|
|
Chapter 18
|
It's not who I am underneath, but what I do that defines me.
Batman Begins – 2005
When you have completed this chapter you will be able to:
• Understand the Single Customer View (SCV) conceptual data model
• Appreciate the benefits of the customer-centered orientation
• Specify the dimensions of customers
• Explain the benefits of customer-related analytics
• Segment customers using the RFM (Recency, Frequency, Monetary Gain) approach
• Categorize data into bands such as age band and education band
• Utilize the concepts of demographics, behavior, psychographics, interests, and transactions.
In Chapter 17, you learned about business intelligence applications and how those applications can help you to achieve benefits. The possibilities for using information to improve results are limitless. Examples of these applications include budgeting, logistics, manufacturing, and employee performance. One area of analytics stands out in the benefits that can be derived – customer analytics.
In this chapter, you will learn about the application of analytics to customers. Most organizations are heavily invested in creating and maintaining positive and profitable relationships with their customers. The importance of customers extends beyond for-profit businesses. The customers of governmental units are tax payers and specific recipients of services. The customers of non-profit organizations are the receivers of benefits.
If you are like most people, you have many questions about your customers. Use of customer analytics is a way to obtain answers to those questions. Table 18-01 shows questions that customer analytics can help you answer. Take time to review these questions and identify questions that are most important to you and your organization.
Table 18-01: Customer-Related Questions
|
Category |
Question |
|---|---|
|
Who |
Who are our customers? • Identifiers –what identifiers set our customers apart? • Profiles – what kind of people are our customers? Who are our best customers? Who are our most profitable customers? Who is delinquent in paying their bills? Who is saying good things about us? (Promoters) Who is saying negative things about us? (Detractors) Who is harming our organization? (Committing fraud, etc.) |
|
What / Which |
What is a customer? What do our customers want? Which products or features are growing or declining in popularity? Which products should we offer a given customer? Which products are purchased together (market basket analysis)? Which offers will be responded to by which customers? |
|
When / How Often |
When is the best time to approach a customer? How often should we approach a customer? |
|
Where |
Where are our customers located geographically? Where do our customers go on our website? Where can we obtain customer data? Where should we open a new location? Where should we close a location? |
|
How |
How can we identify our customers? How do our customers learn about us? How can we sell more? How can we retain our customers? How can we better serve our customers? How can we detect when a customer is likely to leave? How are our customers related to each other? How do our customers want to be contacted? |
|
How Much |
How much money is spent by our customers? How much of a market share / wallet share / mind share? How much money was spent on each marketing campaign last month? How much should we charge for our products? How much credit should we extend to a given customer? |
|
How Many |
How many customers do we have and in what categories? |
|
Why |
Why is it a benefit to understand customers? Why do customers buy? Why do customers not buy? Why do customers leave? Reasons for attrition. |
Customer Analytics is an approach to understanding customer behavior whose critical information is integrated in master data management, data warehousing, and other systems. This approach has some excellent benefits and uses:
• Improved customer service
• Cross sell and up sell opportunities
• Customer segmentation to enable targeted marketing and sales campaigns
• Improving customer loyalty
• Avoiding loss of profitable customers (retention).
For example, Larry Selden and Geoffrey Colvin found that the top 20% of a company's customers generate 150% of the company's profit, while the bottom 20% of customers drain 80% of the profits. (Selden 2003).
Successful organizations take action to understand their customers. ESPN, the sports network, has studied its customers and improved its offerings. Caesars Entertainment, the casino company, has built excellent relationships with its customers through analytics.
Single Customer View
The major obstacle that stands in the way of understanding and serving customers is that data about each customer is gathered in multiple places and cannot be integrated easily, as illustrated in Figure 18-01.
Figure 18-01: Customer Interactions
Single Customer View (SCV) is an approach that unifies information about customers so that all relevant activity and facts about each customer can be seen at a single point. One way to achieve this is Customer Data Integration (CDI), which is the process of integrating information about customers from multiple data sources into a single database.
Integrating customer data means associating all data about each customer to a common identifier or identifiers. This is a challenge, because customers may be associated with different identifiers in different systems. In addition, spellings of names and addresses may differ by system. Table 18-02 shows how data for a single customer can vary between and within systems.
