Why brands need to capture deeper customer insight from unstructured data

Why brands need to capture deeper customer insight from unstructured data

Published on: April 10, 2018
Author: Taoufik Massoussi - Product Manager & Head of AI

Understanding customer needs and emotions is central to delivering the experience that today’s consumers demand. However, most current Voice of the Customer (VoC) programs focus on structured data, from sources controlled by the company, such as feedback surveys sent out to consumers. 

This is ignoring a goldmine of information. Gartner estimates that 80% of an enterprise’s data is unstructured, living in emails, social media posts or other documents. That means that most VoC programs are missing out on the full picture – essentially businesses are trying to understand their customers using just 20% of the available information. And the data they do use is normally quantitative rather than qualitative – perfect if you want to track metrics such as Net Promoter Score (NPS), but less useful if you want to find out why a customer thinks, feels or acts in a particular way. Essentially, brands are getting a partial view that is internally focused, rather than customer-led.

Many organisations understand that this is a drawback, but until now, have been held back from fixing it due to three key issues:

  1. It is difficult to access unstructured data as it spread across multiple systems, such as email, and on platforms that may be out of a brand’s direct control, such as social media.
  2. Analysing unstructured data is resource-intensive to do manually, particularly at scale.
  3. Drawing conclusions from unstructured, qualitative data requires specialist skills, is time-consuming to do and easy to get wrong.

The good news is that this is now changing, thanks to new technologies:

  • Text analytics, which automates the process of analyzing unstructured data to enable it to be better understood.
  • Natural Language Processing, which applies semantic analysis to understand the context of the message, and the sentiment within it. For example, it can understand the difference between words that have multiple meanings (e.g. a bat is both a flying mammal and something you play sports with) by looking at the wider context.
  • Artificial intelligence, which can automatically group together relevant statements to provide a holistic view of the themes and topics that customers are talking about.

Once this analysis has taken place, you have access to a much more complete picture. This customer intelligence goes beyond simple insight (the who, what, when and where) to unlock the why – why is the customer behaving in this way? Why am I suffering increased churn?

By capturing intelligence that was previously inaccessible, brands can then identify problems or opportunities and take steps to implement solutions. This not only improves customer experience and opens up potential new revenue streams, but it also demonstrates to customers that you are listening to them – whatever channel they are using, and without them having to fill in feedback surveys. It shows you are being open and transparent and putting the customer first, while capturing the voice of consumers as they interact with you.

Given the pressures that brands currently face due to greater competition and increased demands from customers,
it is vital that their customer experience programs are based on a complete picture. That’s why they need to ensure that they are analyzing unstructured as well as structured data if they are to deliver real customer intelligence that drives results.

Tags: vecko, unstructured, structured, data, customer intelligence, insight, Net Promoter Score, NPS, Customer Service, Customer experience, AI, Artificial intelligence, NLP, Natural Language Processing, text analytics
Categories: Best Practice, Product

You might also be interested in these posts: