Increasing Customer Value: Harness the Power of Predictive CRM
Use analytical CRM to find out things you’ve never known about your customers – and to say just the right thing at just the right time.
A successful Customer Relationship Management (CRM) strategy relies on three critical components: first, a company needs a CRM vision; second, a strategy to center its operations around its customers; and third, the technology to support that strategy.
To date, most CRM strategies focus on improving customer connections by developing operational efficiency. And, the technology to support that strategy included operational CRM systems such as sales force automation and marketing automation. While operational systems may lead to faster and more efficient customer interactions, they do not enhance the quality of interactions. Operational systems do not help a business say the right thing, or sell the right product to the right customer in the right place at the right time; in other words, they do not help optimize customer interactions.
The only way companies can optimize every interaction with every customer is by using the next generation of CRM: the predictive technologies found in a complete analytical CRM solution. With analytical CRM, companies identify their most valuable customers and predict what customers will want next. Companies gain insights into the factors that drive customer value, such as customer share and loyalty, and use this information to improve customer interactions and optimize revenue, thus providing a new basis upon which they can measure their success.
Using Analytical CRM to Increase Customer Value
Customer Relationship Management (CRM) is not a new topic in today's business marketplace. For years, companies have implemented systems that help them interact with their customers. From enhancing call center operations to improving sales transactions, companies are keenly aware of the need to improve customer connections in order to increase their bottom line.
The payback is clear. Increasing loyalty and keeping current customers costs five to seven times less than finding new ones. Over a lifetime, loyal customers purchase more, cost less to sell to, and studies show they will refer five other people to a business.
But how do you create loyal customers? Build a positive and favorable customer experience. Of course, making sure that on-hold times are short, that customer service representatives have access to a record of previous problems and that the Web site is easy to navigate are all critical to improving customer loyalty.
However, these operational systems are not being used to their full potential until a company can predict the right thing to say, or sell, to the right customer at the right time - in other words, until they begin to optimize individual customer interactions.
The best way to optimize customer transactions - and the only way to become a truly customer-centered organization - is through the knowledge and insight provided by analytical CRM. Analytical CRM involves using data to understand customers, measure and monitor key customer metrics, and aligning the organization around building customer value. Analytical CRM applies a variety of data analysis and modeling techniques to discover patterns and trends in customer data. It predicts potential variations so organizations can react to changes before it is too late. Decision makers and front line employees use these deep insights to understand what their customers want and predict what they will do next.
Armed with this information, companies can increase customer loyalty, share and future value - not only increase the total revenue from each customer but also increase more intangible benefits such as the number of referrals that a customer sends to the company.
Optimizing Customer Interactions Using Analytical CRM
In today's world, optimizing customer interactions is in many ways like replicating the old-fashioned general store experience. When one general store served everyone in a town, the owner knew everyone that came into the store. When a certain individual walked in, the owner knew that he was most likely looking for a certain product based on past buying habits. The owner knew that person's personality, what products he usually bought, and what likely would convince him to buy additional products.
Today companies can act more personally by using analytical CRM. There are four stages of using the analytical CRM process to understand and improve communications with customers: learn, interact, monitor and align.
In order to more fully explain these four stages, we apply each to the website business of the fictional company Rugged Lifestyles. Although the example involves a website, the principles behind analytical CRM are critical to the success of any type of business. (For a real life example of analytics in use in a traditional banking application, see "Banking on the Power of Analytical CRM").
Step One: Learn
Learning involves developing an initial understanding of your customers from the data you currently have. It involves segmenting the entire customer base into multi-dimensional groupings that incorporate characteristics such as value, loyalty and life stage, as well as demographics. Another key part of the learning process is to determine your best customer. Best customers are often defined as those that are most profitable, or those who spend the most money. However, they may also be identified by other characteristics such as the amount of valuable feedback or referrals that a customer provides a business within a year.
Once a company builds segments and determines who its best customers are, it can look for behaviors that are common about the segment, such as the tendency to buy certain products. It can develop target customer profiles, using demographics and customer behaviors, to predict what type of new individual could be a best customer in the future. The company can also use these profiles to predict what products its best customers are likely to buy to increase the total number or value of sales to those customers across their lifetime. It can then more effectively focus its resources on retaining those customers and attracting more like them.
