5 Lessons from a CX Analyst


5 Lessons from a CX Analyst

As companies continue to recognize the importance of Customer Experience and focus more resources towards improving CX within their organizations, it has become vital that analytics play a big role in helping prioritize efforts.    Having a great measurement program in place to capture customer feedback at key “moments that matter” is only part of the solution.  As wonderful as it is for helping you keep a pulse on performance, real-time monitoring is not able to tell you where your focus needs to be to improve.  Here’s where the importance of analytics comes in.

In the course of helping many clients analyze and act on their CX results, we’ve learned a few things about how to approach analytics within CX.   We’ve included five of these “lessons” or pieces of advice for those wanting to get the most from their CX data.

1. The measurement tool can impact analytics

The most popular way of capturing customer feedback is through surveys.   Directly asking customers how they feel about a particular experience, product, or brand is still one of the best ways to understand how your company is being perceived.  However, in order to gain the best analytic results, there are some key things to consider when designing your program:

  1. Include close-ended questions that are relevant, actionable and are known to influence Customer Satisfaction. To do this, consider piloting potential survey attributes/questions prior to implementing the survey more broadly.   Use key driver modeling to inform which questions will be most relevant to your organization to ensure you are monitoring metrics that truly matter to customer satisfaction.  Keep this list of attributes short and focused.
  2. Representativeness of the survey. Ensure those being invited to take the survey truly represent the population. Not surveying key groups could lead to misleading results. Along these lines, keep survey removals on the back end to a minimum.  It’s never ideal to be asked why customer satisfaction is so high but so is attrition — and have the reason be that your program is not including an important group of customers or removing relevant feedback.
  3. Once designed, pilot any changes made to the survey. Adding or removing questions, changing the wording of an existing question, or changing the scale used could have unforeseen effects to metrics on the survey.  So when opting to make any changes, it is advised to run a side-by-side A/B pilot to measure any impacts.  This will help you disentangle what was a true performance change versus a methodology change.
  4. Keep the survey current and relevant to stakeholders. Analysts are often asked questions from stakeholders that aren’t necessarily captured in the current measurement tool.  “Hot topics” are a good way to probe into certain areas that are top-of-mind and to get additional value from an existing program.

2. Don’t underestimate the value of Text Analytics

Closed-ended survey questions are valuable for helping obtain measurable and quantitative data, which analysts, of course, love.   They provide a way to standardize across respondents and allow companies to understand how customers perceive them on key metrics.  They are easier to understand and track and they provide straightforward ways to build models.  However, they don’t allow respondents to expound on their experiences.  Open-ended questions are going to provide the detail – it’s where respondents will convey their emotions and feelings and truly give their opinions.  This is where analysts go to understand what the closed-ended metrics really mean to a respondent.  To help define “Show a desire to resolve” or “Be knowledgeable”, for example.   And many of the actionable recommendations that can come from a CX program will be informed by what customers talk about when given the chance.    But there is a lot of value in going beyond just using open-ended data as “supplementary”.   Text Analytics provides ways to quantify themes and we’ve found it very useful to use in modeling.  Building simulation models around text categories, volume of mentions, and sentiment has proven to be very insightful for prioritizing focus areas – turning this “qualitative” data into quantitative insights.

3. Cross-program insights are key

Many companies have done a great job standing up their measurement programs to capture every aspect of the customer journey – every transactional touch point a customer might have, feedback about products and services, brand health, end-to-end journey experiences – but this results in an overload of incoming data that can oftentimes lead to more confusion than clarity on what action needs to be taken.  Don’t get us wrong – programs that measure calls to a contact center or digital interactions provide necessary information to individual business leaders on how to improve these specific channels.   And much can be learned at this level.   However, when advising senior leadership on where to be investing efforts, a more comprehensive view needs to be adopted.   For example, will there be a bigger improvement in CX – and ultimately in growth of the company – if more focus is placed on improving the digital Bill Pay experience or improving the Fraud journey for a customer?    To truly understand this, multiple data sources need to be considered given experiences are not limited to a single channel.  Impact models need to consider how customers are affected across all the various places they might have interacted as well as how the relationship with your organization in general is affected.  Where possible, competitive information should be considered to understand how key competitors perform on these experiences.   Data sources outside of customer feedback should also be considered and, on that note, we come to our next “lesson”.

4. CX analytics cannot end with attitudinal feedback only

Not only are cross-program insights key but going beyond the attitudinal feedback to understand customer behavior is a must to truly understand CX.  Going back to the previous example of customers experiencing Bill Pay challenges versus those who went through a Fraud experience, what long-term impact might each of these have to a customer’s loyalty to your company?  And not just stated loyalty – but behaviorally as well.   Are customers who experience Fraud more likely to leave in the next year compared with those who might face Bill Pay issues?   Are they less likely to invest more with you or acquire more products from you?  How will their Lifetime Value change as a result of the experience they had?   Linking customer attitudes and experiences measured by a CX program to what they ultimately do will help in prioritizing what needs the most attention in order to retain or acquire the most business.

5. Keep evolving

The CX landscape is evolving rapidly and CX Analysts need to keep up.    Data is more accessible and “real time” is now the expectation.  Gone is the desire to wait for a “quarterly review” of data.  Analysts are being asked for deep dive insights faster.  CX Analysts need to constantly find new ways to wrangle, mine, and dissect data to uncover meaningful insights – both in the tools used and the methodologies employed.    CX Analysts need to be even more agile and nimble with an ear towards what is most important to their organizations.  They need to be creative in the way they approach data and the way they use it to answer the questions stakeholders are asking.   With more data available, proactivity is crucial.   Communicating what the organization needs to know before they ask has become the expectation.