Tuesday, August 2, 2016

No Text Analysis = No Voice of the Customer

A bit of a provocative title to this post.  That said....

Having been part of the customer survey business for a number of years, I believed that quantitative customer surveys, and analyses of survey data, could provide a reasonable approximation of the customer's "voice". In particular Net Promoter Surveys did a good job of that. Over time, my opinion on the subject has changed, to the point where I think that quantitative customer surveys cannot any longer be exclusively relied on to approximate the "voice of the customer" (VOC). In my opinion there are two reasons why quantitative approaches don't work as well as they did: The necessity of concise surveys and the increase in alternative non-quantitative channels being used by customers to provide feedback.

"Back in the day", when customers could be relied on to complete questionnaires with fairly large numbers of questions and would do so in numbers large enough to give a reasonable sample, surveys could in fact approximate VOC. Today though, customer surveys have to be VERY concise in order to generate reasonable response rates, and verbatim questions have to be substituted for the questions we used to, but cannot any longer, ask. Customer surveys have become so concise that many businesses today use surveys with a single question and a single verbatim option. With such short surveys I don't think its possible to approximate VOC unless some sort of text analysis is being applied to the verbatim responses being received.

Additionally, social media has been adopted by many businesses (particularly in retail) as a channel for customer interaction. Here again, verbatim comments are the mechanism being used to provide opinions, make complaints, suggest changes, share experiences, etc.. Without using text analysis there's no easy way to aggregate and understand what they are talking about, which topics are trending, which are positive and which are negative.

Needless to say, for people in the VOC space, none of what I'm saying is a revelation. In fact, text analytics has been around for a while now. Customer surveys are shorter today and rely on more verbatim responses than ever before.  Survey response rates are generally lower across the board than they've been in the past. All the conditions exist for a sea change in how businesses ascertain VOC. So, to me, its clear that Text Analysis is something more and more businesses will need to adopt going forward.

For large businesses, with the necessary resources, implementing a combined social media, customer survey and text analytics based VOC process isn't difficult. It costs a lot of money, but ROI is good given the scale of their needs. Smaller businesses though have to consider lower cost and easier-to-implement approaches. Fortunately those exist, often as component based multi-vendor solutions. The key in these situations is easy-to-implement APIs so that customer feedback can flow through various channels into the text analysis tool and then on to customer dashboards or other visualization environments, as needed.

Voice of the Customer today without text analysis isn't representing the customer's voice very well. But, Text Analysis is getting easier to do and lower in cost all the time. Businesses that do not today use Text Analysis in their VOC process should consider doing so.


- Stewart Nash (www.linkedin.com/in/stewartnash)








No comments:

Post a Comment