For instance, using modeling, businesses can improve their acquisition programs by predicting what type of campaign will appeal to their best customer profile - even without sending promotions to those individuals who they have predicted are unlikely to reply. (See "Retaining Good Customers Using Analytical CRM").
The learning process also extends to the product arena, where companies try to determine which product to offer which customer. In this case, companies can uncover associations between certain products - thus helping determine which type of products are likely to be purchased by which individuals. They can also look at sequences, or the order of purchases, to determine buying patterns over time, such as the likelihood that a person that purchases one type of product will come back to buy another related product later.
Let's take a look at Rugged Lifestyles, a company that sells outdoor gear for city dwellers via the Web. Rugged Lifestyles realizes that it can increase revenue by personalizing the buying experience for its customers.
Rugged Lifestyles first step toward personalization is to uncover associations and discover which products group together in purchases naturally. Some of the clusters are obvious, such as shirts and pants. Others clusters are more subtle, such as books about desert hiking and snakebite kits.
Rugged Lifestyles also models sequences to determine purchase patterns and deliver the right product at the right time. For example, they found that individuals who buy camping tents later return to the site to buy portable gas heaters.
Finally, Rugged Lifestyles uses the customer data it has captured to build predictions that identify customers with a high likelihood to buy the new products that it frequently adds to the catalog. Next, they will build detailed profiles to determine what types of customers are its "best" website customers.
Step Two: Interact
The second step, interact, recognizes that knowing the customer and managing a relationship are two very different things. After all, one presumably can know someone very well, but it is only by applying this knowledge and reacting to that person differently based on, for instance, that person's emotional state, that a relationship improves. Likewise, knowing your customers is not useful unless your business applies this knowledge to better interact with them by reacting to them differently in different contexts.
This concept is often termed "personalization," where companies use information they have gathered on customers to improve their interactions with those customers. Personalization may be as simple as remembering what an individual has purchased in the past so that the business does not keep recommending a product that individual has already purchased. Or it may involve using that information to predict that a certain customer will react to an offer differently than another customer - and using a specific message to communicate with them.
At Rugged Lifestyles, the company uses customer and product data - and the predictive models it has created based on this data - to present the most likely action for each customer and increase sales. It then uses product associations to make additional recommendations whenever someone views a product Web page. It uses customer profile information to send emails about new products to specific customers with a high likelihood to be interested in those products. And, it uses sequence models to send emails for items such as portable gas heaters at the right time - a few weeks after a customer buys a camping tent.
Interacting with customers also means identifying the right place to sell to customers. For instance, before launching its Web site, Rugged Lifestyles researched its brick-and-mortar shopping population to ensure it would not simply be cannibalizing sales from this group by creating a Web site. It also polled the individuals shopping its catalog to determine if it might save money by launching the site because it could then send out fewer catalogs.
By combining additional research data to the customer data and applying analytics, Rugged Lifestyles found that individual who shopped using the catalog showed a high propensity to shop the Web site instead - thus leading to lower catalog costs. On-site shoppers, alternatively, liked to see their merchandise before purchasing it, meaning that there would be little attrition from the brick-and-mortar customers.
| Banking on the Power of Analytical CRM |
| Prior to implementing an analytical CRM program, a large U.S. bank often promoted its products to both new and existing customers using lifestyle segmentation information purchased from market research companies, such as studies that predict income and buying behavior based upon neighborhood or subdivision.