Table 18-02: Customer Data Varies by System
|
System |
Example Data |
|---|---|
|
Help Desk |
User Id = Bob29431 Email Address = [email protected] Mobile Phone = 612-555-4567 Name = Robert Smith |
|
Sales Order |
Account Number = 89431 Name = Robert A Smith Telephone = 612-555-3579 Address = 100 East Main Street, Anoka, MN 55443 Email Address = [email protected] |
|
Warranty |
Warranty Number = 100456 Name = Robert A Smith Telephone = 612-555-3579 Address = 100 East Main Street, Anoka, MN 55443 Warranty Number = 5431099 Name = Bob Smith Telephone = 612-555-3579 Address = 100 E Main St, Anoka, MN 55443 |
|
Syndicated Data |
Name = Robert A Smith Address = 100 East Main Street, Anoka, MN 55443 Age Band = 5 (30-35 years) Income Band = 4 ($50,000 to $60,000) Lifestyle Segment = Upcoming Married Professional |
|
Survey System |
Email Address = [email protected] |
CDI software includes matching capabilities that enable matching data variations, such as spelling differences, format differences, address differences, nick names, and abbreviations. The integration point is the SCV Hub, illustrated in Figure 18-02.
Figure 18-02: SCV Integrates Customer Data from Multiple Sources
Common Customer Data Model
At the heart of SCV is a data model that supports the integration of customer data. Raw data is gathered for use in later analysis for profiling and segmenting customers. This includes detailed information that is accumulated in the SCV Hub and/or the data warehouse, such as identifiers, demographics, psychographics, location, products, measures, transactions, and interactions. See Figure 18-03 for examples of this.
Figure 18-03: SCV Data Model
Identifiers
Identifiers are data elements that establish the unique identity of customers. They are the connecting points that enable the building of the single customer view using data from many sources. Frequently used identifiers include account number, DUNs number, and driver license number. Sometimes identifiers must be combined with other information such as birth date, name, and zip code to establish a unique identifier for an individual.
Demographics
Demographics are descriptive information about a person that are the sorts of information gathered through the census to understand a population. Helpful demographic data elements include birth date, gender, marital status, number of dependents, occupation, and education level.
Psychographics
Psychographics are data inferred about the inner customer. This includes attitudes, beliefs, lifestyles, opinions, interests, motivations, preferences, and other psychological dimensions. This information is often discovered through surveys and social media research.
Locations (Geographics)
Location data is the answer to the “where” question. Postal address, work location, telephone numbers, and email addresses describe physical and virtual locations.
Products
Product data describes the goods and services that the customer has purchased or has expressed an interest in. For a bank, this may be a listing of the types of accounts held by the customer. It can also include a list of products ordered. This information is useful for market basket analysis, as well as cross sell and up sell efforts.
Interactions
Interactions are customer touch points, where the organization and the customer are in contact. Interactions could be initiated by the customer or by the organization. Complaints, website visits, and warranty service requests are examples of customer initiated interactions. Outgoing sales calls and marketing campaigns are examples of organization initiated interactions. Building positive interactions is a key to building customer loyalty and retention.
Transactions
A transaction is a unit of business exchange such as sales, shipments, deposits, withdrawals, and investments. Traditional data warehousing tended to focus on transactions, however, there is a marketing tendency to focus on interactions and relationships, rather than transactions.
Measures
Measures are quantitative metrics such as account balances, expenses, fees, and overdraft amounts. Some measures, such as Lifetime Customer Value (LCV), are calculated. For example, LCV may be calculated as the sum of projected revenues minus the sum of projected expenses discounted by a specified interest rate.
Relationships / Hierarchies
Relationships are associations between customers. For example, an individual may be a holder of a small account at a bank, while at the same time may be a CEO of a company with a large account. In this case, it would be an advantage to the bank to provide excellent service on the small account.
Single Customer View Tips and Traps
Following best practices in the area of Single Customer View increases the likelihood of success. Table 18-03 highlights tips and trips of SCV.
