Making Your CX Data Meaningful to your Business…
To those of us in the Customer Experience (CX) business, it is encouraging to see the continued interest and usage in CX metrics within corporate environments. However, as we recently wrote in a published article, not all metrics are created equal. Many CX metrics still struggle to have a clear, quantifiable value to Executive Boards, or even to one’s own customers.
The key reason for the metric “vagueness” resides in the underlying methodology. Some CX methodologies take a silo-approach to customer feedback, while others exclusively look at agent action (key word “action”), assuming that either of these perspectives will provide a comprehensive understanding in order to develop a scalable approach to serving customers. These methods refuse to recognize that any singular perspective will be full of misunderstandings and a lack of root cause. The only CX strategy more dangerous to your customers than no strategy is one that attempts to optimize their experience based on a shallow view of their engagement. These limited views may result in false expectations as they omit perspective from all aspects that impact a customer touch point.
And what are those impacts to each touch point? Twenty years of study have proven to us that the list is complete at four:
- Personnel proficiency
- Business process efficiency
- Appropriate response to customer reaction
- A focus on the desired outcome of the interaction (from the customer’s perspective)
While we live in a world clamoring for simplicity, twenty years of Customer Experience diagnostics have taught those of us at TPG that simplicity must be balanced with comprehensive understanding. If you lose the root cause behind the metric you create, you are destined to develop a standard that lacks corporate support or understanding. It will simply be a numeric value on a dashboard seeking, but missing, its purpose among the corporate KPIs that our Executive Boards leverage to drive business change and success.
If you have ever had an experience where you present a Customer Experience metric that receives no vocal support, a host of voiced challenges surrounding the data collection method rather than the operating gap identified, or a quiet set of blank stares – then you know the feeling of leading a “Quality effort” that is in need of a new metric and overall transformation.
When you have had enough of such an experience, the thought-provoking questions to ask yourself come down to:
- Am I thoroughly tired of presenting data that generates eye-rolls, disbelief and no executive buy-in?
- What REALLY matters when it comes to ensuring that my Customer Experience data is relevant?
These questions are not for the faint-hearted. They will create an introspective moment and a challenge that could involve revamping legacy processes. You have to feel conviction about the future impact that Customer Experience Insight can and should have on your business in order to embrace the transformation that lies ahead.
Measuring Agent Skills and Actions
Identifying the proper elements to measure begins with realigning some previously held quality monitoring beliefs. TPG has found it important to measure customer reaction and skill proficiency as well as agent action. Many times, operational teams measure agent actions, “agent answered phone with proper greeting” but fail to measure the agent skills utilized for such a delivery (i.e., “agent answers without delay” or “agent answers prepared to respond to customer”). The agent may use the choice verbiage but fail to connect properly with the customer. While they have followed a desired procedure, they did not display a behavior that conveys goodwill.
Redesigning the perspective of agent delivery involves separating the ACTION from the SKILL. An agent’s proficiency in a skill is radically different than their ability to adhere to a company process. Such a separation demarks the beginning of a Scorecard that will serve to be customer centric, but is not quite complete.
The completion of such a Customer-Centric Scorecard is the incorporation of two more areas of understanding: Voice of Customer Reaction to the experience as well as the Interaction Outcome.
We consider these ‘4 Quadrants of Learning’™ to be the required structure to transform a Quality Program into an Executive-sponsored Customer Experience Initiative, supporting the needs of both Customer Listening Post and Performance Change Management.
The design of a ‘4 Quadrant Learning’ includes the construct of behavioral measurements that are actionable and clearly identifiable versus a “feel good” instinct of what transpired. Incorporating abstract behavior concepts makes it as difficult to calibrate scoring as to coach improvements. The result will be a metric reported to an Executive Board that experiences unexplained swings without correlation to business performance. While the metric will now have a broader perspective, it will remain unreliable, and therefore, non-adopted.
While calibration can be a subject warranting a dedicated article, and one that most companies would admit to resisting and struggling with, the practice requires a pre-requisite environment that is worthy to discuss (and yes, a future blog about calibration will follow!).
One of the greatest barriers to calibrating CX data, beyond the ability to clearly define behavioral expectations, is the ability to follow a structured process in the collection and analysis phases of measurement. We leverage Six Sigma’s DMAIC principle because it is an internationally recognized standard to allow us to replicate our success across clients. One of our earliest learnings as we embarked on the Six Sigma journey involved sampling. So many companies will sample a percentage of calls with limited understanding of the statistical relevance behind the ratio. Data confidence and confidence intervals became our foundation and the journey to define Customer Experience diagnostics began.
Data Integrity Factor: Sampling parameters
Sampling is the foundational requisite toward Customer Experience transformation. Understanding the difference between a random representative sample and an agent-level sample should be the starting point to decide the purpose of your Customer Experience diagnostic. This decision will determine how you can leverage the metric generated and analyze the learnings collected.
- Representative sample: A random selection of calls to capture the customer’s experience with your company at each lifecycle touchpoint.
- Agent-level sample: A sample of calls by agent that provides equal ‘weighting’ across agents involving the experience delivered. Since all agents do not take the same number of calls with your customer, this method does not pristinely define the experience of your customer, but it does provide an “equal footing” when applying CX metrics to agent incentive compensation.
Either choice is a good one. They key is to understand the path taken. Once the path is chosen, the measurement volume determination follows the same methodology for both. That calculation involves understanding your Data Confidence and Confidence Interval to ensure the statistical validity associated with your data.
Data Integrity Factor: Closed loop process to replicate outcomes
As I mentioned above, we leverage Six Sigma principles as the standard for applying controls to ensure pristine learnings and replicate outcomes. Over the last 18 years, we have refined how we incorporate standards to ensure success for all of our clients. Through the years, the concepts of “Define > Measure > Analyze > Improve > Control” (DMAIC) still apply. Whether you follow the over-arching concepts, or become a student of Six Sigma, the common-sense approach is what ensures you generate much more than a metric on a dashboard. Unfortunately, many people wish for transformation and overnight miracles, without an investment in time or intellectual thought. While we all live in an extremely rushed world, the only path to ensure that your CX Metric resonates with both your Customers and Executive Board is to invest the time to:
- Define your CX objectives in firm business outcome language
- Measure your experience from a ‘4 Quadrant Learning’ perspective
- Analyze the statistical relationship between CX metrics and tangible business outcomes
- Improve the experience through strategic and tactical personnel, product and process change
- Control the environment to quantify the improvements made