"That kind of external segmentation scheme has its place and can be valuable when soliciting new customers. However, we realized that we already had much more specific and potentially valuable information on the buying habits and needs of our 1.4 million customers already locked in our database," explained the manager of customer acquisition and research. "It was just a matter of mining the data and analyzing the patterns to learn more about who needs what and when. This predictive analysis helps us contact the right people at the right time with the right offer." The key is using analytical CRM to find customers that "look like" those that have demonstrated a particular buying behavior in the past. Take investment products, for example. Which personal characteristics and patterns in deposits to a checking account indicate a customer might be interested in a higher yielding investment? This buying behavior has occurred thousands of times in the past and can help predict buying behavior in the future, if the bank knows where to look. Mining sales data helps the bank uncover statistical relationships, and, more importantly, shows them the strength of those relationships, so they can instantly see what is meaningful. This helps the bank optimize resources in building an effective marketing strategy. Working with the bank's individual product departments, the research department has created successful marketing strategies based upon analytical CRM predictive models. In just three years, sales are up by up to 50% across the bank's numerous product lines. By targeting customers more accurately, the bank not only uncovers the most promising prospects for a particular product, they also save money by not contacting customers who do not fit the predictive profile. "If I'm doing a direct mail campaign, for example, I can mail out fewer pieces in a more targeted manner, get a higher percentage response rate and bring in about the same amount of money," the bank reported. "Avoiding the traditional 'shotgun' approach, in one recent promotion, we reduced a mail program by more than a third, saving thousands in postage and printing costs, but still brought in 95% of the revenues of the previous mailing. So our ROI was enormously improved."
|
Step Three: Monitor
Monitoring involves tracking metrics such as sales, customer loyalty and share to determine the impact of CRM activities on the business and customer value. Metrics include Key Performance Indicators (KPIs), which give decisions makers a quick view of what has happened, and Key Performance Predictors (KPP), which provide at-a-glance predictions for a view of the future.
An example KPI is retention rate, or the ratio of customers at the end of a period to the total customers during a period. Another example KPI is total number of sales. Each can be further broken down by customer segment, region, sales team, etc. An example KPP is churn propensity, or the likelihood that a current customer will stop buying. It is important for a company to examine both KPIs and KPPs, for the first determines current success and the second predicts future success - both of which are critical for a company's long-term survival.
For instance, Rugged Lifestyles' sales may be high, meaning its KPI looks good. However, when the company examines its KPP that measures churn propensity, or the likelihood that all of the new customers driving these high sales volumes will not buy again, it may find that this number is high. The low KPP predicts that the company's high sales KPI likely will not continue in the future. Rugged Lifestyles can then determine why this is happening and try to change it. Such predictive model-based indexes can be used to predict a customer's propensity to buy, churn or steal - all extremely critical information for a business.
A critical part of the monitoring process also involves validating the models and business rules that are established as part of the analytical CRM process. For example, if Rugged Lifestyles has established through its modeling that individuals who buy books about desert hiking also are likely to buy snakebite kits - but no customers who bought desert hiking books actually responded to a cross-selling pitch about snakebite kits - Rugged Lifestyles might determine that its model was ineffective.
| Retaining Good Customers Via Analytical CRM |
|
For almost every company, the cost of acquiring new customers exceeds the cost of keeping good customers. This was the challenge faced by one Internet service provider (ISP), which experienced the industry-average attrition rate, eight percent per month. With a customer base of 1 million, this eight percent churn rate meant that 80,000 customers left the company each month. The cost to replace these customers was $200 each or $16 million - plenty of incentive to start an attrition management program. To begin the attrition management program, first the ISP prepared the data used to predict which customers were likely to leave. The company needed to select the variables from its customer database and transform them. Since the bulk of the ISP's users were dial-in clients (as opposed to clients that are always connected through a T1 or DSL line), the company knew how long each customer was connected to the Web. The company also knew the volume of data transferred to and from a user's computer, the number of email accounts a user had, the number of email messages sent and received along with the customer's service and billing history. In addition, the company had demographic data that customers provided at sign-up. Next, the company needed to identify who were "good" customers. This is not an analytical question but a business definition followed by a calculation. The company built a model to profile its both its profitable and unprofitable customers. It then used this model not only for customer retention but also to identify customers who were not yet profitable, but might become so in the future. The company then built a model to predict which of its profitable customers would leave. As in most analytical CRM problems, determining which data to use and how to combine existing data provided much of the model development challenge . For example, this ISP needed to look at time-series data such as the month usage. Rather than using the raw time-series data, it smoothed the data by taking rolling three-month averages. The company also calculated the change in the three-month average and tried that as a predictor. Some of the factors that were good predictors, such as declining usage, were symptoms rather than causes that could be directly addressed. Other predictors, such as the average number of service calls and the change in the average number of service calls, the company thought might be indicative of customer satisfaction problems worth investigating. Predicting who would churn, however, wasn't enough. Based on the results of the modeling, the company identified some potential programs and offers that it believed would entice people to stay. For example, the company found that some churners were exceeding even the largest amount of usage available for a fixed fee and were paying substantial incremental usage fees. The company offered these users a higher-fee service that included more bundled time. Some users were offered additional free disk space to store personal Web pages. The ISP used models to predict which would be the most effective offer for a particular user. In short, to reduce churn, a company can make use of three models. One model identifies likely churners, the next model picks the profitable churners worth keeping and the third model matches the potential churners with the most appropriate offer. In the case of this particular ISP, the net result was a reduction in the company's churn rate from eight percent to 7.5 percent, which allowed the company to save $1 million per month in customer acquisition costs.
