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

While many companies have failed at managing customer segments on a day-to-day basis, National Australia Bank (NAB) succeeded at this task. The lead franchise in NAB was National, a Melbourne-based bank serving more than 4.5 million customers. With the help of Beth Ann Konves of the Teradata Division of NCR, Fernando Ricardo, head of customer knowledge management at the bank, identified six major customer segments composed of individuals and businesses. Each segment was managed as a global business. The system, called National Leads, identified leads within the bank’s database based on significant changes (events) in a customer’s transaction pattern, such as balance changes, large transactions, or loan renewals. The system also calculated each customer’s propensity to buy a product or to respond to an offer. The bank’s database was maintained on an NCR Teradata computer using Teradata’s CRM analytical software, which was used for the National Leads system.

The National Leads system prioritized events and alerted the appropriate bankers each morning so that they could take appropriate action. The system also served as a communications gatekeeper, managing the frequency, content, and channel used for customer interactions during marketing campaigns. The goal was to ensure that customers were contacted with meaningful opportunities at the right times through the right channels.

The six major customer segments used by the bank were common to many other banks and financial institutions:

• Custom business

• Packaged business

• Agribusiness

• Private

• Premium

• Retail

Each segment was managed by a vice president. Every National customer was assigned to one of the segments on the basis of a financial needs analysis and profitability assessment. Within each of these major segments, further refinements were made based on current profitability, the bank’s share of the customer’s wallet potential, and the customer’s demonstrated financial needs. As a result of this analysis, there were more than 200 subsegments that were used for target marketing. The system worked this way:

• The Teradata warehouse was updated with an average of 2.2 million transactions per day.

• Every night, queries “trawled” the warehouse to search for any unusual changes. The queries involved complex logic, scanning hundreds of millions of rows in the database and joining from six to ten tables at a time.

• Once a month, each customer was scored, using a model that predicted the customer’s propensity to purchase various products and propensity to respond to product offers. The best leads were selected and sent to the bankers.

• Every night 250 communication vectors combining the events detected and the propensity predictions were run. The software recommended the action to be taken via ATM, email, mail, or leads to call centers, branches, or package business bankers, who followed up with a phone call or personal mail.

• The system captured all feedback and responses, measuring them against the opportunities developed. Each offer was placed in a location in a three-dimensional “cube,” with the three axes being the segment, the channel, and the offer. The system moved customers from one location in the cube to another based on their response to offers.

In a typical scenario, the system might direct a banker to contact a customer with an offer, such as a financial planning appointment or the renewal of a loan. Armed with this lead, the banker would contact the customer to gather financial profile information, using the standard financial needs assessment process developed by the bank. Bankers were given incentives to gather pieces of profile information every time they spoke to their customers. The bankers typed the collected data into a customer needs record in the database, using a terminal on their desk.

The way it worked was this: A customer might tell a banker that he was currently renting, but that he planned to purchase a home in about 9 to 12 months. The banker would enter this information into the system. Eight months later, the banker would receive a message generated by the database to recontact this customer to discuss home loan options.

The Results

• During the first year, more than 1 million leads and $4 billion in growth opportunities were sent to National bankers.

• During the next 6 months, 570,000 new leads were sent, which resulted in the closing of $4.4 billion worth of new loans.

• During the second year of the system, premium sales of banking products increased by 25 percent over the previous year, while sales of wealth management products increased by 40 percent.

• The close rate for leads increased by five times over the close rate before the new system began.

• The bank achieved a $391 million return on investment on one campaign.