In an article on Oct. 9, 2000, in the Financial Times’ Mastering Management series, Wharton marketing professor
Banks have traditionally treated retail customers identically: “Deposit your money and you get a check book.” Yet when banks began to collect data that allowed them to compute profitability for each customer, there were marked differences. Some kept a lot of money in their checking accounts and rarely accessed it, or if they did, withdrew the cash through ATMs. These customers were very profitable. Others kept low balances and made frequent withdrawals, often through contact with tellers. Such customers ended up costing the bank more, and were not profitable. Identifying profitability helped banks devise strategies to suit different types of customers.
Customer Portfolio Management
The analysis banks used to track profitability was customer portfolio management. It is a technique that has proved useful for many businesses, those dealing with both consumer and business markets. To determine which customers were most valuable, banks had to allocate revenues and costs at the customer level. Although revenues had traditionally been allocated at the customer level (banks always knew who had the most money in their accounts), it was not typical to treat costs in this way. In banks, as in most companies, costs were handled at the product or operations level.
By allocating costs to each customer, banks were able to find out which customers were most profitable. It then became easier to devote resources and time to the most profitable customers. While revenues were known to follow an 80/20 law, so that 80% of revenues were generated by 20% of customers, costs can follow a 90/10 law, so 90% of costs are generated by 10% of “whiny” customers. If you reduce the costs of serving customers whose business does not warrant additional attention, and spend more time serving customers whose business does, the overall customer portfolio can become more profitable.
The trick is to figure out how to allocate costs to each customer. Costs can vary by units (such as inventory, physical handling, cost of goods sold). They can vary at the transaction level (order processing, shipping, order frequency costs); or at the relationship level (maintaining an account, contact time, service costs). There are also selling and maintenance costs (promotional mailings, free samples) and enterprise level costs (warehousing). These costs need to be assigned as appropriate to each customer. When costs are broken down and compared with revenues from each customer, profitability can be computed.
This profitability determines the asset value of the customer relationship, or the customer equity. One way to think about customer equity is to create a grid with revenues (high/low) on the vertical axis and customer costs (low/medium/high) on the horizontal. Customers can be designated to one of the boxes on the grid and profitability assessed.
When customers are designated in a portfolio system, management goals become clear. With unprofitable customers, managers need to do three things. First, reduce the costs of dealing with them, for example by encouraging electronic banking. Second, increase the revenues generated, perhaps through fees for services or price increases. Finally, consider ending the relationship, although this is frequently an undesirable or unethical solution.
The problem with profitable customers is retaining them, because they will attract the attention of your competitors. Building customer relationships may be the answer to both types of problem.
Relationship Marketing
Relationship marketing is grounded in the idea of establishing a learning relationship with customers. At the lower end, building a relationship can create cross-selling opportunities that may make the overall relationship profitable. For example, some retail banks have tried selling credit cards to less profitable customers. With valuable customers, customer relationship management may make them more loyal and willing to invest additional funds. In banking, these high-end relationships are often managed through private bankers, whose goals are not only to increase customer satisfaction and retention, but also to cross-sell and bring in investment.
In determining which customers are worth the cost of long-term relationships, it is useful to consider their lifetime value. This depends on:
- current profitability computed at the customer level
- the propensity of those customers to stay loyal
- expected revenues and costs of servicing such customers over the lifetime of the relationship
Building relationships makes most sense for customers whose lifetime value to the company is the highest. Thus, building relationships should focus on customers who are currently the most profitable, likely to be the most profitable in the future, or likely to remain with the company for the foreseeable future and have acceptable levels of profitability.
Relationship Goals
The goal of relationship management is to increase customer satisfaction and to minimize any problems. By engaging in “smarter” relationships, a company can learn customers’ preferences and develop trust. Every contact point with the customer can be seen as an chance to record information and learn preferences. Complaints and errors must be recorded, not just fixed and forgotten. Contact with customers in every medium, whether over the internet, through a call center, or through personal contact, is recorded and centralized.
Many companies are beginning to achieve this goal by using customer relationship management (CRM) software. Data, once collected and centralized, can be used to customize service. In addition, the database can be analyzed to detect patterns that can suggest better ways to serve customers in general. A key aspect of this dialogue is to learn and record preferences. There are two ways to determine customers’ preferences: transparently and collaboratively.
