Share this content

How to use text analytics to completely measure the customer experience

15th Feb 2013
Share this content

Practical advice on how text analytics can form part of a customer strategy dashboard.

Standard approaches to measuring customer experience suffer from flaws introduced by quali-quantification. This is the process of turning something that is qualitative and emotion-based (in this case a customer’s experience) into a quantitative measure (such as a customer satisfaction score) so that it can be compared, aggregated, graphed and tracked. This process of score generation highlights the need to improve but removes the insight necessary for so doing. Or as the Porter in Macbeth put it, “It provokes, and unprovokes; it provokes the desire, but it takes away the performance,” (albeit he was referring to alcohol and not the Net Promoter Score).
Being fair to the Net Promoter Score, it is as good a way of quali-quantifying as any other. Equally many businesses that employ it collect a qualitative explanation of the score given. This is where the value in the process really resides. As web metric guru Avinash Kaushik summarises, “experience analysis allows us to get in to the heads of our customers and gain insight or an ‘a-ha’ about why they do the things they do.”   
The problem is that many businesses still see score generation as the primary focus of customer satisfaction initiatives, particularly if they have a large customer base. In part this is because of their failure to grasp the potential of text analytics software to draw succinct insights from the comments of thousands of customers. 
Any mass market business intent on delivering a superior customer experience should be developing its text analytics capability for mining complaints, voice of the customer feedback and the notes taken by customer service representatives during customer conversations (that would otherwise sit untouched in the CRM system). 
If such an investment is not being made, a key feedback mechanism is not in place. As a result, how serious the business is about delivering an excellent customer experience has to be called into question. That may stick in the craw of some customer experience managers. But if you are not rigorously assessing the qualitative explanations of what people like or don’t like about the service you provide, your ability to delight and differentiate is highly compromised. 
Such tools can also be used to measure the customer experience as a complement to any formal score capture. Firstly this requires translating the vision for the experience into the words that customers would use to describe the service they had received if that vision had been met. The second stage is then tracking how many customers use those words (or their synonyms) when describing the service they have received versus how many use antonyms. 
Such a measure would form part of a comprehensive customer strategy dashboard. This would have three layers (See Figure 1 below).

  1. The first layer comprises metrics that track the service levels that the business is delivering (e.g. first time call resolution, on-time deliveries, etc.). These measures cover what the business has control over. Being inward facing they tend track performance on potential drivers of customer dissatisfaction – those that cause customers problems if they go wrong – rather than drivers of delight (e.g. the degree of empathy shown when resolving a problem), which are difficult to measure using process controls. Ultimately these process metrics help a business get the basics right (at least most of the time), providing the foundation for a differentiated experience.
  2. The second layer tracks customers’ reactions to the experience received. This would include the formal customer satisfaction or Net Promoter measurement process along with the word-based scoring approach using text analytics outlined above. This layer tracks the emotional response to the service provided, helping to identify both moments of truth (typically interactions where there is significant emotion invested in the outcome) and how best to deal with such instances. 
  3. The third layer then looks at how the combination of performance in the first two layers translates into outcomes for the business. This layer covers how the customer experience being delivered translates into performance on key corporate objectives – revenue and profitability (in aggregate and per customer); brand recognition and preference scores; total customers and customer acquisition, retention and cross-sell rates.
A customer strategy dashboard of this type enables a business to hypothesise cause and effect relationships between variables in different layers, define how each will be measured, track correlations and pro-actively experiment to determine how to enhance the customer experience in ways that most enhance business performance.
Jack Springman is head of customer analytics for Sopra Group in the UK. He can be contacted at [email protected]

Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.