Customer understanding and insight enable superior service.

Delivering a satisfying customer experience that enhances loyalty should be simple. If we handle their call efficiently, treat them with respect, and resolve their issue, the experience should be positive. However, too many firms continue to evaluate subjective agent behaviors such as tone of voice rather than listening to what the customer is saying, the voice of the customer.

Ensuring customer engagement and loyalty requires an understanding of customers’ mandatory requirements, their service preferences, and potential service differentiators. Some are as simple, others more complex.

The chart below identifies the options for generating customer interaction insights.

While all of the activities listed can deliver some insights, the most effective approaches, located in the top right quadrant, rely upon statistically valid customer data and avoid the limitations associated with the less effective approaches: small sample size, anecdotal data, statistical skew associated with poor survey design, and dated (potentially irrelevant) information.

Kano methodology, an approach that leverages paired responses to functional and dysfunctional questions, is the best practice approach for defining customer requirements, service preferences, and potential differentiators. Responses to paired questions are mapped to a response scale and clearly identify whether a customer requirement is a Must-Be (a mandatory requirement with resulting dissatisfaction if not met), One-Dimensional (preferences for service with the ability to satisfy or dissatisfy the customer based upon performance), Attractive (unexpected and thus potentially differentiating), or Indifferent (irrelevant to the customer, potentially an invalid bias held by the organization or leadership).

The charts below demonstrate how this methodology is used providing the functional and dysfunctional question forms, response options, and how the response options are matched and drive behavior classifications.

In addition to the functional/dysfunctional questions, a self-rated ten-point importance scale allows the organization to identify the relative priority of requirements. The self-rated importance scale drives potential behavior prioritization and focus and is illustrated below with a sample question and results.

An alternate methodology, the call observation (monitoring) and survey approach, relies upon objective call observations and Likert surveys (agreement response scale) to correlate identified call handling activities and outcomes with customer perceptions of service. Extensive data is collected from call observations cataloging details on agent and customer behaviors as well as call outcomes. Customers whose calls have been monitored are then surveyed within 24-48 hours of the observed interaction to gauge the customer’s perception of service and provide the basis for correlating call activities and outcomes with customer preferences.

It is important to note that not all customers have the same requirements and preferences. With McIntosh, you will gain the ability to identify through Kano and other tools, customer segmentation opportunities providing the operations team the ability to vary routing, automation options, and call handling treatment by customer segment.

For both Kano and the monitor/survey processes, a minimum of 1,080 surveys must be completed to achieve a 95% confidence level, +/- 3% margin of error. For customers with less stringent requirements, we can generate a margin of error of +/-5% with a sample size of 384.

On an ongoing basis, best practice firms continue to measure customer satisfaction, engagement and loyalty, and to adjust their customer experience and service strategy based upon actionable customer data driven by analytics.

Ongoing measurements of satisfaction and loyalty can be obtained through surveys, customer churn metrics, and annual revenue changes.

Forrester’s Customer Experience Index (CSI) and Bain’s Net Promoters Score (NPS) are two recognized measures of customer engagement and loyalty that are considered best practice approaches. McIntosh prefers the NPS approach but can support any tool preferred by the customer.

Whatever methodology is selected for measuring enterprise customer satisfaction and loyalty, it should be done on a monthly basis so that early identification of negative trends can be acted upon. Accompanying corporate measures should be a call center specific survey that can be completed and reported from site level through agent level. This customer engagement/loyalty score would become a part of the call center scorecard for all levels of the call center organization.

The call center survey should be no more than five questions, one of which is the likely to recommend question, and should be conducted over the channel the customer used to contact the company (e.g., email survey for email queries, online survey at the end of a chat session, post-call VRU survey for voice calls).  Customers should be offered a survey no more often than twice annually.

Results should be aggregated for site leadership and managers and also reported on a daily/monthly basis.

Success Story

A major hotel chain encountered agent resistance to survey scores with agents questioning the validity of the data and complaining that one or two difficult customers could damage their results. The firm expanded the sample size and dropped the lowest score on a monthly basis to eliminate the random occurrence of a single, impossible to please customer.

As with all metrics, immediate reporting including MTD performance is critical to providing actionable data.

Beyond agent through leadership reporting, the customer surveys can also provide valuable product related insights, competitive insights, and may be an early warning system for identifying evolving customer requirements and preferences.

While big data is conceptually appealing at the enterprise level, the allure of actionable analytics is even greater for call center operators.

On a daily basis call centers organize massive data streams into useable information, creating knowledge bases that inform decisions and enable action. The pragmatic appeal of actionable analytics is our ability to use data to transform operations and differentiate service.

Analytics enabling call center decisions and action include:

• Performance dashboards/scorecards, enabling us to maintain focus and identify trends at theindividual through site levels as well as product trends.
• Speech analytics, the ability to analyze every call for customer, operational, product, and competitive insights.
• Text analytics, the analysis of electronic content regardless of source; most useful in a technology support center where chat is a common tool.
• Predictive analytics, the ability to analyze, model, and mine current and historical data to predict the future.
• Social analytics, the ability to collect and analyze data from evolving social platforms, a turbulent space made frothy by hype. Data indicates that customers who engage on social platforms are more intense in their loyalty and drive higher annual revenues per customers than non-social customers.
• Customer segmentation modeling, the ability to append historical or known preference data to a customer’s profile and thereby drive treatment options ensuring premium treatment for premium customers whether individuals or businesses.
• Multivariate testing, the ability make informed decisions by testing multiple variables and only deploying based on proven results.