|
Step Four: Align
Aligning means changing a company's business processes to reflect what it has learned - particularly the insights it has gained through its KPIs, KPPs and predictive models. At Rugged Lifestyles, once the company determined that its desert hiking books/snakebite kits offer was not working, it stopped making that cross-sell product recommendation to customers. When the company's predictive models indicated that the customers it was attracting were likely to leave in just a few months, it implemented a update to its marketing programs to focus on a different type of customer.
| Pop Quiz: Are You A Customer-Centered Organization? |
|
To determine if your company really knows its customers, see if you can answer the following questions with specific examples: Do you know who your best customers are? Do you identify customers and treat them differently? Do you personalize your customers' experiences? How? Do you learn from customer interactions and improve processes based on their feedback? Do you know if you are the sole provider for each of your customers? Do you measure customer share of wallet or individual potential customer revenue? Do you measure the success of your company in terms of customer value? |
Aligning also means changing a company's business processes to incorporate its new focus on the customer - and measuring success in terms of that new focus. For instance, Rugged Lifestyles traditionally measured the performance of its inventory managers by how quickly it sells off inventory, or its inventory turn. However, in addition to being counter-productive - itencouraged the manager to keep inventory levels low and resulted in lost sales due to inventory not being available - inventory turn was not a driver of customer loyalty, share or value.
Using analytical CRM, Rugged Lifestyles reacted by changing the way it measures the performance of inventory managers. They found that measuring inventory fill - or the number of orders that are successfully filled - was related to customer loyalty and thus a more critical gauge of success. By using analytical CRM to change its business processes to reflect the needs of the customer, a company can become truly customer centric.
Conclusion
Analytical CRM offers the missing link to truly understanding customers: prediction. Prediction helps companies take their CRM systems to the next level, using the value of customer information to optimize each interaction with every customer. Armed with information about the future, companies can use analytical CRM to find out things they've never known about their customers, and use it to say just the right thing at just the right time. (To see when your company can improve its business using analytical CRM, read the "Pop Quiz: Are you a Customer-Centered Organization?")
In order to achieve this goal, companies must follow a four-step process to learn about their customers, using this knowledge to optimize interactions with customers, monitoring these interactions to ensure they are working and aligning their business processes to reinvent their business around customer interactions. The result will be a truly customer centricorganization that makes more revenue by keeping the right customers over a longer period of time.
About CustomerCentric Solutions
CustomerCentric Solutions, a division of SPSS, Inc., provides a combination of software and analytic consulting services that drive greater value from Customer Relationship Management (CRM) initiatives. Its flagship product, the CustomerCentric analytical CRM soluton, is designed for companies that recognize customers are their greatest asset. Companies that use CustomerCentric optimize revenue by increasing the number of customers, their share, and their loyalty. CustomerCentric aligns an organization around customer value, provides deeper customer understanding, and drives more effective interactions with customers. To learn more, call the SPSS toll-free number, 800-543-2185 ext. 3673, or visit CustomerCentric. Send email to ccsinfo@spss.com.
