Etuma Blog

Lists by Topic

see all

Subscribe to Email Updates

Net Promoter System: How to Extract the Root Cause for a NPS Score Change

[fa icon="calendar"] Mar 16, 2015 11:24:00 AM / by Matti Airas

Is it possible to hear the true voice of the customer, if the most genuine form of customer feedback -open text comment - is turned into statistical information?

Statistical information might be valuable when pinpointing problem areas but it doesn't reflect the nuances of human interaction. It doesn't tell the reason - the why . The process of finding the reason is called root-cause analysis.

etuma_demo_carts.png

CASE: Problems with CARTS 

There are 6306 customer comments in this grocery store chain Transactional Net Promoter Score (TNPS) system dataset.

The NPS score of comments about CARTS is -46 during the week of July 25, 2015. This is quite a bit lower than the average NPS score accross the whole feedback process (31). This is a clear signal that something is wrong with CARTS during the week of July 25. 

Filtering by topic CARTS and time (week of July 25), you can narrow the  analysis down to the 28 relevant comments. Read those comments and you will find out that the reason. In this case the main reason is that the staff is leaving stacking carts on the aisles during the business hours. 

Now, instead of reading 6306 comments and sorting out the ones in which customers talk about CARTS you only need to read 28 comments to find out the reason for the deviation in the NPS score.


When customers' open-ended comments  are categorized and sentiment is detected on topic level, it is easy to drill down to topic and sentiment specific sentences and find out the root-cause for a  problem (or an opportunity).


Want to become an expert in feedback analytics?

WEBINAR VIDEO   

Topics: Feedback Analysis, Customer Experience Management, customer loyalty, NPS, root cause analysis, retail cx, loyalty management, transactional NPS

Matti Airas

Written by Matti Airas

My passion is to figure out how to turn open-text feedback into well structured usable information.