Table 18-03: Single Customer View Tips and Traps
|
Tips (Do This) |
Traps (Don't Do this) |
|
• “Start small, think big” • Get multiple perspectives on customers • Use Customer Data Integration (CDI) technologies • Create a common customer model • Include relationships between customers and their accounts • Consider external data sources |
• Attempt to integrate all customer data at once • Fail to promote use by people who interact with customers • Fail to think from the customer perspective • Leave out the full picture when calculating customer value • Confuse CDI with DW and BI • Use duplicate and disparate data |
Customer Segmentation
Customer Segmentation is the classification of customers for the purpose of improving profitability and/or achieving the organization's mission. Uses of customer segmentation can include:
• Determining the level of customer service
• Determining what to offer to the customer
• Determining whether to drop the customer.
There are a number of ways to segment customers. One of the most effective methods is through the use of segmentation factors that score or rank customers. The following steps are performed to segment customers using RFM (Recency, Frequency, and Monetary) analysis:
• Determine quintile ranking for each factor (quintiles are calculated by breaking data into five groups. These factors include:
o Recency – by date of most recent purchase
o Frequency – count of purchases over some time period
o Monetary – Dollar amount of purchases.
• This results in 3 sets of numbers with values 1 to 5.
• Each combination is a cell within a cube. This results in 125 cells for analysis.
• Each cell can be assigned a customer score and a treatment approach.
• Customers in cell 1-1-1 are the best customers. They buy the most goods, most often, and have bought most recently.
• Customers in cell 5-5-5 are the lowest ranked customers. They buy the fewest goods least often and have not purchased for a long time.
There are many characteristics which may be used to segment customers. Table 18-04 shows dimensions of demographics, behaviors, and interests that are supported by individual characteristics. The ≪xxx≫ entries are variables which apply to a particular customer group such as ≪plans to purchase a home ≪ or ≫ wants to retire early ≪.
Bands
A band is a grouping based on a range of values like those in Table 18-05. Bands support customer segmentation and simplify statistical analysis.
Figure 18-05: Customer Band Examples
|
Age Bands |
Income Bands |
|
Education Level Bands |
|
Under 18 18 – 25 26 – 35 36 – 45 46 – 55 56 and above
|
Less than $10,000 $10,000 – $35,000 $35,001 – $50,000 $50,001 – $70,000 $70,001 – $100,000 More than $100,000
|
|
Not high school graduate High school graduate Some college Bachelor's Master's Ph.D. |
Table 18-04: Customer Dimensions and Attributes
|
Demographics |
Behavior |
Interests |
|
|
Name Gender Birthdate Income Level Education Level IQ Marital Status Number of Children Household Size Living Parents Language Ethnicity Race Occupation Net Worth Credit Score Clothing Size Height Weight Accident Count Felony Count Relationships with other parties Telephone Number Address Length of Residence Zip Code |
ATM Usage Count Returns products Requests service Mail Order Buyer Files claim Engagement Level Eats out Loyalty Level Purchases by internet Purchases by direct mail Answers email Pays by credit card Pays by debit card Pays by check Voting pattern Uses coupons Pays in full Pays minimum balance Uses Overdraft Risky behavior Visits to branch/store Files complaint Answers survey
|
Automobile Preference Hobbies Volunteer Products Purchased Business Owner Favorite brands Plans to xxx Wants to xxx Likes xxx Dislikes xxx Owns xxx Reads xxx Watches xxx Invests in xxx Fears xxx
|
|
Clusters
Clustering is a statistical technique for segmenting a population. This can include target market and customer service clusters. Acxiom PersonicX® provides segmentation on millions of United States consumers based on detailed information gathered about the American population. It has clustered people into 21 life stages divided into 70 segments/clusters. Examples of life stages include:
• “Gen X Singles” – Single households without children and low-middle income.
• “Cash & Careers” – Affluent people born in the mid-1960's and early 1970's who are childless and are aggressive money earners and investors.
• “Boomer Barons” – Wealthy baby boomers with high education levels who enjoy luxury homes.
• “Mature Rustics” – Near retirement blue-collar people with country values.
This information can be helpful in both understanding an overall market and in targeting offers to individual consumers. For example, a bank may target younger consumers who have the potential to become valuable customers.
