Maru/edr attended the NGCX conference at the end of March, which included three days of scheduled sessions with CX leaders, senior executives focused on making their organization more customer-centric, and service providers in the CX space.

While a fair number of topics were covered, they could largely be categorized into one of two broad areas:

  1. Best practices in relation to CX feedback programs
  2. Practical implementation of next-generation technology in CX programs

Discussions related to CX feedback program best practices were varied and included matching CX initiatives to company goals, breaking down corporate silos, turning data into insights, and prioritizing CX budgets, among others.

In contrast, topics related to next-generation CX tools and methodologies were narrowly-focused. These conversations largely put the spotlight on natural language processing (NLP) in one form or another – intelligent chatbots, text analytics, intelligent searches, etc. Understandably, NLP is a hot topic and a much-needed service for synthesizing and understanding vast amounts of qualitative data. However, I was surprised about the lack of discussion surrounding predictive analytics in a CX context across much of the presented content.

Customer-centric predictive analytics is possible

In the one presentation that briefly touched on predicting future outcomes, the speaker mentioned that ‘there is a need for organizations to build a CX-oriented culture [to see positive business results], because tying CX to future lagging results is harder to do’. While I do not disagree with the former portion of this statement, I could not disagree more with the latter.

Sure, building a CX-oriented culture can be difficult to do for several reasons – with key ones including gaining executive buy-in and being open to the organizational restructuring needed to serve customers in a more efficient manner.

However, implementing predictive analytics is not difficult, or at least it doesn’t need to be.

A lack of understanding and awareness

Here at Maru/edr, we have robust predictive analytics capabilities in-house. With similar offerings picking up steam with other suppliers as well, the potential for more widespread use of predictive analytics is increasing.

Given its availability, it seems as if the hesitation in discussing predictive analytics stems from the organizational side, rather than the supplier side. Perhaps the perceived complexity of the topic is preventing teams from having open conversations about it altogether.

The solution to this issue is simple. Initiate discussions with analytics and research professionals to understand how predictive analytics work in layperson’s terms, and to learn what data inputs are required and what outcomes can be expected from it. Focus on solutions that are not black boxes, because black boxes transform your data in hidden ways, making it impossible to explain and understand. It’s the simpler solutions that suggest which metrics are driving future predicted outcomes and provide the business with specific, actionable steps forward.

While I believe the aforementioned barriers of education and awareness are likely having the biggest impact, another potential challenge to talking about such analytics is the lack of reliable data, particularly data that is longitudinal in nature (i.e. data from multiple timeframes for the same customers/employees). In conversing with several clients, there seems to be a general distrust of basic data elements stored in databases, like demographics, and it is often preferred to rely on information provided directly from the customer or employee in surveys. Add to that a gated data source that you need to prove your objective’s worth to get to, and you’ve got a bigger issue.

But these are elements that can be solved by persistent conversations with database teams to understand data inputs, including how the different dimensions were collected and how they are defined, and then communicate transparently the need and benefit of tracking multi-year longitudinal data.

How it can benefit you

While CX conferences play an important catalyzing role for driving conversation about what is next in the world of customer experience research and management, in the absence of such representation, it is imperative to start conversations around how to incorporate predictive capabilities into feedback programs.

Predictive analytics can help you:

  1. Identify opportunities where customers are being underserviced, shortening the gap between customer needs and fulfillment.
  2. Focus targeted and more effective save strategies on high probability churn customers.
  3. Understand what drives future positive behavior, which doesn’t necessarily need to be spend, but could also be elements like desired channel usage and interactions that will provide customers with a better experience.
  4. Measure a tangible ROI. Quantify customer experiences in the form of bottom line numbers, an initiative that will give your CX program the internal support you need.

Break the mold and start to talk about how to take your CX program into the next generation of feedback.