketing), there is less need to consider the usual market segmentation schemes that contain large groups of customers (e.g., women 18-24 years of age). Rather, there is increased attention being paid to understanding each “row” of the database—that is. understanding each customer and what he or she can deliver to the company in terms of profits and then, depending on the nature of the product or service, addressing either customers individually or in small clusters.
As a result, a new term, “lifetime customer value” (LCV), has been introduced into the lexicon of marketers. The idea is that each row/customer of the database should be analyzed in terms of current and future profitability to the firm. When a profit figure can be assigned to each customer, the marketing manager can then decide which customers to target. The past profit that a customer has produced for the firm is the sum of the margins of all the products purchased over time less the cost of reaching that customer.
These costs include any that can be broken out at the individual customer level, through such efforts as direct mail and sales calls. Note that mass advertising would not be part of this formula. The cost could be assigned to individual customers by computing a per customer dollar amount; but because it is the same for each customer, it would not affect the rank ordering of the customers in terms of their profitability. LCV is calculated by adding forecasts for the major parameters and discounting back. This obviously requires assumptions about future purchasing, product and marketing costs, as well as how long the customer can be expected to remain with the firm. Generally, this will result in a number of scenarios for each customer depending upon these assumptions.
The LCV formula can also be used to show where additional profits can be obtained from customers. Increased profits can result from:
• increasing the number of products purchased, by cross-selling;
• increasing the price paid, by up-selling or charging higher prices;
• reducing product marginal costs; or
• reducing customer acquisition costs.
Other kinds of data analyses besides LCV are appropriate for CRM purposes. Marketers are interested in what products are ofien purchased together, often referred to as market basket analysis. Complementary products can ihen be displayed on the same physical page in a hard-copy catalogue or virtual page
on a Web site.
As noted, a new kind of analysis born from the lnlernel is the clickstream analysis. In this kind of data analysis, patterns of mouse “clicks” are examined from cyberstore visits and purchases in order to better understand and predict customer behavior.'” The goal is to increase “conversion” rates, the percentage of browsing customers to actual buyers. For example, companies such as Blue Martini and Net Perceptions sell software that enables Web-based stores to customize their sites in real time depending upon the type of customer visiting—that is, their previous buying patterns, other sites visited during the current session, and their search pattern in the cyberstore.
Customer Selection
Given the construction and analysis of the customer information contained in the database, the next step is to consider which customers to target with the firm’s marketing programs. The results from the analysis can be of various types. If segmentation-iype analyses are performed on purchasing or related behavior, the customers in the most desired segments (e.g., highest purchasing rates, greatest brand loyally) would normally be selected first for retention programs. Other segments can also be chosen depending upon additional factors. For example, for promotions or other purchase-inducing tactical decisions, if the customers in the heaviest purchasing segment already huy at a raic that implies further purchasing is unlikely, a second tier with more potential would also be attractive. The descriptor variables for these segments (e.g., age, industry type) provide information for deploying the marketing tools. In addition, these variables can be matched with commercially available databases of names to lind additional customers matching the profiles of those chosen from the database.
If individual customer-based profitability is also available through LCV or similar analysis, it would seem to be a simple task to determine on which customers to focus. The marketing manager can use a number of criteria such as simply choosing those customers that are profitable (or projected to be) or imposing an ROI hurdle. The goal is to use the customer profiiability analysis to separate customers that will provide the most long-term profits from those that are currently hurting profits. This allows the manager to “fire’ customers that are loo costly to serve relative to the revenues being produced. While this may seem contrary to being customer-oriented, the basis of the time-honored “marketing concept,” in fact, there is nothing that says that marketing and profits are contradiaions in terms. The 80/20 rule often holds in approximation: most of a company’s profits are derived from a small percentage of their customers.
For example:
• AT&T offers different levels of customer service depending upon a customer’s profitability in their long-distance telephone business. For highly profitable customers, they offer “hot towel,” personalized service. For less profitable customers, you get automated, menu-driven service.
• The wireless provider PageNet raised monthly rates for unprofitable subscribers. Clearly, the intent was to drive them away.
• Similarly, Federal Express raised shipping rates for residential customers in expensive-to-serve areas where their volume did not justify normal rates.
The point is thai without understanding customer profitability, these kinds of decisions cannot be made.
On what basis should these customer selection decisions be made? One approach would be to take the current profitability based on the above equation.
An obvious problem is that by not accounting for a customer’s possible growth in purchasing, you could be eliminating a potentially important customer. Customers with high LCV could be chosen, as this does a better job incorporating potential purchases. However, these customers are difficult to predict and you might include a large number of unprofitable customers in the selected group.
No matter what criterion is employed, de-selected customers need to be chosen with care. Once driven away or ignored, unhappy customers can spread negative word-of-mouth quickly, particularly in today’s Intemet age.
Targeting the Customers
Mass marketing approaches such as television, radio, or print advertising are useful for generating awareness and achieving other comimunications objectives, but they are poorly-suited for CRM due to their impersonal nature. More conventional approaches for targeting seleaed customers include a portfolio of direa marketing methods such as telemarketing, direct mail, and, when the nature of the produa is suitable, direct sales. Writers such as Peppers and Rogers’^ have urged companies to begin to dialogue with their customers through these targeted approaches rather than talking “at” customers with mass media.
In panicular, the new mantra, “l-to-l” marketing, has come to mean using the Intemet to facilitate individual relationship building with customers.”
An extremely popular form of Internet-based direct marketing is the use of personalized e-mails. When this form of direa marketing first appeared, customers considered it no different than “junk” mail that they receive at home and treated il as such with quick hits on the delete button on the keyboard. However,
sparked by Godin’s call for ”permission”-based programs whereby customers must first “opl-in” or agree to receive messages from a company, direct e-mail has become a very popular and effective method for targeting customers for CRM purposes.'” Companies such as Kana and Digital Impact can send very sophisticated e-mails including video, audio, and web pages. Targeted e-mails have become so popular that Jupiter Media Metrix projects that over 50 billion of them will be sent in 2001.
A study by Forrester Research shows why this is so.” Exhibit 4 demonstrates that e-mail is a very cost-effertive approach to customer retention.
Through lower cost per 1,000 names by using the company’s own database (the “house” list) and greater clickthrough rates than those afforded by banner advertisements and e-mails sent to lists rented from suppliers, companies can reduce their cost per sale dramatically.
Some examples are the following:
• Southwest Airlines’s e-mail-based Click ‘n Save program has 2.7 million subscribers. Every Tuesday, the airline sends out e-mails to this database of loyal users containing special fare offers.
• The bookseller Borders (Borders.com, Borders and Waldenbooks offline retailers) collected all of its customer information into a single database.
The company then uses e-mails tailored to the customer’s reading interests to alert them about upcoming releases.
• The Phoenix Suns basketball team sends streaming video messages from its players promoting new ticket packages and pointing them to the team’s Web site.
by Russell S. Winer