Customer Analytics Terms
Customer analytics has its own terms, as defined in Table 18-06.
Table 18-06: Customer Analytics Terms
|
Term |
Definition / Descriptions |
|---|---|
|
Band |
A band is a grouping based on a range of values such as age band and income band. Bands simplify statistical analysis. |
|
Behavior |
Actions taken by a party that are evaluated through analytics to gain insight into potential future behavior. |
|
Campaign |
A series of activities intended to market a brand, product, or service. Ideas for campaigns may be the result of customer analytics. |
|
Communication Channel |
A medium for communicating, such as telephone, email, mail, television, radio, or in person. Determining the effectiveness of communication channels enables more effective allocation of resources. |
|
Clustering |
Clustering is a statistical technique for grouping a population based on a shared characteristic. Factors commonly used in clustering include age, purchasing habits, income level, wealth level, education level, and preferences for luxury goods. |
|
Cross-Sell |
Cross-selling is the practice of selling multiple products and services to existing customers. Opportunities for cross-selling are the result of evaluating previous purchases and other customer characteristics. |
|
Customer Data Integration |
Customer Data Integration (CDI) is the process of bringing together information about customers from multiple data sources into a single database. |
|
Customer Relationship Management |
Customer Relationship Management (CRM) is an enterprise-wide strategy that seeks to make organizations customer-centric by finding, attracting, retaining, and serving customers. CRM mobilizes people, processes, and technologies to enhance customer relationships. It is a way of life and thinking, not just a software package. |
|
Demographics |
Information that describes characteristics of people such as gender, birth date, income level, and education level that may be used in analytics to segment the population to evaluate and predict future business opportunities. |
|
Detractor |
A party who communicates negative information about an enterprise, brand, product, or service. Detractors feel negatively about an enterprise and are unlikely to recommend it to others. It is important to understand what they dislike to determine if there are factors within the business that need to be remediated. |
|
Engagement |
Activities involving a party interacting with an enterprise. A party is engaged when they interact by visiting a website or store, for example. Analytics evaluate the cause and results of a customer's engagement with an organization. |
|
FICO |
A credit score produced by the FICO corporation. The FICO score is frequently used to evaluate credit worthiness. |
|
Hierarchy |
A categorization that uses ranked order. For example, a geographic hierarchy could be organized with increasing rank by city, state, region, nation, and continent and used to determine similarities and differences in customers at the same or different levels of the hierarchy. |
|
Interaction |
An event during which a customer was in contact with the organization. There are a wide variety of interaction types, based on purpose and communication medium. Each interaction is an opportunity to gather more information about the customer and to provide an enhanced experience for them. |
|
Interests |
Something that holds the attention of somebody on an extended basis, such as a subject, hobby, or sport. Knowing a customer's interests provides some insight into how they may behave in relation to what the organization has to offer. |
|
Lifetime Customer Value |
Lifetime Customer Value (LCV) is a marketing metric that is calculated by summing the net present value of revenues and expenses projected for a customer relationship. |
|
Customer Loyalty |
Customer loyalty is a marketing metric that may be measured by: • Customer retention/defection rates • Referral count • Net Promoter Score • Tendency to purchase again or purchase different products |
|
Marketing Channel |
The path where products flow from producer to consumer. This path may be direct, where consumers buy directly from the producer, or complex, where many intermediaries are involved, such as wholesalers. Financial products might have a number of marketing channels such as: banks, wire houses, broker dealers and insurance agencies. Understanding the needs of marketing channels can lead to more effective product design, for example. |
|
Personalization |
The process of tailoring an interaction to the person involved and their preferences. Personalization could include showing information and products of interest. |
|
Preference |
A choice made by a person, such as the type of preferred communication like email versus telephone or the frequency of communication like daily, weekly or monthly. The use of preferences requires capturing and storing those preferences. |
|
Profile |
A set of descriptors about a party that may have a specific subject such as demographics profile, interests profile, and behavior profile. Profiles are an aid to understanding customers and categorizing customers. |
|
Promoter |