Discovering preferences transparently means that the marketer learns the customers’ needs without actually involving them. For example, the Ritz Carlton hotel makes a point of observing the choices that guests make and recording them. If a guest requests extra pillows, then extra pillows will be provided every time that person visits. At upmarket retailers, personal shoppers will record customers’ preferences in sizes, styles, brands, colors and price ranges and notify them when new merchandise appears or help them choose accessories.
With increased use of the internet, companies have more opportunities to learn customers’ preferences and behavior transparently. Information that can be easily collected includes:
- words used in search engines to find your site. These can help understand how to categorize your services
- source of visit (search engine, other sites, directory sites, portals, banner advertising and so on)
- individual customer measures such as number of monthly visits; time spent per visit and search patterns; orders from site; conversion (from visitor to purchaser); spending per order; spending per visit.
When marketers learn customers’ preferences collaboratively, they engage in dialogue to help customers articulate their needs and identify how to meet those needs. Ultimately, this method should result in an ideal product, but it can take time. The goal of the marketer in learning preferences collaboratively is to determine a way to maximize learning without frustrating customers. Again, the web can be useful with online questionnaires and forms.
Permission marketing can be used also. This involves asking customers for permission to gather personalized information or to contact them in the future. Finally, collaborative filtering is effective for learning customers’ preferences and helping them choose items they may enjoy. The book retailer Amazon.com uses this approach. The company tracks customers’ preferences and purchase patterns, matches them with those of similar customers, and recommends other products based on purchases by similar customers.
When customers collaborate with marketers, the customers are also learning their own preferences. Customers can thus have more control and ensure they get what they want. A product designed with the customer is by definition not wrong, so having a customer collaborate on discovering his or her own needs encourages the customer to commit. Finally, just being part of the process seems to increase satisfaction. Levi-Strauss found this to be the case with customized jeans. Customers were more likely to be satisfied with the final product if they actually tried on a pair of jeans (although this wasn’t necessary in the process).
Learning Helps Cross-selling
One advantage of learning a customer’s preferences is that the company then has a record that can be used to cross-sell other products or services. For example, if a customer used personal finance software, such as Quicken, then a marketer could (as a service) offer to provide credit card information in that format. This information would be useful to the customer for financial planning and would provide a rich information bank for the marketer.
As well as helping the customer to control his or her finances, the marketer (with permission, of course) could learn about the customer’s recent purchases. For example, if a customer began buying children’s clothing, the marketer could propose related goods; or if plane tickets were bought, hotels could be suggested.
Other customer histories can help the marketer. For example, health organizations could keep a record of incidents and patient lifestyles that might suggest diagnostic testing and treatments. Insurance companies, such as USAA, which has traditionally focused on military officers and their families, already do this. USAA follows marriages, births and other life-changing patterns so it can advise customers on changing needs. By responding to these patterns, USAA increases revenues by selling more insurance and financial services. Furthermore, it increases customer satisfaction and loyalty by monitoring and adapting to their needs.
Marketers need to be careful with this information and build a relationship based on trust. Some businesses have natural advantages: People are more likely to trust a doctor or a bank than a supermarket. However, if relationships can be built and if the marketer provides valuable suggestions, the customer is likely to be loyal. It is easier for a customer to stay with a trusted company than to switch to another.
A Virtuous Circle
By learning customer preferences and focusing on long-term relationships, managers can provide products and services that fit customers’ needs. They can also do this in a way that ensures loyalty. If a company earns a customer’s trust and if, as a result of that trust, customers share strategic information about their preferences and needs, it will be difficult for competitors to duplicate the relationship.
As relationships develop, customers will tend to buy more from the company. Further, the more a customer buys, the more likely he or she will be to buy from that company again. This virtuous circle is reinforced because the more a customer buys from the trusted company, the less likely he or she is to turn to another supplier. Finally, the regular customer is more likely to switch to a premium product or service.
The ultimate reward in managing customized relationships will come if a company can transform customers into advocates. A recent example has been seen among Palm Pilot owners. The Palm Pilot, a pocket organizer, can be customized and so has fanatical supporters. Owners are willing to submit their names and preferences, to be on mailing lists, and are frequent visitors to the company’s web pages. In addition, these loyal customers extol the virtues of the product to potential buyers.
Although price strategies may be effective in the short-term, they rarely come out best in the long run. A better strategy to transform customers into advocates is to try to meet the needs of each customer more precisely. Learning customers’ preferences can not only help meet their needs better than the competition, but can also help marketers forge an enduring relationship.
