Customer feedback professionals are asked to demonstrate the value of their customer feedback programs. They are asked: Does the customer feedback program measure attitudes that are related to real customer behavior? How do we set operational goals to ensure we maximize customer satisfaction? Are the customer feedback metrics predictive of our future financial performance and business growth? Do customers who report higher loyalty spend more than customers who report lower levels of loyalty? To answer these questions, companies look to a process called business linkage analysis.
Business Linkage Analysis is the process of combining different sources of data (e.g., customer, employee, partner, financial, and operational) toÂ uncover important relationships among important variables (e.g., call handle time and customer satisfaction). For our context, linkage analysis will refer to the linking of other data sources to customer feedback metrics (e.g., customer satisfaction, customer loyalty).
Business Case for Linkage Analyses
Based on a recent study on customer feedback programs best practices (Hayes, 2009), I found that companies who regularly conduct operational linkages analyses with their customer feedback data had higher customer loyalty (72nd percentile) compared to companies who do conduct linkage analyses (50th percentile). Furthermore,Â customer feedback executives were substantially more satisfied with their customer feedback program in helping them manage customer relationships when linkage analyses (e.g., operational, financial, constituency) were a part of the program (~90% satisfied)Â compared to their peers in companies who did not use linkage analyses (~55% satisfied). Figure 1 presents the effect size for VOC operational linkage analyses.
Linkage analyses appears to have a positive impact on customer loyalty by providing executives the insights they need to manage customer relationships. These insights give loyalty leaders an advantage over loyalty laggards.Â Loyalty leaders apply linkage analyses results in a variety of ways to build a more customer-centric company: Determine the ROI of different improvement effort, create customer-centric operational metrics (important to customers) and set employee training standards to ensure customer loyalty, to name a few. In upcoming posts, I will present specific examples of linkage analyses using customer feedback data.
Linkage Analysis: A Data Management and Analysis Problem
You can think of linkage analysis as a two-step process: 1 ) organizing two disparate data sources into one coherent dataset and 2) conducting analyses on that aggregated dataset.Â The primary hurdle in any linkage analysis is organizing the data in an appropriate way where the resulting linked dataset make logical sense for our analyses (appropriate unit of analysis). Therefore, data management and statistical skills are essential in conducting a linkage analysis study. More on that later.
Once the data are organized, the researcher is able to conduct nearly any kind of statistical analyses he/she want (e.g., Regression, ANOVA, Multivariate), as long as it makes sense given the types of variables (e.g., nominal, interval) you are using.
Types of Linkage Analyses
In business, linkage analyses are conducted using the following types of dataÂ (see Figure 2):
- Customer Feedback
Even though I discuss these data sources as if they are distinct, separate sources of data, it is important to note that some companies have some of these data sources housed in one dataset (e.g., call center system can house transaction details including operational metrics and customer satisfaction with that transaction). While this is an advantage, these companies still need to ensure their data are organized together in an appropriate way.
With these data sources, we can conduct three general types of linkage analyses:
- Financial: linking customer feedback to financial metrics
- Operational: linking customer feedback to operational metrics
- Constituency: linking customer feedback to employee and partner variables
Before we go further, I need to make an important distinction between two different types of customer feedback sources: 1) relationship-based and 2) transaction-based. In relationship-based feedback, customer ratings (data) reflect their overall experience with and loyalty towards the company. In transaction-based feedback, customer ratings (data) reflect their experience with a specific event or transaction. This distinction is necessary because different types of linkage analyses require different types of customer feedback data (See Figure 3). Relationship-based customer feedback is needed to conduct financial linkage analyses and transaction-based customer feedback is needed to conduct operational linkage analyses.
The term “linkage analysis” is actually a misnomer. Linkage analysis is not really a type of analysis; it is used to denote that two different data sources have been “linked” together. In fact, several types of analyses can be employed after two data sources have been linked together. Three general types of analyses that I use in linkage analyses are:
- Factor analysis of the customer survey items: This analysis helps us create indices from the customer surveys. These indices will be used in the analyses. These indices, because they are made up of several survey questions, are more reliable than any single survey question. Therefore, if there is a real relationship between customer attitudes and financial performance, the chances of finding this relationship greatly improves when we use metrics rather than single items.
- Correlational analysis (e.g., Pearson correlations, regression analysis): This class of analyses helps us identify the linear relationship between customer satisfaction/loyalty metrics and other business metrics.
- Analysis of Variance (ANOVA): This type of analysis helps us identify the potentially non-linear relationships between the customer satisfaction/loyalty metrics and other business metrics. For example, it is possible that increases in customer satisfaction/loyalty will not translate into improved business metrics until customer satisfaction/loyalty reaches a critical level. When ANOVA is used, the independent variables in the model (x) will be the customer satisfaction/loyalty metrics and the dependent variables will be the financial business metrics (y).
Business linkage analysis is the process ofÂ combining different sources of data toÂ uncover important insights about the causes and consequence of customer satisfaction and loyalty. For VOC programs, linkage analyses fall into three general types: financial, operational, and constituency. Each of these types of linkage analyses provide useful insight that can help senior executives better manage customer relationships and improve business growth. I will provide examples of each type of linkage analyses in following posts.
Download a free white paper titled, “Linkage Analysis in your Voice of the Customer Program.”