A party who communicates positive information about an enterprise, brand, product, or service. Promoters feel positively about an enterprise and are likely to recommend it to others. The Net Promoter Score (NPS) is a method used to determine the promotion and detraction levels. |
|
Psychographics |
A customer segmentation method that uses criteria such as feelings, lifestyle, attitudes, personality, and motivation. |
|
Relationship Marketing |
A marketing approach that emphasizes a long term association with retention, rather than individual sales transactions. |
|
Recency, Frequency and Monetary Analysis |
Recency, Frequency, and Monetary (RFM) analysis results in a marketing metric that segments customers based on: • Recency of purchase – the more recent, the higher the score • Frequency – the more often, the higher the score • Monetary – the larger the purchase, the higher the score |
|
Segment |
A method of categorizing customers by divided them into groupings for analysis and marketing purposes. |
|
Single Customer View |
An approach that unifies information about customers so that all relevant activity can be seen at a single point. This often includes use of a database where customer data is integrated. |
|
Syndicated Customer Data |
Data about parties (individuals and/or organizations) provided by an external organization. For example, some organizations gather profiles of United States residents and make that data available to marketers. |
|
Touch Point |
A touch point is the same as an interaction. |
|
Up Sell |
Attempting to sell a higher end product or service to somebody who has purchased a lesser product or service. Customer analytics may be used to determine who to up sell to and which products to offer. |
Analysis Types
There are many kinds of customer analysis. Table 18-07 provides an overview of several of these.
Table 18-07: Customer Analysis Types
|
Analysis Type |
Description |
|---|---|
|
Campaign Analysis |
Campaign Analysis is an investigation into the effectiveness of marketing efforts. It tracks efforts to improve sales and the results of those efforts with a goal of making the best use of organization resources. |
|
Call Center Management |
Call Center Management is the direction and control of the call center, which is a group that provides service to customers over the telephone or internet. |
|
Customer Churn Analysis |
Customer Churn Analysis is an assessment of the degree that customers leave a company. Churn rate is another name for retention rate and is the inverse of customer loyalty. |
|
Customer Satisfaction Analysis |
Customer Satisfaction Analysis is an investigation about the feelings and perceptions of customers about a supplier. This analysis seeks to find the factors that cause satisfaction and dissatisfaction. |
|
Market Basket Analysis |
Market Basket Analysis is an investigation of the combinations of products that a customer purchases. |
|
Market Share Analysis |
Market Share Analysis is a procedure for determining the percentage of sales that a company, brand, or product has within a particular market area. |
|
Market Analysis |
An investigation of a market to understand its characteristics, including market size, growth rate, trends, and critical success factors. |
|
Sales and Profitability Analysis |
An assessment that compares the costs of sales with the sales revenue. |
|
Sales, Marketing, & Channel Management |
Analysis and management needs of marketing channels can lead to more effective allocation of marketing resources. |
|
Store Operations Analysis |
The use of analytic techniques to improve management of retail stores. |
|
Subscriber Usage Pattern Discovery |
The use of data analysis to understand how customers who subscribe to websites use those websites and the services within them. This includes navigation patterns, download usage, and timing. |
|
Warranty Analysis |
The analysis of repairs and services. |
Learn More
Build your know-how in the areas of marketing and customer using these resources.
|
Visit a Website! |
Wikipedia provides a good introduction to customer analytics: http://en.wikipedia.org/wiki/Customer_analytics |
|
Get Research! |
Search the web for research reports (filetype=pdf): • Forrester Wave Marketing • Gartner Magic Quadrant CRM
|
|
Read about it! |
Try this book: Blattberg, Robert C. Database Marketing: Analyzing and Managing Customers. Springer, 2009.
|
|
Key Points • Understanding customers can result in improved customer service and loyalty. • Customer analysis can help answer many questions about the customers, such as “What is the life-time value of this customer?” • Single Customer View provides an integrated view of customer information. Customer identifiers such as account number and DUNs number support building an integrated customer view from multiple data sources. • Demographic information such a birth date, gender and marital status are important to analyzing customers. • Psychographic dimensions are also significant factors when analyzing customers. • Customer segmentation is a method of classifying customers in order to improve profitability and achieve enterprise goals. • RFM (Recency, Frequency, and Monetary) Analysis is a proven method for segmenting customers for many environments. |
|