Planning for Customer Analytics (Part 3 of 4)
Analytics can help you improve customer experience and customer value for your enterprise, but it requires both data management and the careful choice of analytics techniques.
- By David Loshin
- April 20, 2016
To determine the best methods and techniques to adopt for your customer analytics, you must first identify specific opportunities to improve the customer's experience along each stage of the customer life cycle.
In my previous article, I looked at improving just one part of the cycle: building product awareness among prospective customers. I included suggestions for optimization such as website visitor/web log analysis, location analytics, prospect profiling, and customer clustering.
Of course, there are more opportunities for improvement across different stages of the customer life cycle -- speeding the time for evaluation and consideration, improving the customer's selection process, speeding the time to purchase, increasing sales through upselling and cross-selling, growing customer loyalty, reducing customer attrition, and engaging for advocacy.
Note that some of these opportunities improve the customer's experience (such as simplifying the evaluation process), others increase the customer's lifetime value to the organization (such as increasing sales), and some may achieve both goals.
To design and engineer an environment that addresses these opportunities, you must consider two key elements: the analytics techniques and the dependent data.
The environment must provide the analytics toolkit -- methods for clustering, segmentation, classification, affinity grouping, market-basket analysis, etc. -- as well as the foundation for managing the data associated with the subjects of analysis, such as customers and products.
Effective planning for customer analytics must balance three capabilities:
- Customer data management focuses on designing master data models for persisting the characteristics of the customer entity employed in analytics processes. This will combine demographic attributes (e.g. sex, birth date, and other inherent descriptive attributes), contextual attributes (such as location, profession, or level of education), and behavioral attributes (such as purchasing preferences, methods of payment, or preferred leisure activities).
- Analytics methods span the technique as well as the tools and technologies used to apply the technique. For example, customer retention analysis is a technique that uses customer profiling (to ensure complete customer records), clustering and segmentation (to organize the population into groups with similar characteristics), and anomaly detection (to identify individuals or behaviors that vary from the norm for each customer segment), among others.
- Operational procedures are used by analysts to direct the ways that business processes incorporate analytics results. This may involve adjusting the way a standard operation procedure uses analytics to improve the expected outcome.
An example that demonstrates the dependence on these three capabilities is call center customer retention. When a disgruntled customer calls, the customer service representative must determine the best way to retain the customer. If you want to optimize customer retention in a way that maximizes the customer lifetime value, you can apply analytics in two different ways.
The first way is to determine customer lifetime value; this requires customer profiles, clustering and segmentation, and projection of future value based on behavioral characteristics. The second way is to identify the best potential retention offers to present to the customer, which relies on historical analysis of retention among others within the same cluster.
The desired outcome is for the customer service representative to present an offer that costs the organization less than the remaining customer lifetime value. With analytics it may also be possible to gain greater insight into the aggregate value of retention -- some customers may be critical to retain but in other cases the analysis may show that the lifetime value is not significant enough to warrant retention!
This example shows how we can use analytics to question traditional presumptions. Such insights may warrant changes in standard procedures and suggest ways that team members can be trained to integrate analytics into their everyday activities.
Other articles in this series:
Part 1: Analytics for Customer Engagement
Part 2: Optimizing the Stages of the Customer Life Cycle
Part 4: Managing Customer Data for Analytics
David Loshin is a recognized thought leader in the areas of data quality and governance, master data management, and business intelligence. David is a prolific author regarding BI best practices via the expert channel at BeyeNETWORK and numerous books on BI and data quality. His valuable MDM insights can be found in his book, Master Data Management, which has been endorsed by data management industry leaders.