The Customer Loyalty Solution : What Works (and What Doesnt) in Customer Loyalty Programs

Verizon Information Services was experiencing high attrition in its sales of Yellow Pages advertising. The annual attrition ranged from 10 percent among long-term advertisers to 27 percent among new advertisers. The attrition amounted to over $250 million per year, which translated to over $1.4 billion over the life of the lost advertisers. This loss plus a lower than average acquisition rate was reducing Verizon’s penetration in the marketplace.

A study by the Verizon Database Marketing staff showed that there appeared to be two reasons for the high attrition rate:

1. Many small- to medium-sized businesses felt that Verizon had not established a satisfactory relationship with them. The Verizon sales representatives would make contact with their clients only once or twice a year. As a result, some business owners felt that they lacked a close working relationship with Verizon.

2. Small businesses often had difficulty understanding and calculating the value of Yellow Pages advertising. Figuring out how to assess their return on investment was difficult for them.

To address these issues, Verizon wanted to create a retention communications program with these objectives:

• Position Verizon as a business partner for the small-business owner, not just another Yellow Pages provider looking to sell some ad space.

• Educate advertisers about the value of Yellow Pages.

• Target offers and messages to customers providing useful business information.

Verizon’s challenge, like that of many other companies, was to get its hands on good customer information in a timely manner. There was an abundance of data on the customers in Verizon’s data warehouse, but it was difficult to make that information available quickly enough to enable the company to carry out time-sensitive campaigns. Some of the issues facing Verizon included the following:

• Verizon was unable to coordinate the data with the sales cycle and customer profile. The sales cycle is difficult because Verizon produces more than 1500 different books during the year, with varying closing dates. Some advertisers are in only one book, but many are in two or more books that have different closing dates.

• It was difficult to determine the correct address for advertisers with multiple locations.

• There was no place in the existing database to store model scores or to segment categories.

• There was no process for communicating profile and research information to the supporting groups and departments.

The Verizon Database Marketing staff decided to launch a new program to solve these problems. When the staff began, several solutions were under development or already in place. Behavioral models had been built to predict advertising purchase behavior, including models showing

• Propensity to acquire

• Propensity to cancel, decrease, or increase current spending

• Propensity to buy additional products

• Lifetime value

Some preliminary work had been done on building a data mart, using periodic feeds from the data warehouse. This data mart would be used to organize, summarize, and analyze customer information. Finally, some direct-mail testing had been done, with varying levels of success. Since there was no formalized program in place, these elements were considered nothing more than research and development projects that might or might not yield some long-term value.

From Zero to Hero

The breakthrough solution came as a result of cooperation between information technology (IT) and marketing. A cross-functional team consisting of marketing and IT professionals came together in order to create a “Best in Class” database marketing program.

In order to launch the program, IT revised the data mart so that it could

• Aggregate customer transaction data to an annual level to match the sales cycle.

• Tag customer performance versus the previous year (increase, decrease, cancel, etc.).

• Overlay business demographics (SIC code, number of employees, annual sales, etc.).

• Attach model scores to each customer record.

• Record campaign results attached at the customer level.

When this data mart was combined with campaign management software, it became a powerful tool for direct marketing and marketing research.

While the data mart was being built, customer models were updated and made ready for use in segmentation and targeting. This gave the marketing team the ability to analyze and segment the customer base, design a database marketing program that would target the segments offering the highest opportunity, and create messaging to address those segments.

The final piece of the puzzle was to add a direct marketing agency to the team. The selection of an agency was important in order to ensure

• Strategic input

• Creative insight

• Tactical performance

• Flawless execution

The addition of a direct-mail fulfillment specialist helped to address these issues. Its ability to work with and provide insight to the creative agency, as well as efficiently executing the program’s primary direct-marketing campaigns, provided the foundation for success.

The End Result

The strategy was to use the database to send a series of direct-mail pieces that would stand out in the small-business owner’s mailbox. The idea was to make them colorful and vibrant, with bold, oversized type on all outer envelopes. Many of the pieces were designed to help the customers calculate their return on investment from a Yellow Pages ad.

Contact frequency and messaging were based on segments defined by the advertisers’ tenure and their propensity scores on the various models. A minimum spending level was set as a qualification, so that the marketers could focus their efforts on the top 50 percent of the customer base. All contacts were timed to match various stages of the sales cycle:

• Tenure. Advertisers were divided into multiple segments on the basis of seniority. Newer segments received more value-reinforcement and relationship-building messages. Segments that had been customers longer received both relationship and up-selling communications.

• Propensity scores. Model scores were then added to the segments in order to further target messaging. Advertisers who were at high risk for attrition received stronger reinforcement communication focusing on educating them on the value of a Yellow Pages ad and how to measure the success of such an ad. Advertisers who had a high possibility of increasing spending in certain areas were given strong sales messages in those specific areas. Those customers with high scores in both lifetime value and risk of defection received messages showing them how to optimize their product mix and encouraging trial. Those who scored low in all areas received more relationship-building messages.

All mailings were sent first class so that any pieces with bad addresses would be returned. The returned mail was keyed into the database. The impact of the direct-mail communications program was tremendous. The goal was to get a 1 percent increase in sales revenue, or $6 million in additional Yellow Pages ad sales. Control groups that did not get the mailings were set up to measure the results. These showed that the program generated a 3.8 percent increase in overall annual sales revenue, with an additional $13,361,000 in incremental Yellow Pages advertising. Comparing these results with the cost of the mailings and the database, Verizon realized a 1681 percent return on investment.

The success of this program dramatically changed the way Verizon communicated with its advertisers. In subsequent sales cycles, a newsletter was added to increase communication with and education of the customer. Email retention messages were incorporated to make the program more robust. Budgets for these programs were increased in the following year.

The marketing data mart became the cornerstone not just for database marketing, but for all loyalty and database marketing programs companywide. It became the preferred source for information on Verizon customers.